| From | Message |
samogon79
4/10/2006
17:03:45
[ report this post ] |
 Subject: Team games!
Message:
Here is how team games are currently being setup:
Team captain will pick a pair, an opponent, from a different Team based on their “Non Team Rating”, and also some of the captains look at opponents “Team Rating”. This is all good, but there is no secret that lots of team players are manipulating their “Non Team Rating” in order to get a weaker opponent for a team game, and get more points for a team.
There is also another problem: If your team rating is higher than your “Non Team Rating” you can easily get disqualified from GK for manipulating you ratings.
Big number of players, value their team games, more than regular ones, and while playing a regular, not team game, they would try new combinations, openings, etc. At the end, their team rating would be higher that a personal one. These kinds of players are in a big risk of getting disqualified from GK, due to rating manipulation.
I want to hear your opinions on Team games, team ratings, and what would be the best/right way to set up team games. Any other inputs would be also appreciated.
Thanks in advance,
Dima (samogon79)
|
eqj2
4/10/2006
17:36:39
[ report this post ] | Ratings
Message:
When a team captain makes a challenge which rating is the captain to use?
|
spymaster
4/10/2006
18:26:03
[ report this post ] | ratings
Message:
eqj2
I use team ratings .... but team ratings do not show next to the players names ... so .... :-(
")
|
samogon79
4/11/2006
10:16:44
[ report this post ] |
Message:
Spymaster, yes, You are right! There is no Team rating next to player’s name.
There comes a question: what is the purpose of a Team rating, if it is not being used for team games?
|
nima_tal
4/11/2006
11:12:59
[ report this post ] | .
Message:
It has been pointed out several times by several people (Some, team captains) that indiviual players team games rating is a meaningless number that has no use what so ever. Personally I never even look at it.
|
samogon79
4/11/2006
15:31:06
[ report this post ] |
Message:
Some of the players from our team got kicked out of GameKnot, for having Team rating Higher that personal, and the response from GK was, that this is a rating manipulation issue.
The question is; how you determine rating manipulations? Just by comparing ratings? In this case, what is the purpose of a team rating, if it is not currently used by GK system to help Team Captains to set up a Team game?
|
eqj2
4/11/2006
19:51:33
[ report this post ] | Wow
Message:
I have ran into a player that has a normal rating of 1700+ and his team rating of 1324 thats pretty scary.
|
dharmapunk
4/12/2006
03:07:59
[ report this post ] | When setting up team games
Message:
I find it's best to look at their current rating as well as the rating graph and their highest rating in their past 50 games. This gives a pretty good idea on someones average playing strength.
A while back I timed out in a lot of team games and regular games dropping both ratings (not on purpose). Since I don't play nearly as many team games it takes much longer for my team rating to get back to where it should be (even more so if I lose a game).
I don't think I've ever matched up a team game based on both players team rating.
|
cash1981
4/12/2006
05:03:57
[ report this post ] | Team rating is BS
Message:
Really team rating is BS. What really matters is the % win and % loss in team or regular games. Some players have so insane difference between these values, that it is so obviously cheat. Of course you concentrate more on team games, but be reasonable. Take a look at my page for instance. The values are just a few per cent different, and I play a LOT of team games and normal games. I don't want to call out names, but people who does this know who I am talking about. I am in favour of gameknot closing these accounts.
|
chuckventimiglia
4/12/2006
06:21:10
[ report this post ] | I believe that everyone's.....
Message:
team rating is far below their normal rating.
I do not understand how team rating is calculated
or of what use it is.
When I had a team here I would look at a
person's Win% in all games and their Win%
in team games. They both should be close.
In some cases like mine my team win % is a
lot less than my overall win %. No red flags!!
As mentioned above when a person's team win%
far exceeds their overall win% that would send up a
red flag for me when I was getting matches.
A person that has an overall win% of say
50% but their team win% is say 70%
for some reason that would look and be very odd.
Especially after a person is established here with
many games played.
I am no longer a team captain so I am no longer
concerned about it.
Chuck
|
fedor
4/12/2006
11:51:34
[ report this post ] |
Message:
Lets use team rating to combine pairs in team games and problem of rating manipulation dissolves!
I was disqualified from team for "rating manipulation". Because my team rating was 1750 and main rating 1650. But I never manipulate of rating! I just play team games for win and other games for fun. And I like risk (but not in team matches).
Is it not my right? After disqualification I finished many games and my rating now 1600!
If captains will make challenges by team rating - rating manipulation will impossible.
|
bob2
4/12/2006
12:32:53
[ report this post ] | Fedor
Message:
Playing team games to win and other games for fun is rating manipulation.
Whether or not it is your right to do it is another matter!!
|
mattdw
4/12/2006
12:49:08
[ report this post ] | Bob
Message:
It may result in a higher team rating playing the way fedor suggests, but does that mean if we join a team we are no longer allowed to play our non-team games in a less competitive manner? I would say non-'competitive' games are an opportunity to try out new openings, styles or to play that risky sacrifice. In a 'competitive' game we may be more inclined to go for the win first, knowing that we have ample opportunity at other times to play games where learning is the prime goal.
|
chuckventimiglia
4/12/2006
13:00:24
[ report this post ] | Whenever a person has....
