Arimaa Forum (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi)
Arimaa >> Events >> 2015 Arimaa Challenge Match results invalid
(Message started by: omar on Apr 29th, 2015, 1:11pm)

Title: 2015 Arimaa Challenge Match results invalid
Post by omar on Apr 29th, 2015, 1:11pm
It has come to my attention that side bets have been placed on the outcome of the 2015 Arimaa Challenge Match by one of the challenge defenders, the developer of the program trying to win the challenge and one of the sponsors of the challenge prize. I'm sorry I have not had the time to follow all the postings in the forum and chatroom to be aware of this sooner. This was brought to my attention yesterday by an anonymous member of the Arimaa community. I am highly disappointed and disturbed to learn about this.

The specific details are as follows:

browni3141 - one of the three challenge match defenders.
lightvector - developer of the program Sharp trying to win the challenge prize.
Fritzlein - has pledged to award $1000 USD to the developer who wins the challenge prize.
omar - founder of Arimaa and the Arimaa Challenge Match; has pledged to award $10,000 USD to the developer who wins the challenge prize.

Fritzlein posted the following forum message (Re: 2015 Arimaa Challenge; Reply #11 on: 04. Mar 2015 at 22:57)
"In the chat room, browni3141 and I negotiated for him to sell me some insurance.  I am on the hook for $1000 if the Arimaa Challenge is won this year, and I am worried about having to pay out to lightvector, but browni3141 rates the odds of that happening as very low.  (Exactly how low, browni?  Less than 100:1 for sure, since you need a premium to justify taking on increased variance.)

In any event, I have sent 200 Arimaa points to browni3141, and he has promised to give me all of his prize money from this year's World Championship if the Challenge is won this year. If he wins the World Championship, that is 23994 Arimaa points of insurance, i.e. almost a quarter of my liability.  Thanks, browni!"

FYI: 200 Arimaa points = $2 USD.

lightvectory and browni3141 discussed the following in the chatroom:
2015-04-12 06:16:21 lightvector: browni, if sharp wins the challenge this year, I'll send you your WC money, so don't worry about it and just play a good game
2015-04-12 06:16:53 browni3141: My ego hurts more than my wallet.
2015-04-12 06:20:40 browni3141: I suppose I could grudgingly accept due from anybody but Fritzlein who I would flat out refuse

1. browni3141 - as a challenge match defender you cannot be involved in any sort of bets on the outcome of the challenge match. This is not explicitly stated in the challenge rules, but it is an unwritten rule that is expected as part of ethics. When I write the rules of Arimaa I don't need to explicitly write that you are not allowed to swipe or misplace an opponents piece while they are not looking; it is understood.
2. Fritzlein - it would have been fine if you placed such a bet privately with anyone else but one of the defenders.
3. lightvector - I see no reason for you to be involved at all. If I were to ask you "If the challenge match is won by your program will you be making a payment to any of the challenge match defenders?" your answer would have be "Yes". This is not good. Perhaps you had good intentions, but being the developer of the bot that is trying to win the challenge prize you need to steer clear of any such involvement.  Again this is not explicitly stated in the challenge rules, but it is an unwritten rule that is expected as part of ethics.

It is clear from browni3141's comments in chat that he regretted having made the bet with Fritzlein since it caused additional pressure for him to win. Although the challenge defenders receive no compensation other than glory for defending the challenge, they are never penalized monetarily for losing. The bet with Fritzlein changed this situation.

It is clear from browni3141's comments in chat that after lightvector offered to cover him if the bot won the challenge prize, he felt less concerned about winning his challenge match series as long as he would get a chance later to show that he is better than Sharp. In fact after the challenge match was over browni3141 has asked me to make Sharp available on the same computer that was used for the challenge match.

I don't think any one of the people involved had bad intentions, but their actions have caused the challenge match to be influenced in ways that were not intended. We go to great extents in the Arimaa community to run fair, clean and unbiased events and we need to preserve that. As such the result of the 2015 Arimaa Challenge Match are declared to be invalid and the challenge match will continue in 2016.

This may give negative publicity to Arimaa and outsiders not familiar with all the details may think that I am trying to avoid having to pay the challenge prize. But I am not phased by this. In the long run they will see that I will pay it when it is won without any issues. Had it not been for these side bets I was prepared to pay the challenge prize and had even planned to announce two new challenges on Saturday. Regardless of this announcement lightvector and I have some exciting announcements to make on Saturday about the future of Arimaa.

http://arimaa.com/arimaa/radio/


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by clojure on Apr 29th, 2015, 1:50pm
I haven't been much following Arimaa lately but I must express that after hearing the good news about Arimaa challenge finally being conquered, it's disappointing and very unexpected to hear such news.

I'm not going to take a stance on the decision but I feel for you Omar, I really do. Good to hear you have already plans for the future.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on Apr 29th, 2015, 3:53pm
Thank you clojure. Yes, this was an extremely hard decision to make. After I was informed of what had happened, I did consider just taking the easy route and ignoring it. But I could not be content with such a decision. I realize that I am putting the reputation of Arimaa and my own reputation at steak by taking this decision. In the long run I think this is the right thing to do. Based on how things have gone this year with the challenge match I'm about 95% sure that the challenge will be claimed before 2020. It will give me a chance to prove that I do stand by my word.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on Apr 29th, 2015, 4:16pm
I've already said this in the chatroom, so this will just be repeated for some of you, but I feel I should post here for posterity.

From a mathematical standpoint none of the bets should cause me to perform worse in the Challenge. My initial bet with Fritzlein was in favor of humans, so losing is not incentivized. Lightvector's offer makes this initial bet null, so both winning and losing my mini-match is $0 expectation, therefore losing is not incentivized.

It could be speculated that I was psychologically influenced by these bets, but I do not think that is the case. In the past I have bet and won/lost hundreds of dollars in single hands of blackjack or poker. I view money much more in terms of expectation than results, therefore I was a little frustrated with myself that I made an apparently bad bet, but I am comfortable with losing money on a good bet. I regretted the decision I made with Fritzlein only because I believed that I gifted him expectation after I got more information about sharp, not because I was worried about losing money.

Some of what I said in chat was a way of dealing with the psychological pressure from attempting to defend the Challenge after so many other good players had lost to sharp before me and from the high stakes other people had in the event (not myself!) While I am not results oriented about my own monetary gains/losses, I feel pressure when other people's money is on the line and I'm expected to win.

Saying that having sharp available after the Challenge made me less concerned about winning was also a psychological trick for me to deal with the pressure. I don't like to admit this, but I was highly concerned with the fact that a bot could defeat me in a match the year I won my WC. It may seem silly and unimportant to others, but I take great pride in being the best at something, and having that taken away a week after I got it is extremely disheartening. I was less concerned about losing my Challenge games if I still had a chance to beat sharp later. Realizing this helped me with the pressure. These have no relation to any of the bets made anyway.

None of the circumstances caused me to give anything but the greatest effort I could muster in the Challenge games. I treated these games as seriously, if not more seriously, than my WC games due to others' stakes involved, and didn't deviate from my own ethical principles at any point.

Although I feel I have done no wrong by my own principles, I am very sorry to lightvector for causing you harm by my actions, and sorry to omar for putting you in a tough spot by compromising your principles.

I still strongly disagree with the decision, and wish you to reconsider.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lightvector on Apr 29th, 2015, 4:52pm
Omar, I am sorry, and can only imagine the surprise and shock you must have felt upon learning this. I did not consider how my action would shine in the light of a clear, principled stand such as the one that you've expressed, and I apologize. I realize now that I made a mistake in not communicating the situation to you.

Indeed, it was not my intention to influence the incentives around the outcome of the challenge adversely. Exactly the opposite - by refunding Mathew Brown, I felt that I would both be helping him out from a situation he greatly regretted and that was negatively impacting him, while also restoring the situation back to the status quo. That is, I would be transforming it back to a situation in which none of the defenders would have any monetary stake in the matter and where for all practical purposes I would simply be pledging to return a small portion of Fritzlein's contribution to the challenge back to Fritzlein in the event that my bot won.

To be honest, I care little for the challenge prize money itself, and feel lucky to be in a position in life where I *can* care little for it. I therefore felt unique in my position to be able to contribute in that situation, given that in the event of a challenge win, I alone would be receiving a windfall for which I would have no particular use, while Mathew would suffer a loss that according to him would significantly impact him. What I stand for is for the fun and enjoyment of the people in this community and the shared discovery of this game. From that stand, I would do the same thing again in that situation if there were no other way to dissolve the agreement in a way acceptable to both parties.

And regardless of what happens now, I continue to stand in support of the the fun and health of the community. As I've speculatively expressed to you before privately, regardless of whether this year or in a future year if I win the challenge I intend to donate the prize back to the community.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on Apr 29th, 2015, 6:05pm
Thank you browni3141 and lightvector for expressing your perspectives. As I mentioned in the chat my decision here is not based on game theory but rather on principle. Everyone who is familiar with the Arimaa community know that we go to great extents to make sure that our events are run fair, clean and unbiased. We have implemented features such as delayed relaying of event games to spectators, resuming of games that encounter problems, deep discussion of formats, seeding, etc. We do all this even though in the grand scale of things almost no one is watching or cares. We go to this level of rigor because Arimaa is important to us and we care about this grand experiment to see how long humans can stay ahead of computers using only a chess set. We should be proud and thankful that we've have lasted almost 18 years beyond chess.  That could never have happened if I was the only one defending the challenge. I am very lucky and thankful that other players much stronger than myself found interest in Arimaa and glory in defending the challenge. The games played by the challenge defenders are priceless and there is no monetary value I can put on them to compensate what they are doing. That is why it's ever so important the the defenders play only for glory and nothing else. We have maintained this clean slate for all these years. So as a matter of principle I cannot be content knowing that bets were placed on the outcome by key participants. We all make mistakes and we've all done things that have had unintended consequences. I hope no one takes this personally. I have had good long term working relations with you guys and hope to continue that in the future. I hope this turbulent phase will come to pass and years from now we can look back on it and say that in principle we did the right thing.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Heyckie on Apr 29th, 2015, 8:05pm
In order to clear others from suspicion I'll make it clear that the "anonymous member" was me.

The reason for being anonymous and not bringing this up in Chat or Forum was that I wished to cause no stirrup in case Omar would take no action. I was there when the bets and deals were made and at the time I saw nothing wrong, but subsequently one of them slowly started to disturb me somewhat. Then just a couple of days ago I realized that Omar possibly wasn't aware of these. At that point, after thinking about it for a shorter while than I would have liked, I came to the conclusion that I had to speak my mind at least to Omar as soon as possible or forever stay quiet and risk having to keep questioning myself whether I should have spoken after all, especially in the case someone else would think about this later and lament nothing was done about it until it was absolutely too late.

For what it's worth I really see no serious wrongdoing from neither browni3141 nor Fritzlein; browni was originally betting for himself and then accepted a deal in which there was no possible downside for him and clearly stated he didn't ask for it. Fritz just made a normal bet.

What bugged me was lightvector essentially paying the opponent of his bot should he lose. I have no doubt about his intentions being bona fide, but for some total outsider his behavior would seem similar as if say Roman Abramovich were to publicly announce to pay Glenn Murray whatever his current goal bonus is should he fail to score a winning or equalizing goal in the Crystal Palace-Chelsea match next Sunday.

I hope this will not harm the great community around a great game.
Heyckie

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Fritzlein on Apr 29th, 2015, 10:12pm
I strongly disagree with Omar's decision to invalidate the Challenge.

First, as browni3141 stated, he never had any monetary incentive to lose.  Having lost, he receives no compensation.

Second, if having money at stake in a tournament setting invalidates the results of games because of the addition of psychological pressure, then every Arimaa World Championship to date is invalid due to the psychological pressure created by the prize money which has been available every year, and all future Arimaa World Championships must be played for glory only (without prize money) so as not to distort the results.

Third, Omar can hardly call it "an unwritten rule that is expected as part of ethics" that a player can't bet on himself, given that Omar himself sanctioned it for six years.  From 2005 through 2009, participants in the Arimaa World Championship were allowed to bet on themselves to win their games in the parallel Spectator Contest, and they were indeed awarded prizes for having correctly bet on themselves if they won.

Fourth, while there is a solid precedent for players being allowed to bet on themselves in the Arimaa gameroom, there is no precedent for players being disallowed from betting on themselves, including in the 2014 Arimaa Challenge in which supersamu, one of the Challenge Defenders, publicly bet on himself to sweep his three games.  When an activity has been officially sanctioned in the past and publicly tolerated up to the present, it is far fetched to invoke an "unwritten rule" against it.

Fifth, Omar himself has a clear conflict of interest in deciding whether the conditions of the Arimaa Challenge have been met, or whether external circumstances have arisen which would invalidate the result.  The Match Director of the event should not be a participant; he should be a disinterested third party.  This logic seemed clear to both Omar and myself when we agreed that I should not participate in the 2013 Arimaa World Championship because I was the Tournament Director.  Indeed, Omar himself intended to appoint a disinterested party for the 2015 Arimaa Challenge as well, as he said in the chat room:


Quote:
2015-04-16 11:12:28 omar Actually I would like Janzert to be the TD, but I forgot to ask him.


