Author |
Topic: 2011 WC - Commentary and Reporting (Read 12668 times) |
|
The_Jeh
Forum Guru
Arimaa player #634
Gender: 
Posts: 460
|
 |
Re: 2011 WC - Commentary and Reporting
« Reply #60 on: Jan 25th, 2011, 8:43pm » |
 Quote  Modify
|
on Jan 25th, 2011, 6:48pm, Fritzlein wrote:There seems to be a problem with the commentary spreadsheet. The time slot 116 (for naveed vs. megamau and The_Jeh vs. chessandgo) corresponds to 1:00 Monday, but the games are scheduled a day earlier in the game room, for 1:00 Sunday. I am afraid this problem is not actually a bug in the spreadsheet, but rather a bug in the way selected time slots are converted into scheduled games in the game room.
P.S. There seems to be some kind of discontinuity around Friday, whereby games scheduled after Friday have all been moved 24 hours earlier. |
|
I hope it's a spreadsheet problem, since I certainly did not select any times on Monday.
|
|
 IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #61 on: Jan 25th, 2011, 9:37pm » |
Quote Modify
|
on Jan 25th, 2011, 8:43pm, The_Jeh wrote:| I hope it's a spreadsheet problem, since I certainly did not select any times on Monday. |
|
Oh, I guess it is a spreadsheet problem. Omar explained to me that the spreadsheet numbers didn't come from the scheduler, but rather were entered manually. So I'll just fix them manually and all will be well.
|
|
IP Logged |
|
|
|
99of9
Forum Guru
Gnobby's creator (player #314)
 
Gender:
Posts: 1413
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #62 on: Jan 26th, 2011, 12:06am » |
Quote Modify
|
on Jan 25th, 2011, 2:58pm, omar wrote:| The preliminary games have T=4m so the moves of the game will appear 12 minutes after they were made. |
|
I agree that synchronization is an improvement, but I think 12 minutes is a bit extreme. Before we only had to skip through about 4 minutes worth of moves once the game was over. Now we will either have to try to immediately wrap up 12 minutes of material if the remaining moves are instantly shown, or the players will have to sit around for 12 minutes if they want to discuss the game with spectators. That's a long wait in the middle of the night. I'd prefer 2*T or a fixed 5 minutes.
|
|
IP Logged |
|
|
|
rbarreira
Forum Guru
Arimaa player #1621
Gender:
Posts: 605
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #63 on: Jan 26th, 2011, 7:58am » |
Quote Modify
|
on Jan 25th, 2011, 2:58pm, omar wrote:I've made changes on the server to implement time based delay of moves visible to the spectators.
For rated event games, the new implementation will delay the game moves by 3*T if the T parameter is available and by 2 moves (4 ply) if it is not available. Postal games will always be delayed by 2 moves. It works OK in the JavaScript and Flash V2 client. It kind of works in the Flash V1 client, but the timers are off.
Since the WC games have a maximum time per turn parameter, they will be delayed by time rather than by 2 turns. The preliminary games have T=4m so the moves of the game will appear 12 minutes after they were made. When the game ends the remaining moves don't all appear at once and continue to be time delayed.
Now we will be able to tell how long the players thought on each move. Also audio recordings can later be synced with the client running in auto-play mode.
|
|
That's really nice (though I agree with 99of9 that the delay could be smaller).
Any chance of getting a similar time delay for the Computer Championship, or even eliminate the delay completely for bot vs bot games? There's no maximum turn time in the CC time control so I guess it would still use the 2-move delay for now.
|
| « Last Edit: Jan 26th, 2011, 8:01am by rbarreira » |
IP Logged |
|
|
|
megajester
Forum Guru
Istanbul, Turkey
Gender:
Posts: 710
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #64 on: Jan 26th, 2011, 5:18pm » |
Quote Modify
|
Due to time differences most of the games this round will be in the small hours of the morning for me. Also I'm not around this weekend, so I'm afraid I'll be of no use this round for commentary/anchoring. FYI
|
|
IP Logged |
|
|
|
omar
Forum Guru
Arimaa player #2
Gender:
Posts: 1003
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #65 on: Jan 27th, 2011, 2:30pm » |
Quote Modify
|
Toby, I thought that 3*T would be the least delay to ensure that even if the players had lots of reserve and used the maximum time per turn there would still be about a two move gap between the move the player is working on and the move being commented about.