Message:
a very wide spread of Win% in favor of team
games there will always be hesitance to pair
that individual in team games.
A person should play to win in any game he/she
plays.
|
fedor
4/12/2006
13:34:59
[ report this post ] | chuckventimiglia
Message:
> A person should play to win in any game he/she plays.
Say me pls - when you play with your children (for example) - you play to win? I think no... And such examples enough (trying new opening etc)
Why my games for fun is problem? Because my team rating will higher that common rating? And so? What is a problem? Rating as is do not give me any advantage. Problem - when I in team game will play with player that have TEAM rating smaller. I propose play team games based on team rating.
And nothing manipulation's problem.
PS. If player's purpose is rating manipulation - he did it clever and he'll never disqualified. Because he think about how to do it.
|
ccmcacollister
4/12/2006
15:28:55
[ report this post ] | It seems simple
Message:
That team games Should be paired by team ratings. Of course GK has never been one to tell us What To Do with its features. But it seems kinda obvious this must be what it is there for, imo. :) [and apparently just hasnt really caught on, from what Ive seen/read about it.] I've just always thought this to be the case, or as Everyone seems to say, what is it even there For?
|
johnrowell
4/13/2006
01:00:51
[ report this post ] | It's not manipulation ..
Message:
... to play team games to win and other games for fun. Rating manipulation is a *conscious* effort at driving down your ELO to be matched up against weaker players in your team games.
|
odor
4/13/2006
01:17:31
[ report this post ] | Unrated games
Message:
If you're not playing to win, then play unrated games...
|
fedor
4/13/2006
01:39:47
[ report this post ] | to win or/and to fun
Message:
IMHO question is not in win/fun.
Question is in disqualification of fair players.
I never play "not to win".
But two ways to play to win: "play to win carefully" and "play to win with fun and risk".
When I play my games - is a my risk. When I play team games - is a risk for all my team.
|
mattdw
4/13/2006
02:01:10
[ report this post ] | odor
Message:
I disagree. Feedback is a very important factor in the learning process, a changing rating over time is feedback upon our progress - we therefore have a more tangible guide as to whether we are making improvements or not. It would be much more difficult to judge our current ability on a handful of recent games.
|
ccmcacollister
4/13/2006
03:55:47
[ report this post ] | fedor ...
Message:
Great point you make
|
ionadowman
4/14/2006
03:31:41
[ report this post ] | Check out the GK ratings thread...
Message:
...I've made some remarks earlier today that are relevant to this discussion. One of the reasons why one's overall ELO is often quite a bit higher than one's team ELO is that the latter measures a subset of the former: one's team games only, as against all one's games. Now, I joined a team having played on GK for about 6 months, and my then GK ELO was roughly 1800. My team rating began in the same way as the GK rating: a provisional of 1200 - a 600-point difference. Since I play team games almost exclusively, the present difference of 300 between my ELOs is likely to endure, bearing in mind results are measured under GK and Team ELO lists respectively. As for which ratings team captains ought to use in order to make sensible pairings, it seems simple enough to me: use the players' team ELOs. Apart from anything else, manipulating ThaT rating will hurt the team, so is probably less likely to occur... It seems that a markedly higher team ELO than GK ELO, however, is an alert that something's amiss... For this and probably other reasons it is likely to be the more reliable measure of players' relative strengths. The actual numbers in any rating list are quite meaningless, you know. Of more importance is how they stand in relation to one another.
Cheers,
Ion
|
mattdw
4/14/2006
06:34:37
[ report this post ] | A thought..
Message:
I've not looked too much into the specific mathematics of the elo system, but I have a thought as to why there may be rating 'deflation'. I think for a start the true rating distributions of the two groups (team & non-team players) will probably be different. It is likely that there are a lot more casual players who are not a member of a team which would suggest the true mean rating for this group will be closer to that of the provisional rating they are given than that of the team players.
I had a look through Idonadowmans past teams games and found an example of how this could cause problems in practice, take a look at the pre & post team rating for Avalon, over his two games against Ionadowman (who still had his provisional rating of 1200 at this time) he lost both games - which is not statistically that unlikely for two players of roughly the same rating, losing him 157 team rating points, a lot more than he would have lost in a normal game. The other problem is that it seems that there are a lot more non-team games played here overall which would indicated that the majority of the corresponding ratings would have stabilised, where as many team players ratings will be much more erratic due to the lower number of games played - thus further inflating (that should probably say 'deflating' ;) ) the problem.
This suggests to me that the provisional rating may be a partial cause for the deflation and that a better solution would be to give everybody a provional rating equal to that of their non-team game rating. I haven't really thought about this too much in mathematical terms so I may be talking rubbish, I'll have a think about it later. But for now I must be off, too much time on the forums at the moment when I could be doing more work! I wont be back for a few weeks, see you all in a while! :)
Matt
|
cash1981
4/14/2006
07:45:16
[ report this post ] | RE: fedor
Message:
Hello Fedor. I am sorry to say this but It looks really fishy when looking at your stats. 95% win is way to much with 24 games played. Also your x team has also very doubtful stats. I am not saying they are cheating or pointing names, but I just point to the stats and let them speak for itself.
|
leo_london
4/14/2006
08:51:04
[ report this post ] | cash...
