My personal take is that the situation described is not a match. But I’m only a lowly club director. But I base it on the fact that the players do not choose who they play and there are multiple players. I would consider the above more like Ladder play than Match play just with multiple games since the players are all playing both white and black each round.
are chosen to play each other multiple times, such as a multi-game team-on-team match.
Due to the bolded Red I was wrong in thinking it wasn’t a Match.
Most, nearly all tournament directors are extremely honest and forthright with a desire to serve chess, USCF, and their local clubs.
The tightening of submission re matches are because of the few scallywag
TDs who for personal reasons chose to abuse the system. Such as
very interesting tournaments where there are more byes than games
played, with multiple players playing each other at least twice, a host of incredible and illogical pairings, and last the TD proclaiming loudly
to all who will listen, their innocence.
Every few years I am “honored” to hear the “persecution” stories of a
few such “maligned” TDs who are “misunderstood”. I tend to be one
who forms the opinion that if it quacks like a duck, looks like a duck,
then it just might be a duck.
Integrity and authenticity in what we do as TDs is a cornerstone. I
understand and appreciate this tightening in regard to tournament
submission. But, I do wish it was not necessary.
Rob Jones
There are valid reasons for having lots of byes in events. (It is the best way to enter a ladder event and have the crosstable resemble the order in which the games were played, I’m told.)
But I do agree that having lots of unplayed games can raise a red flag.
The reasons for match rules have to do with honoring the statistical assumptions behind the rating system (ie, many games against a large pool of opponents) and protecting the rating system from manipulation. (Suppose some near-master played several matches against some 1600 players in order to try to get over 2200, for example.)
Whether short team-on-team matches trigger either of those conditions is not clear, and the Board has been asked to review that portion of the match rules.
Of course there is a difference between, for example, the New Hampshire-Maine matches, in which the rosters are fixed by the captains far in advance, and the Oklahoma-Texas matches, in which the rosters are determined by whoever shows up to the site that day. IMO, the former is a match and the latter isn’t, but I’ll be #### if I can tell from looking at a crosstable which is which. FWIW, I’ve directed both of these.
Alex Relyea
A local “very strong expert” decided to only play house games against much lower rated players (not sure if he would consider a 1600 player or if that was too high) until he went over 2200. I’m not sure how much money he spent on gas getting to all the area tournaments.
Alex Relyea
Several years ago I suggested that players have a ratings cap based on their opponents in that event.
For example, an A player plays nobody higher than 1600 in an event. Should he be allowed to achieve an Expert rating from that event, or should there be a cap for that event of, say, 300 points higher than the highest rated opponent he faced?
Another type of ratings abuse we may need to be more concerned about is stair-stepping. Jones plays a bunch of games against Brown and does well enough to move up 200 points. Then Brown does well enough to move up 200 points. Jones will probably lose less than the number of points he gained because he now has a lower K due to the higher rating. (I’ve seen examples of players who faced another opponent 50 or more times in 12 months, that certainly has the potential of violating the statistical assumption of having a large pool of opponents.)
If he gives chess lessons then the money differential between what masters are paid and what experts are paid may quickly cover the money he spent on gas with any lessons after that bringing in an additional profit.
I think that some consideration should be given to having the trigger be three games vs two simply because playing a game as white and black equalizes the color difference.
Equalization of colors is a red herring. If they play 10 games, 5 as white and 5 as black, the colors are equalized. The more important question is, how many games between the same two players does it take before either the potential for abuse or the potential for violation of the statistical assumptions behind the rating system is high enough to require enforcement of the match limitations?
No matter where the line is drawn, there will always be events that someone thinks belongs on the other side of it, not necessarily for exploitative reasons.
I don’t agree that two games making equal colors is a red herring. We have double Swiss’s, double round robins etc and those aren’t matches.
The thing I actually find strange with the wording from the rules is that a team on team match with just one game seems to require multiple games to be a match. Wouldn’t a team on team with just one game be just as much of a match as a person vs person match with one game?
Two player matches have a lot more potential to be used for ratings manipulation than multi-board team-on-team matches. And one goal of the limitations on matches is to forestall attempts at ratings manipulation. With multi-game matches, the other goal comes into play–limiting impact upon the statistical assumptions of the rating system of multiple games against just one opponent.
With regard to maintaining statistical purity…
I play 95% of my games at the local club, where the pool of players is relatively small (10-15 people at each tournament, sometimes up to 25), and (in a Swiss format) due to rating distribution and expected results, I usually end up playing the same 5-6 people pretty much every round in the 2nd half of a tournament. Whereas a chess club match occurs once a year, and the opponent is not typically one I see at my club.
Which event - weekly play at a small local club or infrequent matches between clubs - can have a greater impact on the validity of the rating system?
From what I’ve seen, having multiple games against just one opponent is more likely to produce ratings anomalies than having games against several opponents. Whether that’s statistically significant is a question for the statisticians.
If two players face each other more than a certain number of times in a certain time span, how about limiting the rating gain/loss for their subsequent encounters?
Especially if the one opponent is a 2200 floored life master!
Here is an interesting one: Would an event run in the same format as the World Cup be a match?
Those are all interesting questions, perhaps you should ask them of the Executive Board, since they set the match rules.
Well, it wouldn’t be a swiss, and there doesn’t seem to be an option for “Knock Out”. I would say that it would depend on how it is submitted. If it is one match at a time, or even one round at a time, then I would think that it would be considered a match, but if it was submitted all together, then no, because some of the players, eg the winner, would be playing many different opponents.
Alex Relyea
I had a management professor who used to say
That’s certainly going to be the case for any set of rules as to what is and what is not considered a match, even if perfectly implemented in the validation programming.
OK, but what is your opinion?