Author |
Topic: 2009 World Championship (Read 9055 times) |
|
Adanac
Forum Guru
Arimaa player #892
Gender: 
Posts: 635
|
 |
Re: 2009 World Championship
« Reply #30 on: Jan 13th, 2009, 5:03am » |
 Quote  Modify
|
on Jan 12th, 2009, 8:38pm, Fritzlein wrote:
But you are discounting the chance of losing. Yes, playing woh will get me better tiebreak points than you will get for your forfeit, but I also survived a chance that I wouldn't even get the win. |
|
Imagine, though, that you had lost a tough 1st round game and then your opponent withdrew from the tournament in the next round. Then you'd get 0 wins and only 1 SoS tie-breaker point despite having been paired up with a difficult opponent. That's where the current system is very unfair, and that has been known to happen in the past.
Or, what if it's the final round of the tournament and you're tied for the eighth spot with another player. The tie-breaker will be decided in the final game of the tournament between a pair of 1-3 or 0-4 players. But maybe the player you need to win doesn't bother to show up because it's a "meaningless" game for him. Or perhaps that player shows up knowing the stakes involved, but wilts under the pressure/obligation. Not everyone likes to be placed in that situation. With my proposal you already know your SoS and the games that affect your final standing will generally involve games between 2-2 and 3-1 players heading into the fifth round.
The current system is excellent if we can know for sure that every player will play every game without forfeit (and always to the best of their ability in the final round). The wild unpredictability going into the final round can even be thought of as an advantage. But with all of the forfeits plaguing the past 2 Open Classics, it just seems too unfair to punish people for being paired against players that forfeit -- especially if the forfeit occurs later in the tournament.
|
|
 IP Logged |
|
|
|
RonWeasley
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: 2009 World Championship
« Reply #31 on: Jan 13th, 2009, 5:31am » |
Quote Modify
|
on Jan 12th, 2009, 9:41am, Fritzlein wrote:| Apparently thefrankinator, Amina, and Emaad all forfeited. According to the rules a "player who forfiets a game is automatically removed from the tournament and will not play in future rounds unless the tournament directors accepts the appeal". Ron, have any of the players appealed to you to continue the tournament? My feeling is that just asking to continue is enough reason for an appeal to be accepted, but the rules clearly leave it to the discretion of the Tournament Director whether or not to allow forfeiting players to continue. |
|
So far, I haven't received any appeals. If players are confused about how to do this, a forfeiting player could send me a private message in this forum, ask Omar for my e-mail through the game site, or even post a message on this forum. If this is too hard, post a message in this forum that you are trying to contact me. What I'm looking for is an assurance that the player will make a good-faith attempt to play the remainder of your games. First offenders will tend to get off easy. But if the player doesn't appeal, I must assume that player has abandoned the tournament.
|
|
IP Logged |
|
|
|
RonWeasley
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: 2009 World Championship
« Reply #32 on: Jan 13th, 2009, 6:11am » |
Quote Modify
|
on Jan 12th, 2009, 9:41am, Fritzlein wrote:| The tiebreaker points have not been explicit, but my understanding is that byes and forfeit wins should earn zero tiebreaker points, no matter which round they occur in. |
|
When I read the current rules, they tell me that a forfeit is scored as a normal game with the forfeiting player losing. If a player leaves the tournament, that player's score (number of wins) is used in the sum of scores for that player's opponents. If that player plays no more games, there is still a score for that player.
I don't see any language about receiving a zero for that player's score. Please correct me if I'm wrong here, but I see the rules describing the tie breaker as the sum of the opponents' scores. In spite of the weaknesses of this procedure that have been described, we won't change this procedure unless a compelling alternative is presented. By that I mean indisputable evidence that the alternative is better. That hasn't happened.