Message:
Nothing suspicious there at all if you take the trouble to look at the actual match-ups in his team games. His team captain seems to find him him some fairly easy match-ups, thats not against the rules...in his other games he has been playing opponents with much higher ratings, unless you find that suspicious ?
|
se_santi
4/14/2006
12:24:39
[ report this post ] | I'm
Message:
already getting tired of the discussion about team games, and non-team games.
I play my games the way i see it fit. It's me who pays the account, not any other. Nobody is going to tell me how much time i should dedicate to this game or the other.
There are times when i blitz thru the games and make a lot of mistakes, other when i spend a lot of times looking at the board, thinking positionally and/or tactically according to my mood.
Sometimes i blitz thru a lot of non-team games only to stop in a couple of team games and spend a lot of times analyzing the position, right after having taken less than 10 secs. to move in non-team games. Or viceversa. I even do that with games within very same tournaments (both MT and the big bi-monthly one)
When i am not in the mood or not playing well, i have resigned games within the 10 or 15 moves just because i went down a piece. After all, if the position was lost, why drag the game on? (and you know by your level at any time when you can "row" the game up or not). One day later i started seeing the board again and being down a rook didn't made me quit.
Lately i have been playing kinda slow. Maybe tomorrow i'll blitz thru every some games again.
Also if i want to try something new on a rated non-team or non-tourney game or whatever. If today i want to open games with 1. a3 and tomorrow play Ruy Lopez, it's MY choice. Next big thing, some will be saying "hey, he opened with 1. c4, while we all know that statistically it has less percentage of wins for white, while on me he opened with 1. e4. He's definitely manipulation the situation in favor of the other player".
I will continue to play the way i play and i suggest others doing the same. Rating manipulation is losing games on purpose. It is NOT
1- dedicating this time to this game and then dedicating less to the other game.
2- testing new combinations on some games
3- resigning early when you see you're below in material / position and you can't just outsmart the opponent, wether at move 10 or move 50.
For instance, someone in this thread, bob2, i think, said that it is rating manipulation, right? Well, to put an example, bob2 is 2000+, and i have only twice gone up 1500... And sometimes i see games between players 2000+ or so that they agree to a draw in what i consider, an early position (having some rooks and pieces on the board). Do i accuse them of cheating? Well, no. To me that position would be far from a draw, but then it gets to my head that maybe their understanding of the game and combinations would head that position to a dead position, even with the best effort to win on both parts, because i have had those position in my playing strength level).
Good luck!
|
ionadowman
4/14/2006
16:04:59
[ report this post ] | Avalon...
Message:
...timed out after 4 or 5 moves in his games against me. A gift 10-20 ELO points (and, more pertinently, a boost of several placings on the rating list) that I didn't deserve...
Time-outs of this nature might have a slightly 'inflationary' effect on the list. It certainly had on my rating!
***
Fedor's 'case' seems to suggest that even a higher team ELO than GK ELO (about 150 difference) is (as I thought) not at all a reliable guide to a player's character.
***
I can appreciate se_santi's irritation, but I think the main question in this thread was which rating list a team captain ought to use as a basis for pairing his members with opposing teams' players. For this purpose, my view is that the player's team rating ought to be preferred.
***
I note some people are voting to put this into the teams forum. If this vote goes though, does it transfer?
Cheers,
Ion
|
ionadowman
4/15/2006
14:41:53
[ report this post ] | Ah...
Message:
I perceive my question has been answered...
:-)
Ion
|
taborov
4/16/2006
15:51:36
[ report this post ] | Fairplay in teamgames
Message:
A chessplayer always plays to win the game !
Trying things out like new openings is fooling yourself.
Each player has his own caracteristics, and he always plays the same line.
Losing regular games in a few moves is not normal ! and playing like a monkey in regular games is also not normal !
When you play your games fair you will learn something from it !
Chess is an honest game.
Greetings from Taborov.
|
ionadowman
4/17/2006
04:26:42
[ report this post ] | I can understand...
Message:
...people trying out the unfamiliar in circumstances that don't impact on others, team mates, say. One trying to develop tactical imagination might decide to play risky openings, quite unlike one's usual 'middle of the road', positional lines. Departing from one's chess comfort zone in this way will no doubt be attended by a higher ratio of losses to wins (given no change in the standard of opposition).
***
However, it seems to me that such an exercise has no value unless one plays the games to a finish. Resigning the game in a slightly inferior position at move 15 isn't going to tell you much - and will defeat the purpose of the exercise in the first place. Further, testing oneself needs to occur within the context of a whole game.
***
So while I can appreciate that maybe one's team games might legitimately be more successful than other games on GK, I don't buy the idea of not taking non-team games seriously. To play to win, to give of one's best as *taborov* suggests, is just ordinary courtesy to one's opponent. To do otherwise is just rude. This is not at all inconsistent with trying out new things.
Cheers,
Ion
|
|
Post a reply to this message:
|