After the game is over, the moves continue to be delayed and are not all shown at once.
Ricardo, bot vs bot games won't be delayed now. Thanks for the suggestion.
|
| « Last Edit: Jan 27th, 2011, 2:32pm by omar » |
IP Logged |
|
|
|
Adanac
Forum Guru
Arimaa player #892
Gender:
Posts: 635
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #66 on: Jan 27th, 2011, 2:57pm » |
Quote Modify
|
on Jan 27th, 2011, 2:30pm, omar wrote:Toby, I thought that 3*T would be the least delay to ensure that even if the players had lots of reserve and used the maximum time per turn there would still be about a two move gap between the move the player is working on and the move being commented about.
After the game is over, the moves continue to be delayed and are not all shown at once.
Ricardo, bot vs bot games won't be delayed now. Thanks for the suggestion.
|
|
Omar, for the finals I assume the delay between moves will increase to 15 minutes? After the Open Classic is finished can we drop the delay down to 2*T? I doubt that anyone finishing in the top 8 will benefit from a 10-minute audio delay and it would reduce the post-game interview delays by 5 minutes.
|
|
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #67 on: Jan 30th, 2011, 4:41pm » |
Quote Modify
|
on Jan 7th, 2011, 10:01am, Fritzlein wrote:| Hmmm, the rules don't say whether we should use the population standard deviation of 319.6455251 or the sample standard deviation of 324.6015652. |
|
Omar's code is dividing by #players+1. I don't understand why, because it should be divided by either #players (population) or #players-1 (sample).
Code:# note that the 'r' field of the player record must be a rating; otherwise this won't work
my(@players, $p, $ef, $avg, $std);
@players = $ev->getPlayers();
$avg = 0;
foreach $p (@players){
$avg += $p->{r};
}
$avg = $avg/($#players + 1);
foreach $p (@players){
$std += ($p->{r} - $avg)**2;
}
$std = sqrt($std/($#players + 1));
$ef = $std/100; |
|
on Jan 7th, 2011, 10:16am, Adanac wrote:| I updated my program yesterday to use 3.746 as the F factor. I agree with you that it probably doesn't matter either way but I'd like to confirm with Omar for consistency purposes (rounding to 3 digits should be close enough?) |
|
Adanac, you are using sample standard deviation. I believe this explains the discrepancy between your SoS and Omar's. Perhaps you can check this although it still leaves open the question of what should be done.
|
| « Last Edit: Jan 30th, 2011, 4:45pm by Fritzlein » |
IP Logged |
|
|
|
aaaa
Forum Guru
Arimaa player #958
Posts: 768
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #68 on: Jan 30th, 2011, 5:25pm » |
Quote Modify
|
"$#" before an array returns the index of its last element and since arrays are zero-based in Perl...
|
|
IP Logged |
|
|
|
chessandgo
Forum Guru
Arimaa player #1889
Gender:
Posts: 1244
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #69 on: Jan 31st, 2011, 10:34am » |
Quote Modify
|
Well, and I thought the 2-mn delay was an over-reaction  More seriously, 12 mn delay (and 15 during the finals)? As far as I can tell, the inconvenience for dozens of spectators during hundreds of games will over-weigh by much the possible inconvenience possibly caused by a possible attempt at cheating through live commentary eavesdropping. That's probably not worth anything but I'm favor of the smallest delay possible.
|
|
IP Logged |
|
|
|
Janzert
Forum Guru
Arimaa player #247
Gender:
Posts: 1016
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #70 on: Jan 31st, 2011, 11:49am » |
Quote Modify
|
I agree with Chessandgo, and find my irritation from the delay growing with each game I watch. I understand some commentators feel constrained and don't want to commentate without it. I wouldn't want to lose any of the commentary; but I really do hope the delay can be lowered from what it is now though.
Janzert
|
| « Last Edit: Jan 31st, 2011, 11:50am by Janzert » |
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #71 on: Jan 31st, 2011, 2:15pm » |
Quote Modify
|
So both of you prefer the previous 2-move (4-ply) delay that we had to the current 12-minute delay? I also prefer the 2-move delay because it is shorter on average, although the synching of thinking times has some benefit too, so I don't have a strong preference between the two.