If Omar had publicly expressed his concerns upon discovering the financial arrangments that had transpired alongside the Arimaa Challenge, and he had belatedly appointed Janzert to adjudicate the matter, I would have presented my first four agruments for Janzert to take under consideration and would have then bowed to Janzert's decision.  Instead, Omar chose to act unilaterally.  Of the potentially dubious actions taken by the four interested parties (Omar, browni3141, lightvector, and myself), I believe that Omar's decision not to defer to a neutral party is the most dubious.

Omar has said in chat that he doesn't think anyone acted in bad faith, but he nevertheless couldn't in good conscience let the result of the Arimaa Challenge stand.  Likewise, I don't think Omar has acted in bad faith, but I can't in good conscience abide by his unilateral decision.  In my judgment, sharp won the Challenge fair and square, and I will therefore pay David Wu the $1000 I pledged to the prize fund.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by aaaa on Apr 29th, 2015, 10:42pm

on 04/29/15 at 22:12:55, Fritzlein wrote:
In my judgment, sharp won the Challenge fair and square, and I will therefore pay David Wu the $1000 I pledged to the prize fund.

Is that not even conditional on him submitting a paper to the ICGA?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Boo on Apr 30th, 2015, 2:08am

Quote:
What bugged me was lightvector essentially paying the opponent of his bot should he lose. I have no doubt about his intentions being bona fide, but for some total outsider his behavior would seem similar as if say Roman Abramovich were to publicly announce to pay Glenn Murray whatever his current goal bonus is should he fail to score a winning or equalizing goal in the Crystal Palace-Chelsea match next Sunday.


This is bad comparison, because it doesn't take into account that browni3141 is -236$ if he loses. Compensation from lightvector doesn't give browni any incentive to lose, only takes off the burden of possible loss of 236$. Essentially this only completes the debt-loop: Fritzlein gives 1000$ to lightvector, lightvector gives 236$ to browni3141, browni3141 gives 236$ back to Fritzlein. Fritzlein wanted insurance, and lightvector says "OK, let it be not 1000$, but 764$". Browni3141 acts only as intermediary here. I see nothing bad in it.

What omar has done is total absurd. Omar penalizes the guy (lightvector) who restored his own condition "challenge defenders receive no compensation other than glory for defending the challenge, they are never penalized monetarily for losing". If there is anybody to be penalized, this is Fritzlein, who disregarded this condition. (The question is if Fritzlein could know of condition's existance)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by jdb on Apr 30th, 2015, 2:26am
The posts of Fritzlein and Boo summarize the facts nicely.

Unbelievable decision by Omar.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by novacat on Apr 30th, 2015, 3:01am
I am glad to see people standing up for what they believe is right, and I have a lot of respect for all involved.  I feel I must point a couple things out though.


on 04/29/15 at 13:11:29, omar wrote:
Although the challenge defenders receive no compensation other than glory for defending the challenge, they are never penalized monetarily for losing. The bet with Fritzlein changed this situation.

I am afraid this is not true.  Fritzlein was a defender last year and in 2011 and stood to lose $1000 in the prize fund.  Omar has been a defender several times and could have lost $10000 by losing his games.  Browni agreeing to be responsible for a part of the prize fund is not new, even if he got paid $2 to do so.


on 04/29/15 at 13:11:29, omar wrote:
It is clear from browni3141's comments in chat that after lightvector offered to cover him if the bot won the challenge prize, he felt less concerned about winning his challenge match series as long as he would get a chance later to show that he is better than Sharp. In fact after the challenge match was over browni3141 has asked me to make Sharp available on the same computer that was used for the challenge match.

Browni expressed his desire to show he was better than Sharp over a longer match before lightvector offered anything.

2015-04-11 22:46:54 browni3141 I still think I'm objectively better than sharp, and will request that it go up after the Challenge so I can do a longer match against it.

After accepting lightvector's offer, browni does make a comment about how he wants to be the best player and beating Sharp in a longer match would make him feel better, but it was immediately preceded by his comment that losing the Challenge would feel like a personal failure (2015-04-11 23:44:43).  I do not get the impression at all that he felt less concerned about winning.

(Edit: Note chatroom timestamps are in EDT)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Heyckie on Apr 30th, 2015, 3:08am

on 04/30/15 at 02:08:30, Boo wrote:
This is bad comparison, because it doesn't take into account that browni3141 is -236$ if he loses. Compensation from lightvector doesn't give browni any incentive to lose, only takes off the burden of possible loss of 236$.


Well I'm sorry it isn't perfect, it was just the closest analogy I could quickly think of. lightvector's offer didn't give browni any incentive to lose, but neither does my example give Mr. Murray any incentive not to score decisive goals (actually he still would have some incentive to score even if you don't think about the team as he could score more goals after the winner). It just took out the monetary incentive to win, of course that is not nearly as serious as actual match fixing but still somewhat questionable.

I've spent more than my two cents here. I don't want to be a judge or even a juror here, I just wanted to explain why I felt like I should let Omar know.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on Apr 30th, 2015, 4:35am
I think the Challenge has been won fairly by bot_sharp and the prize money should be payed to David.

To me, voiding this year's challenge is the worst decision for arimaa. From second-best to second-worst, here are some other possibilities:

- Belatedly appoint a Tournament Director and let him make a final decision. Janzert would be a prime candidate, and we would all happily abide by his ruling, no matter what. Even if he made the same decision as Omar, the arimaa community would, I think, be much happier about it coming from a respected neutral 3rd party.

- Declare browni's mini-match void and make him replay it. [Harvey's and mine weren't influenced by the bet in any manner. It would be unfair for sharp not to have a chance at the challenge this year in spite of qualifying.]

As for the bet itself, I'd say that browni's bet with Fritz is fine given that it gives no incentive to browni to change any decision he makes during the challenge defense, and that no other bet was made on the challenge (contrary to a situation in http://joshuaspodek.com/non-judgmental-ethics-sunday-athlete-wants-bet).

In my eyes, the only relevant part in Omar's argument is lightvector's pledge to refund browni in case of a loss. Although I'm 100% sure that David meant well by making this pledge, reducing the incentive of your opponent to beat you is going to be controversial. I don't see this as warranting a cancellation of this year's challenge though.

Jean

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on Apr 30th, 2015, 7:11am

on 04/30/15 at 04:35:43, chessandgo wrote:
- Declare browni's mini-match void and make him replay it. [Harvey's and mine weren't influenced by the bet in any manner. It would be unfair for sharp not to have a chance at the challenge this year in spite of qualifying.]


Great idea. But rather than lightvector compensating browni3141 if he loses, I will do that. This solves a lot of problems. Fritzlein gets his insurance, browni3141 has no external pressure on how he performs and lightvector is not involved. I also like that this whole matter will be decided over the board rather then anyone's judgement. I've already rented the server for another month, so if browni3141 is up for it, we can move forward with it.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on Apr 30th, 2015, 7:14am

on 04/30/15 at 04:35:43, chessandgo wrote:
- Declare browni's mini-match void and make him replay it. [Harvey's and mine weren't influenced by the bet in any manner. It would be unfair for sharp not to have a chance at the challenge this year in spite of qualifying.]


Great idea. But rather than lightvector compensating browni3141 if he loses, I will do that. This solves a lot of problems. Fritzlein gets his insurance, browni3141 has no external pressure on how he performs and lightvector is not involved. I also like that this whole matter will be decided over the board rather then anyone's judgement. I've already rented the server for another month, so if browni3141 is up for it, we can move forward with it.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by robinz on Apr 30th, 2015, 2:25pm
I had no idea about this. I don't even play Arimaa actively any more, and haven't followed the Challenge (but was aware of the result because I visit this forum on a roughly weekly basis) - I only found out about this decision because supersamu posted a link on BoardGameGeek (a site on which I am much more active than on here).

So, while I accept that, since I have not had any involvement in the Challenge this year, it is fair to say that I do not have full awareness of the facts, I think my opinion as an "outsider" (but one who loves the game of Arimaa and would like it to have good publicity) might be valuable. And I agree 100% with Fritzlein - I read through all of Omar's opening post and could not see that anything he stated would make it reasonable to invalidate the Challenge result. And if what Fritzlein says about Omar having sanctioned side bets by Challenge defenders in previous years is true (I would not know this myself, but have no reason to doubt his recollections), then that would to me make Omar's decision completely indefensible. I am surprised that he has since left a comment on this thread without addressing this.

And it is worth adding that I completely respect Omar's right to make this decision, as the game's inventor and as the man who is ultimately in charge of the Arimaa Challenge (and indeed all things related to this website). And I am totally willing to accept that his decision was made on grounds of principle (however flawed I think they might be in this instance), and not because he is trying to find a way to avoid paying the $10,000. However, I think plenty of outsiders, on hearing what has happened here, will reach this conclusion - and I must admit that it is not easy to argue against them.

I don't want to get involved in a big debate here, I just felt strongly enough to give my own thoughts. Fritzlein, chessandgo, and others, have already expressed them, with much greater knowledge and greater authority in terms of being longtime (and very strong) Arimaa players who have been involved in the Challenge for many years. But I thought perhaps my "outsider"'s perspective may also be welcome. If it is not, then please just ignore me and continue the debate :)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by kzb52 on Apr 30th, 2015, 3:30pm
Assuming that browni3141's challenge games are hopelessly tainted, and something has to be done about them, I have a quick thought to share:


I don't think browni can replay his challenge games in good faith; he has a serious conflict of interest.  He has said that he feels the current challenge results (sharp winning) are valid.  And so it's possible that a desire to get the "correct" result could lead him to play less than his best in a redo.  I'm not saying browni would even consider throwing the games, but his personal feelings about the challenge are a real incentive.  I think it's a bad idea to put anyone in a position where they have that kind of incentive to lose, even if it's just a possibility.  We need a different defender, somebody who hasn't stood up and said "sharp deserves to win the challenge."  If we care about ethical principles, than we need to be certain the human defender is in a position where they want to try their best and nothing more.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by SilverMitt on Apr 30th, 2015, 5:02pm
I see no point in replaying a mini-match.  I would much rather just see browni3141 play more games against sharp.  They can be event games with a delay, if that's what everyone would like, but their outcome should not decide the Arimaa Challenge.

Some other random thoughts:
  • The side bet between browni3141 and Fritzlein did not seem serious at the time, at least to me, since Fritzlein's bet was only 200 Arimaa points, and browni3141 essentially responded with a value of his own choosing.
  • While we may have expected chessandgo or harvestsnow to have a reasonable chance of winning their mini-matches, I think we would all say that if browni3141 lost his mini-match, the challenge was likely to fall.  In other words, these events were not independent, but all conditional on sharp's strength, about which we still had a lot of uncertainty.
  • It is contradictory to argue that browni3141 was adversely affected both by believing he had money at risk and by it no longer being so.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on Apr 30th, 2015, 6:53pm
Regarding the future:

The easiest way for this mess to have been avoided was if Omar, and the Arimaa community in general, had been aware of the bets when they were made. Unfortunately, the bets were negotiated in the chatroom, which, despite being archived, is largely ephemeral. No one can be expected to comb through the whole thing looking for unsavoury bets.

To prevent a situation like this from occurring in the future, I think we should create a forum thread called “Bets”. No bet on Arimaa would be valid unless each bettor agrees to it in this thread. This way, no potentially objectionable bet could get buried in the avalanche of the chatroom archive. Anyone would be welcome to read the thread and object to any bet they believe is illegal or unsavoury.

If this system had been in place for Fritzlein's, Browni's, and LightVector's bets, Omar would have been able to nix them long before Browni played his Challenge games, and everything would be fine. I don't mean to lament what could have been. I only want to prevent a situation like this from happening again.



Regarding the present:

In my view, two ethically dubious events have occurred:

A) LightVector offered to compensate Browni if he lost to LightVector's bot.

B) The invalidation of the Challenge Match results was not done by a disinterested third party.

Revalidating the Challenge Match results would solve B, but not A. I'm confident that LightVector's intention was wholly virtuous, but that does not make his action virtuous. I agree with Omar when he says, “If I were to ask you [LightVector] 'If the challenge match is won by your program will you be making a payment to any of the challenge match defenders?' your answer would have be 'Yes'. This is not good.”

The only proposal that solves both A and B is Browni replaying his games against Sharp and Omar taking LightVector's position in his bet with Browni. LightVector would not compensate anyone for losing to his bot. Browni would have no incentive other than glory and pride. Sharp would have a fair opportunity to win the Challenge. And the Challenge would be decided on the board, not by someone who has a sizable financial stake in the outcome.

Perhaps, as Kzb said, Browni's view that the orignial Challenge Match results are valid would incentivise him to lose. This would be unfortunate, but I'm not sure it's true. Instead of speculating how Browni would feel about a rematch, I'll wait for him to let us know.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 6:31am
First of all I want to say that I am not aware of supersamu being involved in any bets relating to the Arimaa Challenge.