I support discussing an improved way to account for withdrawals and would consider a change in this tournament if a timely proposal is made. Realistically, I expect this discussion to have an effect on next year's rules. One objection I have to getting a zero for a forfeit is that it simply increases the penalty one gets for being unlucky enough to have that opponent. It's bad enough the that opponent might leave the tournament and not increase his score.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2009 World Championship
« Reply #33 on: Jan 13th, 2009, 7:34am » |
Quote Modify
|
on Jan 13th, 2009, 6:11am, RonWeasley wrote:| I don't see any language about receiving a zero for that player's score. |
|
I don't think anyone is proposing that. A forfeit win is worth one point. What we are discussing is purely how the forfeit win affects one's strength of schedule.
Quote:| In spite of the weaknesses of this procedure that have been described, we won't change this procedure unless a compelling alternative is presented. |
|
I'm comfortable with that. The discussion can pertain to next year.
Quote:| One objection I have to getting a zero for a forfeit is that it simply increases the penalty one gets for being unlucky enough to have that opponent. |
|
Increases the penalty? What penalty? If there is no difference between a forfeit win and a regular win, in terms of strength of schedule, then the luckiest you can be is to get a forfeit win in every round.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2009 World Championship
« Reply #34 on: Jan 13th, 2009, 7:39am » |
Quote Modify
|
on Jan 13th, 2009, 5:03am, Adanac wrote:| Imagine, though, that you had lost a tough 1st round game and then your opponent withdrew from the tournament in the next round. Then you'd get 0 wins and only 1 SoS tie-breaker point despite having been paired up with a difficult opponent. That's where the current system is very unfair, and that has been known to happen in the past. |
|
Yes, the current tiebreakers become unfair when someone withdraws in the middle of the tournament. I don't really have a good solution for that. But I do think it is better than giving different people different tiebreaker points for playing the same opponent.
In any case, my proposal was about changing the way tiebreaker points are given that for forfeit wins, which I think is a slight improvement even though it doesn't address the problem of tournament withdrawals.
|
|
IP Logged |
|
|
|
RonWeasley
Forum Guru
Harry's friend (Arimaa player #441)
Gender:
Posts: 882
|
|
Re: 2009 World Championship
« Reply #35 on: Jan 13th, 2009, 9:17am » |
Quote Modify
|
on Jan 13th, 2009, 7:34am, Fritzlein wrote:
Increases the penalty? What penalty? If there is no difference between a forfeit win and a regular win, in terms of strength of schedule, then the luckiest you can be is to get a forfeit win in every round.
|
|
The current penalty is that of a forfeiting player being likely to also leave the tournament, or to forfeit again. This means that player's contribution to strength of schedule may be less, but never greater, than it would have been because that player gets no more wins.
|
|
IP Logged |
|
|
|
grey_0x2A
Forum Full Member
Arimaa player #1358
Gender:
Posts: 14
|
|
Re: 2009 World Championship
« Reply #36 on: Jan 13th, 2009, 9:29am » |
Quote Modify
|
The problem of tie-breaks, is always a major issue in swiss tournaments. Every game than I have played seriously has this issue. Esp. in these small 8-32 player size where you can play every game well and still get hooped by the system. Once there where two players with one lost at the end of the turnament but I came second to the player I beat, my first round pairing played poorly. In larger tournaments getting done in by tie-breaks is less of an issue if you are at the top in my experience, and if your on the bubble luck is a major factor either way so who cares. BTW if SOS is the first tie break what is the second and third?
|
|
IP Logged |
|
|
|
Adanac
Forum Guru
Arimaa player #892
Gender:
Posts: 635
|
|
Re: 2009 World Championship
« Reply #37 on: Jan 13th, 2009, 10:08am » |
Quote Modify
|
on Jan 13th, 2009, 7:39am, Fritzlein wrote:
Yes, the current tiebreakers become unfair when someone withdraws in the middle of the tournament. I don't really have a good solution for that. But I do think it is better than giving different people different tiebreaker points for playing the same opponent.
|
|
I don't believe the "common opponents" would make a difference in practice:
If player A's W/L by round are: L-W-W-L-W
and player B: W-W-L-L-W
then by my proposed system player B should have a slightly better Strength of Schedule, usually by an SoS score of 7-5 **. In fact, player B should win by that same margin regardless of whether or not they had any common opponents, or which order they faced their common opponents.