What I definitely wouldn't like to see is a fixed four-minute delay. If I were giving commentary, it would bug me to know that any time a cheater wanted to know what I thought of his move 17g before he made move 18g, he only has to dip into reserve a bit to hear what I have to say about it. He can choose to get my opinion about his last move before making the present move. That is too much potential benefit for cheating, so I think I would simply refrain from giving commentary in that situation.
Yes, the two-move delay works out to a four-minute delay on average, but the difference is that with a two-move delay a cheater, before hearing what I think about any of his decisions, would have to make two additional decisions, usually making it too late to benefit from my opinion. Having this buffer matters far more to me than having the displayed thinking times synch up with the actual thinking times.
|
| « Last Edit: Jan 31st, 2011, 2:18pm by Fritzlein » |
IP Logged |
|
|
|
Nombril
Forum Guru
Arimaa player #4509
Gender:
Posts: 292
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #72 on: Jan 31st, 2011, 2:24pm » |
Quote Modify
|
It seems 2*T (for the preliminaries at least) would be long enough to satisfy the cheating concerns, and would lessen the wait for post game interviews.
Actually, the wait didn't bother me, there were some benefits. I found it very interesting to hear the last minutes of commentary. It also gives a chance for post game interviews to be more collected, as I also had a chance to quickly review the game.
|
|
IP Logged |
|
|
|
The_Jeh
Forum Guru
Arimaa player #634
Gender:
Posts: 460
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #73 on: Jan 31st, 2011, 2:30pm » |
Quote Modify
|
Forgive my computer illiteracy if this is too complicated to implement, but would it be possible to use a short time delay that then pauses momentarily after a move is displayed if the actual game has not moved at least two moves ahead? (Heh, seems like we're always demanding more and more sophistication. I actually was okay with the two-move delay. It allowed me to see the game without fear of cheating while still having a live connection to it.)
|
| « Last Edit: Jan 31st, 2011, 2:49pm by The_Jeh » |
IP Logged |
|
|
|
Fritzlein
Forum Guru
Arimaa player #706
Gender:
Posts: 5928
|
|
Re: 2011 WC - Commentary and Reporting
« Reply #74 on: Jan 31st, 2011, 2:52pm » |
Quote Modify
|
on Jan 31st, 2011, 2:24pm, Nombril wrote:| It seems 2*T (for the preliminaries at least) would be long enough to satisfy the cheating concerns, and would lessen the wait for post game interviews. |
|
I guess everyone will have a different estimation of (a) how much a cheater could benefit from listening to live commentary, (b) how likely anyone is to take advantage of that benefit by cheating, (c) how damaging it is to the World Championship and to the Arimaa community if cheating does occur, and (d) how much of a burden it is on the spectators to see a delayed game.
I view the choice between a 2-move delay and a 3*T delay mostly as a discussion of (d), i.e. both options seem to minimize (a), so it is a question of which kind of delay is better for the spectators. It seems, however, that the discussion is unlikely to stay confined these two options. Reducing the delay to 2*T makes it a question of balancing (a) against (d), which will inevitably lead back to disagreement about (b) and (c).
I have a very high estimate for (c), and while I think (b) is relatively low, it would be absurd to think the probability is negligible given that it has already happened at least once. I am comfortable with either a two-move delay or a 3*T delay, but if there is a reduction in either, I will refrain from giving live commentary.
I understand that all of (a), (b), (c), and (d) are very subjective and that reasonable people may differ. Everyone will have a personal level of comfort. Nombril will give live commentary if the delay is 2*T. Chessandgo will give live commentary with no delay. If either of these happens, I will do my best to refrain from any bitterness or rancor. I will not accuse Nombril and chessandgo of aiding and abetting cheaters. I hope, however, that in return no one will accuse me of "burning down the village" if I refrain from giving commentary when the delay is set too short for my comfort.
Quote:| Actually, the wait didn't bother me, there were some benefits. I found it very interesting to hear the last minutes of commentary. It also gives a chance for post game interviews to be more collected, as I also had a chance to quickly review the game. |
|
So to you, (d) was relatively small? The spectator delay is not annoying you more every game as it is Janzert?
|
|
IP Logged |
|
|
|
|
Home
Help
Search
Members
Login
Register
Reply
Notify of replies
Send Topic
Print