It is easier for most people to see the ethical mistake lightvector made by trying to cover browni3141's possible loss. The ethical mistakes made by Fritzlein and browni3141 are more subtle.

When Fritzlein placed the bet with browni3141 to reduce his own risk he did it at the expense of increasing the risk of the other challenge match sponsors. I would argue that he increased the physiological  stress on browni3141 causing him to lose his 1st round game (when lightvector had not yet been involved). If browni3141 had won that first game the challenge match would not have been lost. After lightvector relieved the stress browni3141 performed noticeably better. I could argue that based on this Fritzlein should compensate the other challenge match sponsors for the damage caused.

When I selected browni3141 to be a challenge match defender I placed him in a position of power. Similar to an elected official or government employee. He should not use that position to benefit himself. So even if he takes bets with him on the winning side he is essentially trying to benefit himself using the special position. If rather than being approached and accepting the bet what if browni3141 had solicited bets that are seemingly legitimate (i.e. with him on the winning side). Surely that would appear wrong. Is it only wrong if he solicits bets and not wrong if he accepts unsolicited bets.

So I do believe I have a legitimate reason for invalidating the challenge match and at the very least the three games played by browni3141.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 7:04am
I just want to stress again that I don't want Fritzlein, browni3141 or lightvector to take this personally and I completely understand that you meant no harm to anyone or the challenge.

BTW: I was a government employee for 10 years and had to go through ethics training many times :-)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by jdb on May 1st, 2015, 7:26am

on 05/01/15 at 06:31:33, omar wrote:
So I do believe I have a legitimate reason for invalidating the challenge match and at the very least the three games played by browni3141.


This is a clear conflict of interest. You have a 10K stake in the decision.

A neutral third party is the standard way to handle this situation.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 7:27am
In the future the challenge match defenders will not be revealed. They will play their games under anonymous accounts like "defender1". They will only be revealed after the match is over. I don't believe this changes the goal post in any significant way, but will prevent complications such as this in the future.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Belteshazzar on May 1st, 2015, 7:59am

on 05/01/15 at 07:27:32, omar wrote:
In the future the challenge match defenders will not be revealed. They will play their games under anonymous accounts like "defender1". They will only be revealed after the match is over.

Interesting.  The defenders will have to be careful not to divulge who they are.  I imagine there would be a fair amount of speculation on that front, although I guess you could forbid that.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 8:13am
A neutral 3rd party is needed if a judgement is required. No judgement is required in this case.

Does anyone dispute that these bets were placed.

Does anyone dispute that we expect key participants to follow ethical standards.

Does anyone dispute that there is a problem with the game if the participant of the game did not follow ethical standards.

Does anyone dispute that a game should be invalidated if there is a problem with the game. The challenge match rules allow for this.

The only thing that could be questioned is if the participants knew if they were breaking ethical standards or not.

Would what I have presented be correct if I only had $1 at steak, but not be correct because I have $10K. Where do you stand on what I have presented if you did not know how much I have at steak.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 8:26am
It does impose an extra burden on the defenders to conceal that information. Fritzlein and I discussed this once several years ago and he convinced me that it was not worth the extra effort and it might make finding defenders harder. I now feel that it is worth the extra effort, although I don't like that maybe some potential defenders might decline the offer because of it.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by rbarreira on May 1st, 2015, 9:02am

on 05/01/15 at 06:31:33, omar wrote:
I would argue that he increased the physiological  stress on browni3141 causing him to lose his 1st round game



on 05/01/15 at 06:31:33, omar wrote:
I could argue that based on this Fritzlein should compensate the other challenge match sponsors for the damage caused.


I just want to say that despite the fact that I see your point with regards to lightvector's offer to browni3141 (which, in hindsight, was misguided even if with good intentions), I think these two statements are too strong and harsh IMO.


on 05/01/15 at 08:13:14, omar wrote:
A neutral 3rd party is needed if a judgement is required. No judgement is required in this case.


And yet a judgment was made! I'm not saying I necessarily disagree with it (I'm on the fence), but I see the points of those who say you didn't handle this in the best possible way.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on May 1st, 2015, 9:04am

on 05/01/15 at 06:31:33, omar wrote:
When Fritzlein placed the bet with browni3141 to reduce his own risk he did it at the expense of increasing the risk of the other challenge match sponsors. I would argue that he increased the physiological  stress on browni3141 causing him to lose his 1st round game (when lightvector had not yet been involved). If browni3141 had won that first game the challenge match would not have been lost. After lightvector relieved the stress browni3141 performed noticeably better. I could argue that based on this Fritzlein should compensate the other challenge match sponsors for the damage caused.


You could argue that, but you can't have it both ways. If Fritzlein's bet decreased your equity by causing browni to play less well, then lightvector's pledge to reverse the bet for browni actually increased your equity Omar, and decreased lightvector's (beyond the $234 he might have to refund). Then lighvector should be commended for winning the challenge in spite of making it harder for himself. The (alleged) blame would only be on Fritz and browni, and it would be very unfair to penalize lightvector for a fault he had no part in, and even tried to correct.

If, on the other hand, you are arguing that lightvector's pledge decreased your equity, then you should thank Fritz, not threaten him, for his bet must have increased your equity to that same extent.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on May 1st, 2015, 9:14am

on 05/01/15 at 08:13:14, omar wrote:
A neutral 3rd party is needed if a judgement is required. No judgement is required in this case.

Does anyone dispute that these bets were placed.

Does anyone dispute that we expect key participants to follow ethical standards.

Does anyone dispute that there is a problem with the game if the participant of the game did not follow ethical standards.

Does anyone dispute that a game should be invalidated if there is a problem with the game. The challenge match rules allow for this.

The only thing that could be questioned is if the participants knew if they were breaking ethical standards or not.

Would what I have presented be correct if I only had $1 at steak, but not be correct because I have $10K. Where do you stand on what I have presented if you did not know how much I have at steak.


I dispute the missing link in the above: that lighvectors's pledge (and the Fritz-browni bet too, even if that looks irrelevant to me) broke ethical standards.

More importantly, I dispute that browni having (allegedly) not performed as well as could have been expected is a cause for invalidating sharp winning the challenge. In 2009, I confused the time of my first challenge game, and you Omar as reserve had to step up and play the game for me. Had you lost that game and had clueless won the challenge, the blame would have been only on myself, and partly on you for selecting an unreliable defender. There would have been no cause for denying clueless its victory. Likewise, if browni broke ethical standards or decreased in any way the chances of defending the challenge by his actions, it is no ground to cancel the challenge. The alledged blame is only on him and on you for selecting him.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Fritzlein on May 1st, 2015, 9:26am

on 05/01/15 at 08:13:14, omar wrote:
I was a government employee for 10 years and had to go through ethics training many times

[...]

A neutral 3rd party is needed if a judgement is required. No judgement is required in this case.

Does anyone dispute that these bets were placed.

Does anyone dispute that we expect key participants to follow ethical standards.

Does anyone dispute that there is a problem with the game if the participant of the game did not follow ethical standards.

Does anyone dispute that a game should be invalidated if there is a problem with the game. The challenge match rules allow for this.

The only thing that could be questioned is if the participants knew if they were breaking ethical standards or not.

Would what I have presented be correct if I only had $1 at steak, but not be correct because I have $10K. Where do you stand on what I have presented if you did not know how much I have at steak.

Omar, I am surprised that the ethics training you completed many times leaves you believing that you have no conflict of interest in this case.  I am stunned that you can read the discussion to this point, which contains sharp disagreements on multiple issues, and yet claim that no judgment is required.  Finally, I am bewildered that you appear to think that you have presented an ironclad train of logic when it contains a glaring logical fallacy that I know you would be capable of spotting if anyone else presented such an argument.

By posing the question of whether "the participants knew if they were breaking ethical standards", you imply it as a fact that they were breaking ethical standards.  You need this to be a fact to complete your logical train.  Yet this very point is quite obviously under dispute, and on this point if nothing else, judgment is required.

In addition to your textbook logical fallacy, there is an essential matter of fact under dispute, namely whether browni3141 would be expected to perform worse given that he had bet on himself.  (It is of course impossible to prove what caused his loss.  It is sufficient to consider what would make him more or less likely to win, and discussing probabilities is the most we can do.)  I would argue that having a financial stake would make browni3141, if anything, more likely to win.  It would tend to encourage him to study the games that other people played in the Screening, to plan ahead for his own strategy, to play practice games against previous versions of sharp, and in the game itself to exert himself to the utmost, including not resigning prematurely, not giving a handicap, and not playing objectively poor moves in an attempt to win gloriously rather than winning surely.

I acknowledge the counter-argument that having a financial stake can increase psychological pressure, which could adversely affect browni3141's play.  I submit, however, that on the balance the incentive would be expected to increase rather than decrease his winning chances.  After-the-fact reasoning that because he lost the game, it must have been due to psychological pressure is another logical fallacy.

At a minimum, the question of whether browni3141's bet on himself would be expected to increase his winning chances or decrease them or not have much effect is another point essential to deciding on the validity of the result of the Arimaa Challenge, another point on which judgment is required, and another reason you have a conflict of interest in making the decision unilaterally.






Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on May 1st, 2015, 9:33am
Finally, I'd like to make it clear that I don't support browni replaying his mini-match (or clyring playing a deciding mini-match).

After hearing thoughts from others, I've changed my mind since my first post quoted below, and would rather have the challenge purely cancelled (wrongly so, in my opinion) than a mini-match being replayed, even if the cancellation is financially more unfair to lightvector.


on 04/30/15 at 04:35:43, chessandgo wrote:
To me, voiding this year's challenge is the worst decision for arimaa. From second-best to second-worst, here are some other possibilities:

- [...]

- Declare browni's mini-match void and make him replay it.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by mattj256 on May 1st, 2015, 10:27am
I have been following this conversation for the past two days.
I was (foolishly?) hoping that this was going to be quickly resolved.

This is my opinion, for what it's worth:

1. Nobody can force Omar to reconsider his decision.  Having said that, in the American legal system it's common for a court to first make a decision and then stay that decision pending appeal.  I think it would be good for Omar to do that here.  (And going forward, as Fritzlein said, the person administering the challenge should be a third party with no financial stake in the outcome.)  I'm not a lawyer and I assume that Omar has the legal authority to invalidate the challenge.  Nevertheless I think it's bad for the community.  As Omar has said, we work hard to create a place that is fair and ethical even though hardly anybody watches the games.  We may disagree about how fairness or ethics is defined but we all believe that the others are acting in good faith, which is how it should be.  If Omar isn't willing to defer to a third party I hope he will agree to mediation or arbitration so there can be SOME third party and some closure other than the closure we were given.  Honestly I think ANY result would be better than Omar unilaterally invalidating the Challenge.

2. I think it is better if future challenge defenders are not anonymous.  All bets should be public (also a previous post) and the Tournament Director for the challenge should have no financial stake in the outcome (Fritzlein's post.)  Taken together, these two measures will ensure that challenge defenders aren't placed under inappropriate pressure.  I think there is value in having a shared community, and value in the defenders being allowed to destress afterwards and discuss their games with others.  Meanwhile Omar phrased this as a decision that he already made, not something open to discussion.  I accept Omar's decision on this.

3. Omar's position (as I understand it) is that people should be held to ethical standards and not every violation of the rules is going to be written down beforehand.  And browni3141's position (as I understand it) is that there was no written rule forbidding betting on himself, and when he bet against himself that only canceled out the first bet so there was nothing ethically wrong with either bet.  This is a thorny problem and I don't expect us to answer it quickly, but I'll ask the question.  SHOULD we have more written rules about inappropriate behavior from Arimaa players?  This won't help the current situation but it will help going forward.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Skarn on May 1st, 2015, 12:07pm
I'm seriously confused by all the people claiming browni didn't have a monetary incentive to loose. It wasn't a single transaction! Browni had already made arrangements that would have left him at -X dollars if he lost. This happened completely without the involvement of the developer. Then the developer comes along and says "if I win and you lose I'll give you X dollars". How do people not see the glaring conflict of interests here?!

And people keep bringing up the precedent of people betting on themselves. But that is completely irrelevant. The problem with betting against oneself is not that it involves money. The problem is that it creates an incentive which goes against the spirit of the game, i.e. "I should/can loose".

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by SilverMitt on May 1st, 2015, 2:38pm
Despite the logical explanations presented why lightvector's proposal did not, in any way, incentivize browni3141 to lose, I see several people repeating it as self-evident that it did.  This is disappointing.  However, having watched a government ethics training video last summer, my biggest takeaway was that even the appearance of impropriety may be unethical.

I am still waiting for the 2014 Arimaa Challenge results also to be declared invalid, as a similar bet (for 100 Arimaa points) was described in the forum in the thread named 2014 Arimaa Challenge.  I assert that omar's position cannot be consistent with only one of the challenge results declared invalid.