On the other hand, suppose player B's first round opponent withdrew in the next round. Now, player A may pass ahead of player B even though he followed the path of lesser resistance and player B was punished for events out of his control.
My system doesn't address the unfair advantage of winning and yet receiving the same credit as a player that had to fight for a win. But that's out of our control, much like winning because the opponent disconnected in a close game and shouldn't be penalized in any way.
** the scores may differ if either player got bumped or down in any round due to an odd number of players. But if so, that change in SoS score is deserved.
|
| « Last Edit: Jan 13th, 2009, 10:15am by Adanac » |
IP Logged |
|
|
|
Tuks
Forum Guru
Arimaa player #2626
Gender:
Posts: 203
|
|
Re: 2009 World Championship
« Reply #38 on: Jan 13th, 2009, 11:25am » |
Quote Modify
|
couldn't you do something with the initial ranks like:
Points scale: (ill make it with 10 players)
when adding up wins:
#1 = 10
#2 = 9
#3 = 8
#4 = 7
#5 = 6
#6 = 5
#7 = 4
#8 = 3
#9 = 2
#10 = 1
When subtracting loses
#1 = 1
#2 = 2
#3 = 3
#4 = 4
#5 = 5
#6 = 6
#7 = 7
#8 = 8
#9 = 9
#10 = 10
we could make forfeits worth half the points because they didn't fight for it but it shouldn't be too much of a hindrance
examples:
Player 1: Beats #8,6,10 Loses against #1,4
Score: 3+5+1-1-4 = 4
Player 2: Beats #10,9,4Loses against #2,6
Score: 1+2+7-2-6 = 2
this approach seemed to favor the person with stronger opponents in the loss section although i haven't tested very much
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: 2009 World Championship
« Reply #39 on: Jan 13th, 2009, 3:52pm » |
Quote Modify
|
on Jan 12th, 2009, 9:55pm, camelback wrote:
Thanks Adanac  . I did sit on the table and it displayed "waiting for opponent". I waited for 20 minutes then I went to chatroom.
|
|
Bizarre why that game didn't get saved as a forfeit like the other ones; there may still be some bugs I have to stamp out. But thanks for confirming this, camelback.
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: 2009 World Championship
« Reply #40 on: Jan 13th, 2009, 4:24pm » |
Quote Modify
|
on Jan 13th, 2009, 11:25am, Tuks wrote:couldn't you do something with the initial ranks like:
|
|
Interesting proposal Tuks. This helps with the forfeit issue, but one could say that it puts too much weight on the initial ranks and the current system tried to minimize the dependence on the initial ranks. Perhaps you might want to consider modifying it to rank the players based on their performance with ties broken by initial ranks and then apply your scoring method on that ranking of players rather than the initial ranking.
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: 2009 World Championship
« Reply #41 on: Jan 13th, 2009, 4:27pm » |
Quote Modify
|
on Jan 13th, 2009, 9:29am, grey_0x2A wrote:BTW if SOS is the first tie break what is the second and third?
|
|
Pre-tournament ratings and then random.
|
|
IP Logged |
|
|
|
grey_0x2A
Forum Full Member
Arimaa player #1358
Gender:
Posts: 14
|
|
Re: 2009 World Championship
« Reply #42 on: Jan 13th, 2009, 5:36pm » |
Quote Modify
|
So the lower pre-tournament player gets in right. After all they are the better player "today". IMHO
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2009 World Championship
« Reply #43 on: Jan 13th, 2009, 5:47pm » |
Quote Modify
|
on Jan 13th, 2009, 9:17am, RonWeasley wrote:| The current penalty is that of a forfeiting player being likely to also leave the tournament, or to forfeit again. This means that player's contribution to strength of schedule may be less, but never greater, than it would have been because that player gets no more wins. |
|
Got it. If the winner by forfeit would get no tiebreak points, that would be greater than the penalty of getting potentially reduced tiebreak points.