Finally, while I disagree with the reasons for declaring the challenge to be invalid, if humanity should get a mulligan due to the culture in the Arimaa gameroom, we should re-examine everything about the challenge to make it as fair and worthy a test of the challenger's ability.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by russ on May 1st, 2015, 3:31pm

on 05/01/15 at 12:07:14, Skarn wrote:
I'm seriously confused by all the people claiming browni didn't have a monetary incentive to loose. It wasn't a single transaction! Browni had already made arrangements that would have left him at -X dollars if he lost. This happened completely without the involvement of the developer. Then the developer comes along and says "if I win and you lose I'll give you X dollars". How do people not see the glaring conflict of interests here?!


As an outside observer, this was my thought as well. It seems like the people involved in the financial transactions are so used to such financial bets/"insurance"/etc that they don't grok how ... odd ... it looks from the outside for the programmer to offer to pay money to one of the human competitors if and only if he loses versus the program.

Justify it all you want as "good motives" or "in the end, the finances cancel out so that there is mathematically no financial incentive to win or lose for the player", but how can "I will pay you if you lose" not look somewhat ethically dubious?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by SilverMitt on May 1st, 2015, 4:03pm

on 05/01/15 at 12:07:14, Skarn wrote:
I'm seriously confused by all the people claiming browni didn't have a monetary incentive to loose. It wasn't a single transaction! Browni had already made arrangements that would have left him at -X dollars if he lost. This happened completely without the involvement of the developer. Then the developer comes along and says "if I win and you lose I'll give you X dollars". How do people not see the glaring conflict of interests here?!

...



on 05/01/15 at 15:31:04, russ wrote:
...

Justify it all you want as "good motives" or "in the end, the finances cancel out so that there is mathematically no financial incentive to win or lose for the player", but how can "I will pay you if you lose" not look somewhat ethically dubious?


The way I can interpret these views is:  "This looks bad and I don't care about the details."  This is perfectly valid, and the fact that the details exonerate lightvector is irrelevant to this case.  As long as we understand one another's views, we can all agree to disagree.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by rbarreira on May 1st, 2015, 5:37pm

on 05/01/15 at 16:03:29, SilverMitt wrote:
the fact that the details exonerate lightvector is irrelevant to this case.


Depends on what "exonerate" means. When I saw the chat between him and browni3141 I did not think anything improper was going on, and I still believe his intentions were good (both towards browni and the challenge). But are good intentions enough?

In hindsight I can see how the action was wrong regardless of the details of why and how it happened. This is why I'm not as much against Omar's decision as others are. The only thing I fault Omar for are some of his arguments in this thread that I disagree with, as well as the fact that he didn't appoint a neutral party to decide it.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 1st, 2015, 5:46pm
I can acknowledge that it's not a good thing if something looks bad at the surface to a number of people, even if the thing itself is fine when looked at closer by them, but I don't think appearances alone is enough to invalidate the Challenge result.

I think that replaying my games now is unfair for several reasons.

First, the replay is conditional on the fact that I lost my mini-match. If Omar can honestly say he would have allowed a replay of my mini-match in the case that I had won, then I would be more open to a replay.

Second, I do not know if I can play my games to the best of my ability. I assure you that my analyses will be hindered by my emotion on this matter, and also that there will be the lingering thought in my mind that if I win I will take away what I consider to be a legitimate win for sharp/lightvector. Part of being a strong player is managing emotions during a game, and I am capable of that, but only if I am dedicated to performing my best in said game, which I am not. By Omar's reasoning, any replay by me must be considered invalid by the same reasoning that my first mini-match was declared invalid, except that in this case I am declaring that I will have a real problem playing my games, not a hypothetical one (in fact I insist that I didn't have any problems before which were caused by the bets)

Thirdly, as mentioned I have played 4 games with sharp already. The Challenge defender is not supposed to have this exposure before his match (unless the developer puts his bot up voluntarily, I suppose)

Finally and least importantly, I simply don't want to play any more Challenge games this year because of the stress and anxiety they caused me (compounded now by recent events) and possibly won't want to in the future. I could deal with this and take one for the team were it my only reason, though.

I decline to replay my mini-match personally and object to someone taking my place (fwiw) until a convincing argument can be constructed against my above reasoning and against other potential objections I have that I can't think of on the spot.

On the topic of psychological stress having caused me to play worse during my first game, this may be true. However, I dispute that it was caused by the bet. There are other more severe contributing factors. After harvestsnow and chessandgo had already lost their first two games and after sharp had performed so well in the screening, I essentially felt that the Challenge defense was up to me. Also, I had not been able to play sharp at all yet, and it is a little bit intimidating to have to play an opponent you don't know much about. It's hard for me to really gauge an opponent's ability only by observing it's games with others. I need to play it.

There are also the possibilities that the perception that I played worse in my first game was caused by my unfamiliarity with sharp, the fact that I played an odd strategy to try to take advantage of it, random variance, me having a bad day for other reasons, sharp playing relatively better than usual, a biased observer, or some other reason or a combination of these reasons.

Others have already said this, but I would like to add that I also dispute that ethical boundaries were crossed by me, Fritzlein and/or lightvector. At the very least I think Omar must concede that this is an ethical gray area. I think relying on an unwritten rule of ethics is not sufficient to justify invalidating the Challenge result when it is clearly not a widely accepted rule of ethics. I dislike the analogy with swiping pieces because almost no-one would dispute that it is clearly unethical. On the other hand many of us here are disputing that the acts of betting/compensation were clearly unethical.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lightvector on May 1st, 2015, 7:54pm
More than one person has privately asked me to publish my opinions on the issue and on the various possible courses of action that have been suggested. I have been avoiding being part of the discussion about what should be done because given my position in the matter, it's simply not my place to be part of the decision-making here.

I enjoy this game and treasure the community around it, and I greatly regret that this event has spawned such controversy and division. I still stand by what I did, because to me it would have been intolerable not to offer my help to someone in such a situation, particularly when doing so simply restored things back to the status quo.

Also, given the way that the discussion has proceeded and how things do not yet appear to be heading to a resolution, it seems clear to me that the event that was to be the closing ceremony on Saturday tomorrow should be postponed. I don't plan to be at this event tomorrow.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 11:11pm
Thank you browni3141 and lightvector for posting your thoughts on this. browni3141, I totally understand you not wanting to play.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 1st, 2015, 11:24pm

Quote:
The only thing that could be questioned is if the participants knew if they were breaking ethical standards or not.


When I said that, I was not trying to rise an issue. We all know that the participants weren't intentionally trying to break any ethical standards. I just meant that if this point was under debate than it would require a 3rd party to make a judgement on this.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 2nd, 2015, 12:22am
Many people are disagreeing with my position because I am one of the challenge match sponsors and so could be considered to have an interest in the outcome. But they are missing a key point.

If one can go from given facts to a conclusion without needing to bring in any judgement then it doesn't matter who carries out the step to go from the given to the conclusion. This is why a tournament coordinator can play in the tournament while a tournament director cannot. The tournament coordinator only carries out the predefined rules of the tournament as various conditions are met. The tournament director is invoked if there is any judgmental decision to be made.

I seems clear to me in this case that anyone can go from the given to the conclusion without needing to invoke any judgement. If that was not the case I would have asked a 3rd party to make the decision.

Where am I invoking a judgement in the following:


Quote:
Does anyone dispute that these bets were placed.

Does anyone dispute that we expect key participants to follow ethical standards.

Does anyone dispute that there is a problem with the game if the participant of the game did not follow ethical standards.

Does anyone dispute that a game should be invalidated if there is a problem with the game. The challenge match rules allow for this.


If anyone can point out where I invoked judgement then I would be happy to have a 3rd party make the judgement call.

In the Arimaa Challenge I have been serving as the tournament coordinator; scheduling the games; starting the bots, etc. In doing so I follow the predefined rules of the Arimaa Challenge. For example we have rules that say that if a game goes wrong it needs to be replayed. I don't need to go to the tournament director if the conditions for replaying the game are met. I only need to go to the tournament director if the situation requires a judgement call.

brownie3141 asked me a the question:

Quote:
If Omar can honestly say he would have allowed a replay of my mini-match in the case that I had won, then I would be more open to a replay.

I can clearly answer yes to this, but it is not Omar answering the question, it is Omar in the role of the tournament coordinator answering yes based on the predefined rules of the Arimaa Challenge.

Quote:
The game must be replayed whether the disadvantaged player won or lost.


I don't think I have made any judgmental decisions in this case. If you can show me where I have, then I can be persuaded to retract the decision or pass it on to a 3rd party.

Argument such as "you are an interested party so any decision you make is judgmental" are baseless.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PhilomathBret on May 2nd, 2015, 12:49am
I would like to add another question that should have been in that list: "Does anyone dispute that ethical standards have been broken?" in which case the answer is clearly "yes". This is why a third party should make the call.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by omar on May 2nd, 2015, 8:31am

on 05/02/15 at 00:49:53, PhilomathBret wrote:
I would like to add another question that should have been in that list: "Does anyone dispute that ethical standards have been broken?" in which case the answer is clearly "yes". This is why a third party should make the call.


Good point Bret. Also I received this PM from bot_sharp:


Quote:
In going from the given (that key participants placed bets) to the conclusion that the effected games should be invalidated and replayed, I do see one technical problem. The Arimaa Challenge rules state that:
"If it is determined after a game is over, but before the next game of the player's series, that either bot or human player has suffered a disadvantage, the game result will be invalidated, the game will be made unrated, and the game will be replayed from the point at which the disadvantage first occurred."
This means that the only game which can be invalidated at this point is the last game of browni3141's series. Since that game did not have any impact on the result of the series, it makes no difference if it is replayed.

Besides I don't want to play browni3141 again; you have no idea how scary it is to play him.


Wow, a very compelling argument from Sharp. I guess this means that the 2015 Arimaa Challenge match results stand. Sharp has his tactics down onboard and off :-)

Guys, I'm really sorry if I caused too much stress for you by creating this suspense and drama. When I found out about the bets that were placed I was quite disturbed and disappointed and I did not know it had happened in a previous Challenge match until I learned about it from this thread. However, I knew it was too late as per the challenge rules to do anything about it. But I thought it would be a good opportunity to add some suspense and drama to the challenge match being won. I hold no hard feeling from what anyone has said in this thread and I'm sorry if I hurt anyone else feelings by what I said. The only person I let in on this was Janzert and only after it seemed that he might have been effected by this thread.

Also I just want to clarify that I made up that PM from bot_sharp.

All is well in Arimaaland !!!

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lightvector on May 2nd, 2015, 10:55am
I have just talked with Omar, and I can confirm that although some aspects of it were genuine (such as his prior lack of knowledge of the betting and some of his feelings towards the matter) and that however in-bad-taste it might have been, according to him this was in large part a play to add some additional drama into the ending of the challenge.

I know that for some people, me included, this event has had a strong impact. I did not enjoy it, and I think it's completely understandable if this affects anyone's desires in continuing to be a part of this community and game. But I'm also glad that things can be resolved, and I just want to move forward.

I continue to stand for fun and community, and will be proceeding with developing my app and rounding up a team to make a new website. I've talked with Omar to reschedule the closing event for next Saturday May 9 to occur at the same time (16:00 GMT) to talk about this and other plans going forward.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Skarn on May 2nd, 2015, 11:23am

on 05/01/15 at 16:03:29, SilverMitt wrote:
The way I can interpret these views is:  "This looks bad and I don't care about the details."  This is perfectly valid, and the fact that the details exonerate lightvector is irrelevant to this case.  As long as we understand one another's views, we can all agree to disagree.


This wasn't supposed to be game-night among friends, it was supposed to be a formal competition. It was supposed to be squeaky clean.

Suppose a businessman offered to pay $1000 to a senator of your country if a bill that benefits his company passes. The senator votes "yes" and the bill passes. Later when confronted about it the senator says that $1000 is very little money compared to the wealth he already has, that he truly cares about politics and that his highschool buddy had also offered him $1000, only if the bill didn't pass. Would you be comfortable with that situation?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by deep_blue on May 2nd, 2015, 11:36am

on 05/02/15 at 11:23:03, Skarn wrote:
Suppose a businessman offered to pay $1000 to a senator of your country if a bill that benefits his company passes. The senator votes "yes" and the bill passes. Later when confronted about it the senator says that $1000 is very little money compared to the wealth he already has, that he truly cares about politics and that his highschool buddy had also offered him $1000, only if the bill didn't pass. Would you be comfortable with that situation?

You can't really compare that. In your situation someone gets money for doing something.
Browni doesn't get any money for losing.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by rbarreira on May 2nd, 2015, 11:46am

on 05/02/15 at 08:31:01, omar wrote:
But I thought it would be a good opportunity to add some suspense and drama to the challenge match being won. I hold no hard feeling from what anyone has said in this thread and I'm sorry if I hurt anyone else feelings by what I said. The only person I let in on this was Janzert and only after it seemed that he might have been effected by this thread.


I'm not sure if I believe this, and even if I do it's still ridiculous and unpleasant.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by russ on May 2nd, 2015, 12:11pm

on 05/02/15 at 11:36:40, deep_blue wrote:
You can't really compare that. In your situation someone gets money for doing something.
Browni doesn't get any money for losing.

Suppose your senator is considering voting for e.g. a tax bill which would cost the senator personally $1000 due to the change in tax calculations which would result. Suppose a lobbyist wants that that tax bill to pass and gives the senator $1000 if he supports it. Now the senator mathematically/financially has no personal incentive either way, because whether or not he supports it, he personally neither loses nor gains money as a result.

Does that mean that there's nothing wrong with the lobbyist giving $1000 to the senator to support the bill?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 2nd, 2015, 12:27pm

on 05/02/15 at 12:11:20, russ wrote:
Suppose your senator is considering voting for e.g. a tax bill which would cost the senator personally $1000 due to the change in tax calculations which would result. Suppose a lobbyist wants that that tax bill to pass and gives the senator $1000 if he supports it. Now the senator mathematically/financially has no personal incentive either way, because whether or not he supports it, he personally neither loses nor gains money as a result.

Does that mean that there's nothing wrong with the lobbyist giving $1000 to the senator to support the bill?


Intent is important.

The lobbyist's intent is to mold legislation to his liking.

Lightvector's intent was to relieve Browni's stress.

                                   

It's moot now, but I think it's worthwhile to explore the ethical contours of this discussion. The arguments have tended to divide along deontological/consequentialist lines. The deontological argument has been that the Challenge should be invalidated because ethical rules were broken, while the consequentialist argument has been that the Challenge should stand because the net consequence of the bets didn't give Browni any incentive to lose. (This excepts the deontological argument that the Challenge should stand because no rules were explicitly broken.)

The two sides have been talking past each other because they've been employing two vastly different ethical frameworks.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 2nd, 2015, 12:41pm

on 05/02/15 at 11:23:03, Skarn wrote:
This wasn't supposed to be game-night among friends, it was supposed to be a formal competition. It was supposed to be squeaky clean.

Suppose a businessman offered to pay $1000 to a senator of your country if a bill that benefits his company passes. The senator votes "yes" and the bill passes. Later when confronted about it the senator says that $1000 is very little money compared to the wealth he already has, that he truly cares about politics and that his highschool buddy had also offered him $1000, only if the bill didn't pass. Would you be comfortable with that situation?


I don't feel like discussing the analogy (although I think there are some flaws with it), but I'll take a stab at the situation.

I would say that the highschool buddy and the businessman both made unethical decisions if they did not know another negating offer would be accepted (although then it would be against their own interest for either of them to offer so I assume they didn't know), and that the senator made an unethical decision if he accepted on offer without intention of accepting the other. The senator acted in his best interest to take advantage of two other bad decisions and make $500 from each party. I'm kind of borderline on whether or not this type of exploitation is okay, but I'm pretty sure enough people would feel it is not okay to make it unacceptable to allow them to know. Therefore the senator made an unethical decision if any of the offers were public knowledge because it looks like corruption, but would be fine if all communications were private.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Skarn on May 2nd, 2015, 12:43pm

on 05/02/15 at 11:36:40, deep_blue wrote:
You can't really compare that. In your situation someone gets money for doing something.
Browni doesn't get any money for losing.


Suppose we have a match in the WC. You promise beforehand to pay my college debt/phone bill if and only if I loose. Totally okay, it's not like I'd be gaining anything, right? Before the agreement with lightvector browni would have been at -$236 if he lost. Lighvector made that number be cero. $0 > -$236.

I know that as a newbie I am a relative outsider, but as much as there might have been a culture of making small bets in the community, the challenge should have been squeaky clean. It was a formal competition, not whoever wins Catan pays for pizza.

No one is being accused of acting in bad faith, but there was definitely bad form. It is a shame that the arimaa challenge will go down in history somewhat tarnished by this.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Skarn on May 2nd, 2015, 12:51pm

on 05/02/15 at 12:27:24, CraggyCornmeal wrote:
The two sides have been talking past each other because they've been employing two vastly different ethical frameworks.


That's not true, I'm a consequentialist. These events discredit the arimaa challenge, the arimaa community and set a terrible precedent of what to expect from future competitions.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 2nd, 2015, 3:24pm

on 05/02/15 at 12:43:29, Skarn wrote:
Suppose we have a match in the WC. You promise beforehand to pay my college debt/phone bill if and only if I loose. Totally okay, it's not like I'd be gaining anything, right? Before the agreement with lightvector browni would have been at -$236 if he lost. Lighvector made that number be cero. $0 > -$236.

I know that as a newbie I am a relative outsider, but as much as there might have been a culture of making small bets in the community, the challenge should have been squeaky clean. It was a formal competition, not whoever wins Catan pays for pizza.

No one is being accused of acting in bad faith, but there was definitely bad form. It is a shame that the arimaa challenge will go down in history somewhat tarnished by this.


This is a completely different scenario. Paying out to Fritzlein is conditional upon the Challenge being lost. Paying my phone bill would not be.

Before I accept your offer to pay my phone bill if I lose, I am -$236 if I lose and -$236 if I win, so both decisions are equal. After I accept your offer, I am +$0 dollars if I lose and -$236 if I win, making me prefer to lose.

Before I accepted lightvector's offer to payout my bet to Fritzlein, I was -$WCev if I lost and +$0 if I won, so I prefer to win. After I accepted I was +$0 if I lost and +$0 if I won, so both decisions are equal monetarily. If I had reason to lose without monetary incentive I should not have accepted lightvector's offer and omar should not have chosen me as a defender. If lightvector thought I would have reason to lose he should not have offered. Neither of these were the case so there was nothing wrong with the offer or acceptance except that other people incorrectly perceive it as wrong.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lucky81 on May 2nd, 2015, 4:15pm
Here is how I see it:

Fritzlein created a monetary incentive for browni to win. I don't see anything wrong with that. He also created a monetary incentive for Sharp to win! It's a common scheme in sports. Chess players play for money too. In fact, I'm surprised Omar hadn't himself allocated at least a small token piece of the Challenge prize money for the defenders. There is no conflict of interest there.

Lightvector then removed browni's monetary incentive to win. The only excuse I see for that is that the amount was small.

Since everybody throws in analogies, I will provide mine as well.

AC Milan plays Manchester United for the UEFA Champions League final. Coca Cola, MU's sponsor, shows up and says to the Manchester United players: we will pay $100,000 to every player that scores a goal in the final.

The manager of AC Milan gets word of this, goes to MU players, and says: we will pay $100,000 to every player that does *not* score a goal in the final.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by mattj256 on May 2nd, 2015, 10:01pm

on 05/02/15 at 08:31:01, omar wrote:
I guess this means that the 2015 Arimaa Challenge match results stand.... I thought it would be a good opportunity to add some suspense and drama to the challenge match being won.
I used to watch the TV show "House" a lot, and I would be annoyed because the medical aspect was already suspenseful and dramatic, but the writers felt they had to inject soap-opera elements into the story line to keep it interesting.

I was one of the 50 people logged into the chat room who watched live as Sharp clinched the challenge by beating browni3141 in game six of the Challenge.  That was supposed to be the memorable part of the tournament.  My coworkers don't know the rules of Arimaa, but they all know that browni3141 won the World Championship and Sharp won the Arimaa challenge.  Arimaa is a great game, and writing a bot that can beat humans (with or without the prize money) is a challenge that is worthy of me.  And also we have a great community, but I wouldn't be part of the community if the game didn't captivate and fascinate me.

I hope we are done with all this soap opera stuff.  If betting on games is an important issue I hope there will be meaningful changes next year.

I hope I haven't said anything that has hurt anyone's feelings, and if anyone is upset at me please reach out to me so I can address what's wrong.

After lightvector publishes in ICGA and wins the prize, I hope we'll have another opportunity to recognize his accomplishments.

To the extent that I make time for Arimaa, I want it to be building my bot, improving the website, improving my skills, and helping the community.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Heyckie on May 3rd, 2015, 4:10pm
Wow this really turned into nearly the worst scenario I had in mind when I decided to send Omar that message. I'm very sorry  :-[

Most of all sorry Fritzlein and browni3141, you didn't do anything wrong (well ok browni maybe could have declined lightvector's offer but nobody can be expected to be that perfect). My message to omar didn't mention your original bet at all, I still just don't see absolutely anything wrong with it.

I guess I should have messaged not omar but lightvector, and point to him that his offer was simply wrong, no matter how well intended, an that maybe he should apologize. But that's hindsight again and having only hindsight and no real time reasoning was what got me dragging the whole community into this the first place.

In my defense:
1. The situation was such that if I decided I should contact someone, the sooner I did it the better. Also I've had a severe Real Life Syndrome for about a week. That meant I simply could not think as much as I would have wanted to about what to do.
2. The only other option I could think of at the moment was to keep quiet and consent feeling a somewhat guilty for an unknown amount of time if no fuss ever came out of lightvector's offer, or b very very guilty if someone else realized it after or during the ceremonies and started to wonder why nobody told omar etc. I still don't know how bad that would have been, on current hindsight probably not as bad as this, but can't know for sure (for example b could have happened). I still don't like the idea of starting to whine and speculate about this in chat behind omar's and lightvector's backs.
3. This drama really is about the worst I could imagine happening when contacting omar (only thing worse still would have been an actual invalidation of the Challenge). I guess I hoped he would maybe just privately scold lightvector a bit or something.

The thing I'm slapping myself in the head hardest about is that I was right then and there in the Chatroom when the inappropriate offer was made and did not in any way object to it at the time.

About all the talk and arguments made about the Challenge in this thread I'm so at one with chessandgo that I really have nothing worthwhile to add.

I tried to apologize yesterday when I saw the drama, but couldn't as I only had an old phone to get online with and it wouldn't send a message of much over a hundred characters; had to shorten my already short message to PerkofBR about our EEE game to get it through.

Really sorry.
Please try to make peace with each other you all. Blame me if you can't stop your angry feelings.
Heyckie

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by ikalyoncu on May 3rd, 2015, 5:12pm

on 05/02/15 at 22:01:30, mattj256 wrote:
I was one of the 50 people logged into the chat room who watched live as Sharp clinched the challenge by beating browni3141 in game six of the Challenge.  That was supposed to be the memorable part of the tournament.


My feelings exactly, Omar.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by supersamu on May 3rd, 2015, 5:38pm
I originally wanted to send the following to Heyckie privately, but I feel this is a better place to share it:

I don't feel you did anything wrong and nobody can expect you to know what Omar would do with the information given to him. I was in the chatroom as well (or I read the chat in the archive, I can't remember), and I didn't raise an eyebrow when the bet was discussed. So I don't think you should feel ashamed for not speaking out when the bet/lightvector's promise happened.
I also don't see you as a snitch because you acted behind the back of the bettors and told Omar about the situation without telling Fritzlein, browni3141 or lightvector.

You saw something that might harm the perception of Arimaa or the view of the legitimacy of the challenge and gave Omar information that was already public.
You are not responsible for the way Omar reacted to the information.
You thought Omar deserves to know what happened (imagine if Omar found out 2 years later) and you gave him the chance to respond to the situation within his own terms.

Again:
You absolutely don't need to apologize for telling Omar of the bet. I fully support your decision, and I would also have fully supported your decision if the challenge match were declared invalid, or any other scenario you can imagine happened.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by kzb52 on May 3rd, 2015, 5:51pm

on 05/03/15 at 17:38:18, supersamu wrote:
Again:
You absolutely don't need to apologize for telling Omar of the bet. I fully support your decision, and I would also have fully supported your decision if the challenge match were declared invalid, or any other scenario you can imagine happened.


Just posting to agree emphatically with this.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lightvector on May 4th, 2015, 12:15am

on 05/03/15 at 17:38:18, supersamu wrote:
You absolutely don't need to apologize for telling Omar of the bet.


Agreed as well.

Setting all questions of right and wrong aside, I feel responsible for having taken an action that, one way or another, contributed and led to the happening of events of the previous several days. And the impact of these few days here has been one of extreme disagreement and discord.

Going forward, I intend to proceed with the projects I've mentioned before and hope to restore some fun and energy to things. I never particularly valued winning the Challenge - writing a bot gave me something interesting to do, but what kept me here was the game and the people. So in keeping with that, if it was at all unclear before: I will be donating the Challenge prize money back to fund future tournaments and prizes and events for Arimaa over the next several years.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on May 4th, 2015, 2:58am
I agree 100% with the above posts. You did the right thing Heyckie, no need to apologize for anything.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 4th, 2015, 1:44pm
I also agree. Heyckie absolutely did the right thing.

I don't think anyone's actions were bad. But by chance, the confluence of these actions produced stressful results. Stuff happens. Once in awhile bad actions produce good results. And once in awhile good actions produce bad results.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PotatoeTheCat on May 9th, 2015, 3:13am
Arimaa does not exist in a vacuum: its participants are subject to the same standards as any other gamers/athletes.  All published codes of ethics are clear on his point: no side betting allowed.  Period.
eg
http://www.egba.eu/pdf/Athletes_COC_A5_EN_v08.pdf

One must accept the participants' statements in good faith that the side-bets were not intended to, and did not affect the motivation or conduct during the challenge games.