on Jan 13th, 2009, 10:08am, Adanac wrote:If player A's W/L by round are: L-W-W-L-W
and player B: W-W-L-L-W
then by my proposed system player B should have a slightly better Strength of Schedule, usually by an SoS score of 7-5 **. In fact, player B should win by that same margin regardless of whether or not they had any common opponents, or which order they faced their common opponents. |
|
OK, you totally lost me. I understand that it is unfair in the current system when your strength of schedule is reduced because an opponent you beat later dropped out of the tournament. But I don't understand why common opponents are irrelevant. Just to clarify, if Player A scores L-W-W-L-W and Player B scores W-W-L-L-W, you think that Player B should have better tiebreakers, even if the two players played exactly the same five opponents? To me it seems that if two players play exactly the same five opponents, and your tiebreaker ranks one above the other, then your tiebreaker can't be called strength of schedule.
In fact, your tiebreaker is very close to what we used to call "cumulative wins" in chess tournaments. As I understand it, that's how swiss-paired chess tournaments used to break ties in the 1970's (at least in Kansas) until the system was displaced by "sum of opponent's wins", which was generally regarded as being more fair. However, it may be that live chess tournaments don't have as much trouble with people withdrawing mid-tournament as the Arimaa World Championship does.
Quote:| On the other hand, suppose player B's first round opponent withdrew in the next round. Now, player A may pass ahead of player B even though he followed the path of lesser resistance and player B was punished for events out of his control. |
|
Yes, this is unfair. If possible, I would like to separate that fairness issue from the issue of whether a win by forfeit should receive any tiebreak points. It is possible that, without addressing the unfairness caused by withdrawals, my proposal would make the current system fairer.
Quote:| My system doesn't address the unfair advantage of winning and yet receiving the same credit as a player that had to fight for a win. But that's out of our control, much like winning because the opponent disconnected in a close game and shouldn't be penalized in any way. |
|
Wait, are you saying that fairness is not the standard for how rewards should be assigned, and the standard is rather what is within a player's control? Or are you saying that it is less unfair to reward a player for something out of his control than it is to penalize him for something out of his control? The latter distinction makes no sense to me, because an unfair reward to one player is equivalent to an unfair penalty to all the other players. My thought is that we try to do the fairest thing possible under adverse circumstances. Giving credit for a win the player did nothing to earn is an unfair reward, while giving the winner no chance at tiebreak points is an unfair penalty. Perhaps it balances out.
on Jan 13th, 2009, 11:25am, Tuks wrote:Player 1: Beats #8,6,10 Loses against #1,4
Score: 3+5+1-1-4 = 4
Player 2: Beats #10,9,4Loses against #2,6
Score: 1+2+7-2-6 = 2 |
|
That proposal is equivalent to adding the seeds of your opponents, and saying that lowest tiebreak is best. A similar approach would be to say that the tiebreak is the sum of the ratings of ones opponents, with highest being best.
I am a huge fan of strength of schedule for the tiebreaker. The trouble is accurately measuring strength of schedule. I submit that "wins of opponents" measures strength of schedule better than "cumulative wins", even given the problem of people withdrawing mid-tournament. However, it may be that "sum of seeds" or "sum of ratings" would measure strength of schedule more accurately than either Adanac's proposal or mine. It seems like a question of which source of unfairness we least abhor.
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
 
Gender:
Posts: 1413
|
|
Re: 2009 World Championship
« Reply #44 on: Jan 13th, 2009, 7:36pm » |
Quote Modify
|
on Jan 13th, 2009, 5:47pm, Fritzlein wrote:| I would like to separate that fairness issue from the issue of whether a win by forfeit should receive any tiebreak points. It is possible that, without addressing the unfairness caused by withdrawals, my proposal would make the current system fairer. |
|
The problem is, your proposal makes the unfairness caused by withdrawals worse (as you've agreed with Ron).
So you can't really split up the two issues. In my opinion the question is whether making your change gives an overall increase in fairness. I agree that it might in a tournament with forfeits but without withdrawals. But in a tournament with forfeits and withdrawals (especially in a tournament where forfeit implies withdrawal...), it is far from clear.
|
|
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print