However, the impartial observer will note that Browni lost his first two games (while the Challenge was still up for grabs) reserving his normal, superlative bot-bashing form for the third game (only after the Challenge was decided).

Unfortunately, the non-Arimaa world will not have the time or patience to wade through a 6-page forum note.  Rather, it will look at the side-bets and sequence of Browni-bot_sharp games, and conclude (perhaps superficially) that the outcome of the Challenge match is tainted, and possibly corrupt.

Sorry guys: that is how it is.  And there is no simple fix for this.

This is a shame for Arimaa, Browni and (especially) for David Wu, who has contributed tirelessly & selflessly to the Arimaa community over a number of years with a number of iterations of bot_sharp.  

bot_sharp could perhaps be considered a little fortunate to have faced an off-form Browni (and chessandgo) in R1 & R2.  

But, so what: sporting encounters often are decided in such fashion!  David fully deserves recognition as the creator of the first programme to defeat a team of humans over the Computer Challenge format.

David's offer to return his prize to fund future Computer Challenges is fair one, and (to any reasonable & informed observer) obviates any allegation of corruption.

Congratulations to David - I am looking forwards to watching the next iteration of bot_sharp fighting it out with Brown & Co in 2016!


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by chessandgo on May 9th, 2015, 4:44am

on 05/09/15 at 03:13:20, PotatoeTheCat wrote:
bot_sharp could perhaps be considered a little fortunate to have faced an off-form Browni (and chessandgo) in R1 & R2.  


And harvestsnow in R1 & R2 & R3. And the rest of humankind in the screening.

Or maybe you have it wrong, and Browni and I are the ones fortunate to have faced an off-form sharp in R3?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 9th, 2015, 4:19pm

on 05/09/15 at 03:13:20, PotatoeTheCat wrote:
Arimaa does not exist in a vacuum: its participants are subject to the same standards as any other gamers/athletes.  All published codes of ethics are clear on his point: no side betting allowed.  Period.
eg
http://www.egba.eu/pdf/Athletes_COC_A5_EN_v08.pdf


The following is assuming you are talking about only my bet with Fritzlein, despite your plural "bets," since the thing with lightvector was not a bet but an offer.

Why should I adhere to ethical guidelines set by any organization with absolutely no jurisdiction over Arimaa? I am NOT subject to their standards or yours. I do not regret my side bet and unless rules regarding them are established I would do it again.

In no way is side-betting disallowed by any written rule, and relying on "unwritten rules" is completely unreasonable as a reasonable person following a reasonable set of ethics may not consider certain side-bets even a minor breach of ethics.


Quote:
One must accept the participants' statements in good faith that the side-bets were not intended to, and did not affect the motivation or conduct during the challenge games.


No one mustn't. Reason can show that I was not affected in any logical decisions, and common sense suggests I shouldn't have been affected psychologically, which is irrelevant anyway, unless people really expect me to make perfect decisions regarding ensuring my top mental condition for my games. Maybe the side bet decreased my performance ability. Maybe not drinking enough water decreased my performance ability.


Quote:
However, the impartial observer will note that Browni lost his first two games (while the Challenge was still up for grabs) reserving his normal, superlative bot-bashing form for the third game (only after the Challenge was decided).

Maybe I should have thrown my final game then, but I'd much rather simply not care what some people might think.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by russ on May 10th, 2015, 5:31am

on 05/09/15 at 16:19:05, browni3141 wrote:
The following is assuming you are talking about only my bet with Fritzlein, despite your plural "bets," since the thing with lightvector was not a bet but an offer.


Sure, but an offer to pay you if and only if you lose to the bot does not exactly look better, or less fishy, than a bet does. :)

(To be clear, I'm not attributing illicit motives to anyone, but I am continually surprised that so many of the people directly involved are seemingly not seeing at all why an offer to pay a competitor if and only if the competitor loses to the payer's bot might not at least look a bit ethically strange...)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by supersamu on May 10th, 2015, 9:54am

on 05/10/15 at 05:31:44, russ wrote:
Sure, but an offer to pay you if and only if you lose to the bot does not exactly look better, or less fishy, than a bet does. :)

(To be clear, I'm not attributing illicit motives to anyone, but I am continually surprised that so many of the people directly involved are seemingly not seeing at all why an offer to pay a competitor if and only if the competitor loses to the payer's bot might not at least look a bit ethically strange...)



It is because browni does not get payed. After accepting lightvector's offer, browni's situation is as follows:

- If browni loses, he gives x$ to Fritzlein and receives x$ from lightvector. Net result: Browni has no monetary incentive to lose.
- If browni wins, he does not receive any amount of money from any individual. Net result: Browni has no monetary incentive to win.

One can also look at the bets differently:
Fritzlein pays browni 2$, and browni gives Fritzlein x$. If the challenge is defended, browni receives x$ from Fritzlein. If the challenge is not defended, browni does not receive any money.

After accepting lightvector's offer:
If the challenge is defended, browni receives x$ from Fritzlein.
If the challenge is not defended, browni receives x$ from lightvector.

Why does that look strange to anyone?

It is true that browni had a monetary incentive to win his game before he accepted lightvector's offer, but before the second game browni played, he knew he had to win 2 games in a row or lose x$, a risk he willingly took himself.
Lightvector's offer gave browni the chance to forget about that risk. How that affects his play, we cannot know.
Lightvector's offer removed a monetary incentive to win, but gave no monetary incentive to lose.

So yes, if you only know about lighvector's offer and nothing else, sure it does look ethically strange. But anyone who reads further  and thinks about the situation will realize that browni had no reason to throw his games.

Actions are not ethically wrong because they look ethically wrong on the surface.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PotatoeTheCat on May 10th, 2015, 12:46pm
The following abstract example might help make some of the issues clearer.

A is an athlete playing for his team in the final of a sporting event.  If his team loses, he stands to personally lose $x.

B is a professional gambler who has bet heavily on the final.  If A's team loses, B stands to win $50*x.

B becomes aware of A's financial incentive, and offers to indemnify any loss to A, should A's team lose.

It's perfectly clear that the sport's governing body, are going to regard both B's offer, and A's acceptance of it as morally squiffy.

The fact that the result of the final is financially neutral to A, assuming that he accepts B's offer, is besides the point.  

The offer (and acceptance of it) will be viewed as unethical precisely because it eliminates A's financial incentive to ensure that his team wins.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by russ on May 10th, 2015, 1:16pm

on 05/10/15 at 09:54:41, supersamu wrote:
It is because browni does not get payed.

Yes, I'm quite aware of how the 2 transactions would cancel out.


Quote:
Why does that look strange to anyone?

If you really sincerely don't get how it wouldn't look strange that the bot's programmer offers to pay a challenger money if and only if the challenger loses to the bot, I honestly don't know what else to say.

Suppose 2 large corporations each offer to donate a million euros/dollars to a politician, one to encourage him to vote for a proposal and one to encourage him to vote against a proposal. He thus has no financial incentive to prefer voting for or against the proposal! So is this perfectly ethical?


Quote:
Actions are not ethically wrong because they look ethically wrong on the surface.

Agreed, but also remember that a large point of the Arimaa challenge was gaining favorable publicity to the game, and surely to create a clearly "fair play" environment. Having the programmer pay a player if and only if he loses to the programmer's bot is pretty obviously working counter to that, from a public relations point of view, even if you personally truly believe that the "cancelled out transactions" mean that logically there was no ethical issue.

It was, in effect, a scientific experiment: can a program beat some of the top human players? In other areas of science, it would be considered quite ethically dubious if, e.g., a tobacco company said "We'll pay you X dollars if you find that our tobacco does not cause health damage" or something like that, even if that financial offer happened to "merely cancel out" some other financial incentive the experiment's participants had encouraging the other direction.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by mattj256 on May 10th, 2015, 2:15pm

on 05/10/15 at 13:16:11, russ wrote:
even if you personally truly believe that the "cancelled out transactions" mean that logically there was no ethical issue.

The problem here is that there was no rule on the books prior to the Challenge.  It has already been said on the forum that Omar and browni3141 were operating from two different ethical frameworks, and that hasn't changed.


on 05/10/15 at 13:16:11, russ wrote:
It was, in effect, a scientific experiment: can a program beat some of the top human players?
If browni were the only defender I would be totally open to considering that some kind of unintentional / unconscious bias affected the results.  The fact is that Sharp beat three different strong human defenders, AND went undefeated in the Computer Championship, AND has won 34/36 of its games since Omar made the bot available to the public.  [1] (And the public version of Sharp2015CC is running on weaker hardware than what was used for the challenge!)

I think it's really unfair to browni to keep dragging and re-dragging and re-re-dragging him through the mud over this.  Do you want changed or clarified rules for next year?  That's a great thing to talk about.  The fact is that Omar was only joking when he called into question the results of the challenge. [2]  The challenge is over, lightvector is going to get his prize money, and I'm more than ready to move on to other things.

I don't want to stifle debate, but I also don't think it's healthy for the community for this debate to continue indefinitely.  Hopefully some older and wiser (and higher-rated) players will help me out here.  I'm not going to fault browni for operating out of a different ethical framework than Russ or Omar.  If there had been a written rule on the books, or if browni had made the bet against himself first, this would be a completely different story.  As it stands, I simply don't know what there is to be done other than agree to disagree and move on.

[1] http://arimaa.com/arimaa/gameroom/playerpage.cgi?id=23751
[2] I personally didn't realize that this was a joke at the time.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 10th, 2015, 3:07pm

on 05/10/15 at 14:15:01, mattj256 wrote:
 [1] (And the public version of Sharp2015CC is running on weaker hardware than what was used for the challenge!)

[1] http://arimaa.com/arimaa/gameroom/playerpage.cgi?id=23751

Actually, I'm pretty sure that bot is running on the Challenge hardware at full strength until the end of the month.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 10th, 2015, 6:22pm

on 05/10/15 at 14:15:01, mattj256 wrote:
Do you want changed or clarified rules for next year? That's a great thing to talk about.


How can we prevent a situation like like this from arising again? There are two things we should try to thwart:

1) Actual misconduct.
2) Apparent misconduct. Because even though nothing truly unethical happened this time, no one wants wants to endure another of these arguments.

I liked the idea of mandating that all bets be posted on the forum for everyone to see. But awhile ago in chat, Browni argued that this would encourage people to bet through private channels, driving the practice underground and making it impossible to police.

This is a good point, so I'd like others to share their ideas.

By the way, since the Challenge concluded, Browni has played three games against Sharp. He's 1-2. So if the Challenge had been replayed, the results would have been exactly the same.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by deep_blue on May 11th, 2015, 8:05am
I agree that we shouldn't keep debating for something that doesn't matter anymore. But then we should think about ways to prevent problems like that in the future.
I quite like the idea of a betting thread and I don't think that would lead to people betting through private channels since the effort to post a bet is close to 0 and there's no reason to bet in chat but not in forum.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Fritzlein on May 11th, 2015, 6:28pm
My previous responses in this thread were directed mostly at Omar.  I would now like to say more comprehensively why I believe that sharp's victory was valid, in light of what many people in addition to Omar have said.

In a court of law, it is foolish to pursue two contradictory lines of argument in the hopes that one or the other will persuade the judge.  Each argument makes the other less persuasive; your case is stronger if you advance only the better argument and stick to it.  That said, if you lose the case with one consistent line of reasoning, you can rest assured that another lawyer will say you could have won with the other line.  In the old joke about commentating chess endgames, you just say that the loser moved the wrong rook to the open file, and leave it at that.  Contrary to the joke, I would like to contend that the Challenge result should stand, whichever rook you move.

Central to Omar's case for invalidating the Challenge result was his contention that browni3141 played worse in his first game when he had money on the line, and better in his second two games when he did not stand to lose money and could relax.  It follows that Omar could only object to lightvector's action on the grounds that it was unethical, but not that it had adversely affected browni3141's play.

I would argue that an unethical act that doesn't adversely affect the course of the Challenge match can't be grounds for invalidating the Challenge.  For example, if lightvector had robbed a bank after browni3141's first game, we could agree that lightvector behaved unethically, but the Challenge result would be just as valid as if he hadn't robbed a bank.  Omar's argument about how browni3141's play was affected by lightvector's action doesn't let lightvector off the hook ethically, but it does mean that what lightvector did can't be grounds for invalidating the Challenge result.

Omar's case then rests entirely on the fact that browni3141 played worse in the first game due to pressure from his bet with me.  I submit that this fact is impossible to establish.  How do we know that he didn't play worse because he was up against an unfamiliar opponent and better later as he learned?  How do we know it wasn't because of what browni3141 ate for breakfast?  For that matter, how do we know that the difference from later games wasn't random variation in sharp's play?  An Arimaa game is such a complicated affair that depends on so many variables, it is absurd to think we can precisely untangle the causes.

The only way to make a case in such a situation is via probabilities that can be substantiated by a great quantity of evidence.  It would be impossible to establish in this single game that being in danger of losing money from losing the game made browni3141 play worse, but one could argue that in general people who play Arimaa (or chess, or tennis, or whatever) play worse when they stand to lose money from losing.

When it comes right down to it, I am not sure what the broad evidence says on this score.  I would be curious if anyone can point to studies of the effects of monetary incentives in sports and games.  Most of my personal experience has come from poker.  Most people, when playing poker only for chips (i.e. when they won't lose any money for losing the game), play far worse than when they are playing for money.  They call all-in on ridiculously weak hands, and make complete bluffs with suicidal frequency.  Only when they stand to lose something do they reign in their loose bets and begin to play somewhat sanely.

There are, of course, exceptions.  For example, my wife plays poker just as fiercely over chips as she does over quarters.  She tries to make the highest-expectancy play according to her understanding of the game no matter what the expectancy is denominated in.  She is always trying to win.  In her case, adding money to a game of poker could well be expected to make her play worse, because she might play too tightly in order to avoid a big loss.

So I admit that there may be some uncertainty and/or variation in the way a financial disincentive to lose affects game players, but I believe that in general they make people less likely to lose.  Before the match took place, I thought that I was increasing browni's chances to win by encouraging him to prepare seriously instead of underestimating his opponent.  If browni3141 had another upcoming match against sharp, and I was on the hook for another $1000 if he lost, I would make another bet with browni3141 in the same direction as last time, because I would still think I was increasing his chances to win.

In my experience, it is only such predictions that carry useful information about cause and effect, because after something has happened, people can explain it in any way that is convenient.  I think a judge or a jury would make the same prediction as I do for a future match.  I reject the idea that my bet with browni3141 would be expected to reduce his chances of winning.  Therefore I reject the notion that my bet with browni3141 is grounds for invalidating the Challenge.

It has been pointed out that in other sports, participants are not allowed to place any bets on their games, even betting on themselves to win.  Yet it is quite common to have bonuses for achievements such as rushing 1000 yards or hitting 30 home runs.  It is routine for a player to take the field knowing he will be thousands of dollars better off for performing well than for performing poorly, and indeed sports contracts are written to heighten rather than diminish this dynamic.  Therefore financial incentives per se can't be the reason for the gambling ban.  I am not sure what the reason is, but I think the reason needs to be elucidated before it can be clear that it applies to Arimaa as well, and in particular to the Challenge.

In any case, this would be an argument about what the rules for Arimaa events should be, not about what the rules were.  The Arimaa Challenge was not held under the rules of football.  I was deemed a perfectly acceptable Challenge Defender in multiple years when I stood to lose money for losing, the same situation browni3141 was in after his bet with me.  He effectively became an additional sponsor of the Challenge Prize.  There was never an Arimaa event in which it had been against the rules to bet on yourself, and there was, on the contrary, a tradition that it was acceptable.  Yes, if a rule (of Arimaa, not football) had been violated, that would in itself be grounds for invalidating the Challenge result in the absence of any other harm, but invoking an "unwritten rule" when there has been no harm is completely insufficient grounds for invalidating a challenge.  

It is now time for me to turn to the "other rook", and address the people who think that my bet with browni3141 is no grounds for invalidating the Challenge, and lightvector's offer to browni3141 is the whole problem.  These people are indeed contradicting what Omar said, but they are not contradicting themselves.  They merely offer the stronger line of reasoning that would have a better chance in court if it were pursued from the start.

In this line of reasoning, lightvector's offer to browni3141 could be expected to increase the chances that browni3141 would lose.  This can only be because my bet with browni3141 increased the chances that browni3141 would win (You are welcome, Omar), but lightvector cancelled out the effect of my bet.

A couple of posters have drawn analogies to influencing legislation.  These are inapplicable because is against the rules (the law) to pay a legislator to vote in either direction.  It's not OK to incentivize a "yes" vote and wrong to incentivize a "no" vote: each promised payment causes harm.  This is unlike the Challenge, where influencing browni3141 to play better is a good thing.  Also legislation is unlike the Challenge in that even if the two incentives cancel out so that there is arguably no harm, they are still both against the rules.

No, the strongest argument that I see is that lightvector's action could be expected to decrease browni3141's level of play, which would in turn benefit lightvector financially by increasing his chances of winning the Arimaa Challenge.  This wasn't against the rules, but it is a tangible harm.

I buy into this line of reasoning.  Yes, russ, it looks "ethically strange".  The only way I give lightvector a pass ethically is that I completely believe he didn't realize he was doing harm.  In my opinion, lightvector thought that he was acting to help browni3141 play his best, ala Omar's argument.  I truly believe that lightvector is such a generous human being that he was willing to both reduce his chances of winning the Challenge prize and reduce the amount of the prize he would win, merely to alleviate browni3141's psychological suffering.

That said, one can give lightvector a pass ethically and still insist that his behavior ought to be against the rules in future years, and that in the present year he (accidentally) invalidated the Challenge.  Clearly it undermines the result of any Challenge game if someone has done something that prevents the players from playing their best, no matter what was intended.

But my final argument is that the degree of harm is also relevant.  Suppose that, instead what actually happened, there had been no financial arrangements on the side, and lightvector had tried to make browni3141 play worse by calling him a hamster.  We could agree that this is unethical behavior.  We could agree to change the rules such that for future years anyone who engages in taunting is ineligible to win any prize.  We could agree that taunting is likely to adversely affect browni3141's play, so lightvector's chances of winning the Challenge were increased by his nefarious behaviour.  But would we therefore say the games didn't count, and the Challenge hadn't been won, because of the hamster taunt?  I think not.

Returning to what actually happened, my opinion is that the harm was real, but insufficient to invalidate the Challenge.  After lightvector took away the monetary incentive that I had given browni3141, browni3141 still had all the incentives he had before my bet, all the incentives that Omar expected when he chose browni3141 as a Challenge Defender, all the incentives that most Challenge Defenders have ever had.  Clearly these incentives are sufficient for a valid Challenge.

Lightvector (or I, depending on "which rook") may have tarnished the Challenge, but it is still plenty shiny in my book.  Sharp won, sharp deserved to win, and I'm glad I'm not the only one who is going to be paying out the promised prize.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 12th, 2015, 3:09am

on 05/11/15 at 18:28:44, Fritzlein wrote:
It has been pointed out that in other sports, participants are not allowed to place any bets on their games, even betting on themselves to win.  Yet it is quite common to have bonuses for achievements such as rushing 1000 yards or hitting 30 home runs.  It is routine for a player to take the field knowing he will be thousands of dollars better off for performing well than for performing poorly, and indeed sports contracts are written to heighten rather than diminish this dynamic.  Therefore financial incentives per se can't be the reason for the gambling ban.  I am not sure what the reason is, but I think the reason needs to be elucidated before it can be clear that it applies to Arimaa as well, and in particular to the Challenge.


Betting on yourself can pollute the market for other bettors, giving an unfair advantage to those who know how you're betting. (http://www.nytimes.com/2014/11/16/magazine/what-if-an-athlete-wants-to-bet-on-himself.html)

Suppose last September I discovered that Ben Roethlisberger bet $5000 that his Steelers would beat the Panthers, but the next week was only betting $50 that they'd beat the Bucs. I would have naturally inferred that, despite the Bucs looking like an easier opponent than the Panthers and the Steelers being 9 point favourites, Roethlisberger was not so confident. I would have capitalized on my insider information by betting on the Bucs to upset the Steelers, and I would've made a boatload of money.

I can see a few salient differences between betting and performance bonuses:

1) Performance bonuses don't vary by opponent. If Roethlisberger negotiates into his contract a bonus for passing for 300+ yards against the Panthers, but not against the Bucs, this could be a problem. If he gets a bonus for passing for 300+ yards against any opponent, that's fine.

2) Unlike bets, performance bonuses are negotiated far in advance of any games being played. Thus, they may reflect a player's general level of confidence, but they won't reflect a player's level of confidence for any particular game.

3) Information about performance bonuses is publicly available. All bettors know what players are incentivised to do. There is no insider information.

So I think it makes sense for any sport with a betting market to prohibit its players from betting on themselves or their teams.

One question remains: is Arimaa's betting market substantial enough for us to worry about people with insider knowledge having an unfair advantage?

If we answer yes, I see two obvious courses of action (though I'd be happy to hear about others):

1) Prohibit players from betting on games they're playing in.

2) Let players bet on themselves, but post all bets publicly. If this rule were instituted in my Roethlisberger example, everyone would have known he wasn't confident the Steelers would beat the Bucs. I would not have had information other bettors weren't privy to and the betting lines would have been adjusted to account for the increased probability of a Bucs victory, making my bet once again a tossup. This could, however, have the unfortunate and ironic side effect of driving betting underground, so I'm not convinced it's our best option.

Of course, if we adopt one of these rules, it could only apply to future bets.

Other than this, I think your judgement is quite convincing, Fritz. I'm sure you would make quite a good lawyer. But thank you for not becoming one. We have plenty as it is.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PotatoeTheCat on May 12th, 2015, 7:36am
If the Arimaa community wants to move on from this unfortunate incident, the first step must be to clarify its policy on side-betting. (Especially for the WC, CC and other high-profile events, with prize money at stake.)

Craggy's principle 1 "Prohibit players from betting on games they're playing in" is a good place to start from.

For reference, the Olympic Committee tightened up its ethics code in preparation for the London Olympics:
http://www.olympic.org/Documents/Commissions_PDFfiles/Ethics/code-ethique-interactif_en_2013.pdf

Side-betting is specifically covered (and prohibited) - please see pp.80-84.  Especially note Article 5 (pp.83-84) which specifies the outcome of the competition and performance/effort of the player concerned is not material to whether a breach has been committed.

I'll leave it to Fritz to write to the Olympic Committee to demand a rationale for their side-betting ban.  :)

Seriously though, the Arimaa community could spend a lot of time debating  & drafting its own policy on side-betting.  But it certainly could do a whole lot worse than following the lead of (arguably) the most senior sporting body in the world.

Surely if a code of conduct is good enough for the Olympic Games, it's good enough for the Arimaa world.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 12th, 2015, 2:44pm

on 05/12/15 at 07:36:11, PotatoeTheCat wrote:
If the Arimaa community wants to move on from this unfortunate incident, the first step must be to clarify its policy on side-betting. (Especially for the WC, CC and other high-profile events, with prize money at stake.)

Craggy's principle 1 "Prohibit players from betting on games they're playing in" is a good place to start from.

For reference, the Olympic Committee tightened up its ethics code in preparation for the London Olympics:
http://www.olympic.org/Documents/Commissions_PDFfiles/Ethics/code-ethique-interactif_en_2013.pdf

Side-betting is specifically covered (and prohibited) - please see pp.80-84.  Especially note Article 5 (pp.83-84) which specifies the outcome of the competition and performance/effort of the player concerned is not material to whether a breach has been committed.

I'll leave it to Fritz to write to the Olympic Committee to demand a rationale for their side-betting ban.  :)

Seriously though, the Arimaa community could spend a lot of time debating  & drafting its own policy on side-betting.  But it certainly could do a whole lot worse than following the lead of (arguably) the most senior sporting body in the world.

Surely if a code of conduct is good enough for the Olympic Games, it's good enough for the Arimaa world.


As far as I remember Craggy in his last post is the only one to give a good reason for prohibiting betting, and prohibiting betting is not the only course of action as he mentioned in his post.

Appealing to outside authorities is not valid reasoning to ban side-betting.

I still see no good reason to ban betting without conflict of interest instead of making bets mandatorily public, for example, and no one has provided a good reason. Personal morals alone is not a good reason if the large majority of other people don't share your morals and "other organizations do it" is not a good reason.

So I would like to know your rationale for wanting a side-betting ban. We need reasons to change the rules rather than leave them as they are.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PotatoeTheCat on May 13th, 2015, 9:47am
Both my argument for a ban on Side-Betting (highest level sporting authorities support the ban) and the counter-argument ("we need reasons to change the rules") are appeals to convention.  It isn't any more logical (or less logical) to migrate the Arimaa code of conduct to that of wider sporting/gaming bodies than it is to continue with existing practice in Arimaa.  Logic does not help us out here.

However, I am not convinced that Arimaa has a "convention" allowing Side-Betting.  As a number of contributors have noted, there is no written rule on Side-Betting.  The most that can be said is that Side-Betting has actually taken place, and no one has strongly objected (previously).  This is not the same thing as a rule or convention, which would require consensus that the practice is acceptable.  A 6-page, and (at times) emotive forum thread, should be enough to convince anyone that no consensus currently exists on this topic.

The introduction of the OC code of ethics would not "change" the rules on Side-Betting.  The heart of the problem is that Arimaa has no rule on Side-Betting.  That is what needs to change.

Regarding the rationale for the ban on Side-Betting, I am sure that I have nothing original to add on top of what a number of authorities have written and concluded on this matter.

These include:

- the Olympic Committee,
- the EU Athletes Association,
- the European Gaming and Betting Association,
- the Remote Gambling Association,
- the European Sports Security Association

Anyone who feels that the rationale for the ban needs to be spelt out further is most welcome to do some further research.



 

Title: Re:  2015 Arimaa Challenge Match results inva
Post by browni3141 on May 13th, 2015, 5:19pm

on 05/13/15 at 09:47:10, PotatoeTheCat wrote:
Both my argument for a ban on Side-Betting (highest level sporting authorities support the ban) and the counter-argument ("we need reasons to change the rules") are appeals to convention.  It isn't any more logical (or less logical) to migrate the Arimaa code of conduct to that of wider sporting/gaming bodies than it is to continue with existing practice in Arimaa.  Logic does not help us out here.

However, I am not convinced that Arimaa has a "convention" allowing Side-Betting.  As a number of contributors have noted, there is no written rule on Side-Betting.  The most that can be said is that Side-Betting has actually taken place, and no one has strongly objected (previously).  This is not the same thing as a rule or convention, which would require consensus that the practice is acceptable.  A 6-page, and (at times) emotive forum thread, should be enough to convince anyone that no consensus currently exists on this topic.

The introduction of the OC code of ethics would not "change" the rules on Side-Betting.  The heart of the problem is that Arimaa has no rule on Side-Betting.  That is what needs to change.

Regarding the rationale for the ban on Side-Betting, I am sure that I have nothing original to add on top of what a number of authorities have written and concluded on this matter.

These include:

- the Olympic Committee,
- the EU Athletes Association,
- the European Gaming and Betting Association,
- the Remote Gambling Association,
- the European Sports Security Association

Anyone who feels that the rationale for the ban needs to be spelt out further is most welcome to do some further research.


I am not arguing for any new rules to be put in place to explicitly allow side betting. I am just asking for some logical reasoning for a new rule to be put in place which is going to restrict people's freedoms. Why is that so much to ask for?

Your argument is fallacious, unless you're not trying to make a logical argument at all, but who here is going to respect an argument without any logical reasoning whatsoever? "Some well-known organizations do this, so we should to."
If you can find sources explaining the reasoning behind the rules of these organizations or if you can supply your own reasoning for these rules then your argument can have value, but you seem to be either lazy, apathetic or unable.
Don't expect others to support your argument for you or maybe just don't make it in the first place if you don't care about following through.

I will be happy to continue if you or anyone else is willing to present a stronger logical argument to introduce rules regarding side bets, but right now I'm just getting frustrated with these barely supported propositions to introduce new rules (with the exception of Craggy)

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 13th, 2015, 10:23pm
I thought of a counter to my argument that betting* on yourself should be regulated or prohibited because it pollutes the betting market. (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi?board=events;action=display;num=1430331089;start=75#76)

Betting on yourself can be a problem when it tips off some but not all other bettors about your confidence level. But this is not the only thing that can tip off people about a player's confidence level.

Heading into the finals of this year's World Championship, Fritz said several times in chat that he believed his chances of beating Browni were low. Suppose I capitalized on this information by betting on Browni with someone who wasn't present in chat at the time and had no idea about Fritz's low confidence. Would that have been ethical?

There are two actions to scrutinize:

1) Fritz divulging his confidence level.

2) Me capitalizing on my insider information by betting with someone who didn't have access to this information.

Let's first address Fritz's action. Should we really have censured him simply for talking about an upcoming game? Should we have said, "Shut up, Fritz. You're polluting the betting market"?

I think it's unreasonable to restrict players' speech this way. No one should bite their tongue out of fear they'll slightly compromise the integrity of the betting market.

Yet if we refuse to prohibit players from talking about their confidence levels, on what grounds can we prohibit them from tipping people off about their confidence levels via their bets? My insider information would have been just as valuable in either case, and the betting market would have been just as polluted.

Perhaps instead of prohibiting players from talking about their confidence levels, we should allow it only if they post about it in the forum. This would negate the advantage of my "insider" information by making it available to all bettors.

Unfortunately, this also restricts freedom of expression, though indirectly. Players are far less likely to chat about their feelings about upcoming games if they have a burden to also post about it in the forum.

So if we refuse to make public all information about players' confidence levels, how can we justify making public a subset of this information, bets?

I've shown we can't have a pure betting market without imposing excessive restrictions on our freedom of expression. But I believe there are two reasons we should continue to let players talk about whatever they like, yet single out betting for regulation:

1) Freedom of expression is more important than the freedom to bet. I would like the betting market to be as pure as possible, but not at any cost. Regulating speech is too high a cost. Regulating bets is a reasonable cost.

2) When players talk about how they feel about upcoming games, the risk of corruption is much lower than when they place bets. An increased ability to dissuade and catch corruption is, I believe, the biggest virtue of making bets public. LightVector's agreement with Browni wasn't corrupt, but it looked odd enough to warrant a few questions. If and when an actual case of corruption arises, I want us to have the tools we need to discover it quickly so we can prevent it from inflicting any damage on Arimaa.

Now let's consider the ethics of me capitalizing on my insider information about Fritz's low confidence. I think we need to recognize that, unlike in pro sports, in Arimaa players and bettors are in very close contact. In fact, all of the bettors are also players.

In pro sports, there's a natural barrier between players and bettors. In Arimaa, people are unlikely to bet on a game between players they don't know. If we choose to prohibit betting on games when a bettor has a piece of insider information about one of the players, we'll have to prohibit most of our bets. In any community as small and cozy as Arimaa, people are going to know stuff about each other, and that information won't be distributed evenly.

Perhaps the close relationship between players and bettors generates so much insider information that my desire for a pure-ish betting market is laughable. I'm beginning to think that my initial argument (http://arimaa.com/arimaa/forum/cgi/YaBB.cgi?board=events;action=display;num=1430331089;start=75#76) only applies when the betting community is so separated from the players that information about how a player bets is a rare insight into their confidence level. In Arimaa, these insights are not rare at all.

So unless someone can resuscitate my argument for maximizing the integrity of the betting market, the only reason I can think of for regulating bets is that it will dissuade corruption and make it easier to catch. In particular, I think making all bets public would help catch unintended corruption, which could occur in an agreement similar to LightVector and Browni's.


*By "betting," I mean any proposed exchange of money or goods conditional on a future event. By some definitions, LightVector's agreement with Browni wasn't a bet since there was no scenario where Browni would have had to pay out. But I think it's clear that if we decide to regulate bets, we should regulate these kinds of agreements in the same way.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by SilverMitt on May 14th, 2015, 12:12am
The last post seems entirely a non sequitur to me.  Do we have a betting market I don't know about?

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by 99of9 on May 14th, 2015, 1:25am
One reason that sporting bodies do well to prohibit betting on oneself is to prevent sharking (sandbagging your apparent ability in previous rounds and then cashing in with a big bet on yourself).

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by PotatoeTheCat on May 14th, 2015, 2:27am
An argument from convention is not a fallacious argument: it is not a logical argument at all.  But that does not mean that we can discount convention.

There is a strong argument that the practice of attempting to settle questions by reason is itself a social construct (originating from 5th century BC Athens).

[ http://plato.stanford.edu/entries/convention/

The seminal work on this subject is by David Lewis (no relation): "Convention" (1969)]

Why is it "better" to try to decide matter by reason, rather than say, trial by combat, or just voting on the matter without discussion?  (Don't say "its more logical", because that's a circular argument.)

Honestly, does anyone think this forum thread (engaging though it is) is more likely to yield a satisfactory answer to the issue of Side-Betting, than say an Arimaa Blitz tournament to decide the matter?

@Browni

You stated above that "You seem to be either lazy, apathetic or unable [to explain the reasoning behind the rules of these organizations]"

Actually, I believe I am none of these.  The rationale is on the websites I have provided links for.  Anyone can read these and comment on them, just as they wish.

However, there is a convention (actually a written rule) which prohibits discourtesy to forum participants:

http://arimaa.com/arimaa/forum/cgi/YaBB.cgi?board=talk;action=display;num=1284316699

I believe your comments above breach that rule, so I decline to contribute further until normal standards of civility are resumed.

Signing off - PTC

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by mattj256 on May 14th, 2015, 9:01am

on 05/14/15 at 00:12:45, SilverMitt wrote:
The last post seems entirely a non sequitur to me.  Do we have a betting market I don't know about?
I agree with SilverMitt.  The recent part of this discussion is talking about the solution to a problem that doesn't exist.

For what it's worth, MY main concern is that Arimaa events don't get canceled or invalidated, and the Tournament Director of an Arimaa event isn't placed in an awkward position.  Therefore I suggest the following rules:

1. For non-event games, all betting is legal and unregulated.  The only exception is that a player may not bet against himself or herself.  (Each bet is considered in isolation, and this would make Browni's behavior illegal going forward.)

2. Every new tournament or Arimaa event must have an explicit policy regarding betting.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by CraggyCornmeal on May 14th, 2015, 10:59am

on 05/14/15 at 00:12:45, SilverMitt wrote:
The last post seems entirely a non sequitur to me.  Do we have a betting market I don't know about?

Sometimes people bet on games. Therefore, there's a betting market.


on 05/14/15 at 09:01:08, mattj256 wrote:
I agree with SilverMitt.  The recent part of this discussion is talking about the solution to a problem that doesn't exist.

Arimaa may not currently have any corruption, but it is a very real danger, and should be the focus of any rule that regulates betting.


on 05/14/15 at 09:01:08, mattj256 wrote:
1. For non-event games, all betting is legal and unregulated.

I'm not convinced that non-event games are less susceptible to corruption than event games.


on 05/14/15 at 09:01:08, mattj256 wrote:
2. Every new tournament or Arimaa event must have an explicit policy regarding betting.

This would repeatedly reopen the debate, which would be redundant and tiring.


on 05/14/15 at 02:27:30, PotatoeTheCat wrote:
An argument from convention is not a fallacious argument

I disagree. (http://rationalwiki.org/wiki/Appeal_to_tradition)


on 05/14/15 at 02:27:30, PotatoeTheCat wrote:
Why is it "better" to try to decide matter by reason, rather than say, trial by combat, or just voting on the matter without discussion?  (Don't say "its more logical", because that's a circular argument.)

Because, historically, deciding matters with reason has tended to bring us closer to the truth. Deciding matters with trial by combat has tended to pile up bodies.


on 05/14/15 at 01:25:54, 99of9 wrote:
One reason that sporting bodies do well to prohibit betting on oneself is to prevent sharking (sandbagging your apparent ability in previous rounds and then cashing in with a big bet on yourself).

Great point, 99of9.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by browni3141 on May 15th, 2015, 1:32am

on 05/14/15 at 02:27:30, PotatoeTheCat wrote:
@Browni

You stated above that "You seem to be either lazy, apathetic or unable [to explain the reasoning behind the rules of these organizations]"

Actually, I believe I am none of these.  The rationale is on the websites I have provided links for.  Anyone can read these and comment on them, just as they wish.

However, there is a convention (actually a written rule) which prohibits discourtesy to forum participants:

http://arimaa.com/arimaa/forum/cgi/YaBB.cgi?board=talk;action=display;num=1284316699

I believe your comments above breach that rule, so I decline to contribute further until normal standards of civility are resumed.

Signing off - PTC


Sorry if I offended you. My phrasing could be seen as antagonistic, perhaps, but I stand by what I said. I would appreciate it if you could post a more direct link or quote which parts of their rationale you think may apply to Arimaa. I see guidelines in your links, but not reasoning behind them which would apply to Arimaa, IMO.

Title: Re: 2015 Arimaa Challenge Match results invalid
Post by lightvector on May 15th, 2015, 10:24pm
Just to chime in briefly - I would support and be happy to see a policy around placing or publicizing of bets or financial arrangements, whatever it may be, for future events.

As far as the details of what that policy is - one can obviously run particular analyses of whether and in what situations certain kinds of bets might result in adverse incentives, and argue about the exact form of the rule. But all that aside, it seems to me that one of the most important considerations is simply - what rule would let as many other people as possible be comfortable and able to move on with future events?

Taking a step back, I would guess most of us are here really just to play and have fun. I don't know about how other people feel, but I can't imagine that it's fun to keep debating this, or that it feels like a terribly valuable use of time. From that stance, it seems obvious to me that any concessions necessary are worth it to come to agreement and move on. A straightforward restriction against on betting or staking money on future event games, for example, seems reasonable to me in that regard, and if that's what it takes, at least personally I would be happy if that's people decide on.

Regardless of what happens here and regardless of any rules that are established, I know that I personally will be considering my actions more carefully in the future. Because while different people have different perceptions of right and wrong, the perceptions themselves are real and can cause an action to have real consequences on other people.


Title: Re: 2015 Arimaa Challenge Match results invalid
Post by Arimabuff on Jun 19th, 2015, 10:34am
What's the final word on all that?



Arimaa Forum » Powered by YaBB 1 Gold - SP 1.3.1!
YaBB © 2000-2003. All Rights Reserved.