Dealing with expelled players

Running Chess Tournaments
1

Situation:

A player has been detected cheating during the 5th round of a 6 round tournament. The player was undefeated at this point, and on the verge of winning the 5th round game when the infraction was detected. The player is immediately removed from the playing hall and taken out of the tournament.

I’ll tell you what I did (Senior TD) and am interested in hearing if there was a better way to have dealt with the aftermath.

My initial ruling was that the players first 5 games would be unrated and treated as forfeit wins.

Rule 28I (Opponents of Expelled Players) does not apply in this case, since that discusses valid games that were played with a player that should not have been in the section due to their rating. Therefore, I did not use the guidance of 28I2a and give draws to the round 1-4 players who had played the removed player.

The tournament organizer (a NTD) called a friend of his (another NTD) and they pointed out to me rule 28P (Unplayed game) where is says in the 2nd paragraph, “Note that a game in which both sides make moves is always rated, even if a player forfeits on time or for an infraction of the rules; this type of forfeit is never marked with a F or circled.”

Based on that input, I changed my decision to treat the round 5 game as a rated win; however, I kept the first 4 rounds as unrated forfeit wins.

My thought process is that it didn’t make sense to me to give the round 1 opponent a rated win. (The removed player had about an 800 point rating advantage over his round 1 opponent; I did not want to give that opponent a rated win over such a higher rated player that he (probably) didn’t deserve.) I REALLY wanted to treat all 5 games the same, preferably as unrated forfeit wins; however, I didn’t see how 28P gave me any flexibility for the 5th round game.

I can see people saying that 28P should also mean that the rounds 1-4 games should be rated; however, as a previously said, it seemed like the best of bad options was to for me to assign the forfeits for round 1-4.

No matter what I do here, the tiebreaks of the players involved are going to take a beating for playing an opponent with a zero score. Unfortunately, his 3rd and 5th round opponents ended up in a 4 way tie for first, and their tiebreaks doomed them to finishing 3rd and 4th. I’d also be interested if anyone can think of a better way to deal with tiebreaks in a situation like this that doesn’t penalize the players for having played against someone who was tossed out for breaking the rules. (I had to console myself with the philosophy that tiebreaks are already not a perfect solution and this is just another example.) Even when you toss a low score, you still hurt them because they can’t toss another honestly obtained low scoring opponent.

2

I’ll add one more data point:

I’m very confident that the player did not cheat in the first game.
I’m very confident that the player cheated in the third game.
I think it is likely that the player cheated in the fourth game.
I’m on the fence on the second round game.

Does that change people’s minds on what games should be rated or not?

3

For the tiebreak, I don’t have the Rulebook in front of me right now to check if this is covered by a rule, but I’d think the logical solution is to credit all of his opponents as having played an opponent with a score equal to the average of their opponents’ scores. That way they’re still essentially scored based on their opponents’ scores rather than suffering a penalty equivalent to the tiebreak treatment of a full point bye. IIRC, this is similar to how debate tournaments (which are typically Swiss System without ratings, so it would be unfair for full point byes to be treated as zero for tiebreaks) are handled.

4

You don’t tell us exactly how he cheated, which might affect how we would rule. Did he use notes of some kind? Was he receiving advice?

When I read your first post, I assumed that the situation was that he was caught cheating during the 5th round and that there was no real evidence that he had cheated in prior rounds (though I can imagine that in any case where a person is caught clandestinely cheating, there would at least be suspicions of prior cheating, which is why the details of how he cheated could be important).

I agree with you that Rule 28I would not apply, because this rule was intended to deal with validly played games against opponents whom the player should not have been facing, whereas here we are dealing with games that were not validly played due to cheating.

We can assume that if the cheating player loses a game, he would have also lost it if he were not cheating, but the same is not true if he drew or won the game, and therefore a win or draw by the cheating player has no validity, either for rating purposes or for purposes of tournament placement.

If the evidence indicated that the player had not cheated during the first 4 rounds, I would be inclined to count those games as rated wins and to count his 5th game as a rated loss. If it seems unfair to give his opponent a win (which could affect both the opponent’s rating and standing with respect to prizes), I would point out that there are any number of situations unrelated to the players’ playing abilities that can affect the outcome of a rated game (e.g., a player, in a moment of anger, knocks over some pieces and is then forced to recontruct the board position from the scoresheets on his own time, causing him to lose on time).

What makes this situation difficult is that (for reasons you don’t spell out), you have reasons to believe that he also cheated on some, though not all, of his prior games. I’m not sure what, exactly, “very confident” means. My inclination would be to report the 5th round game as a rated loss for the cheating player, and to treat the other 4 games as though he had not shown up for the game. If you’re certain he didn’t cheat in the first round, you could, I suppose, count that as a rated win for him, but you would risk being unfair to his first round opponent if (1) you’re wrong and he really did cheat on that game, or (2) he really didn’t cheat on the second and/or the fourth round game but you didn’t treat those games as rated losses for his opponent(s) as you did for his first round opponent. So I’d simply throw them all out so as to be as fair as possible.

For tie-breaking, I’d follow the procedures stipulated in each tie-breaking scheme with regard to unplayed games.

Bob

5

At the HB Global, a player who was expelled for having used a confederate and computer advice was treated as having been in the wrong section (players had in-section non-played game scores of one-half point higher than their actual result and the actual games were moved to an extra games section).

6

Here is one example of how a player removed for cheating was handled at a big tournament. Look at player #2 in this crosstable. This particular player shows as having a full-point bye in the last round. Player #2 was originally paired against Player X, who was leading this section going into the last round. Player X was determined to be cheating (due, in this case, to an overwhelming preponderance of circumstantial evidence). This was discovered by the TD staff during Player X’s round game.

Player X was thrown out of the tournament, and his name was stricken from the crosstable. His last-round opponent was given a full-point bye, and each of his previous opponents in the tournament was given a bye that was half a point higher than the result of the played game. (I believe Player X was 7.5/8 at the time his cheating was discovered; his seven beaten opponents were given half-point byes, and his one drawn opponent was given a full-point bye.)

As an aside, you can look at player #99 in this crosstable for an example of how a player whose games were invalidated post-tournament was treated.

7

That is the process for Rule 28I, which I did not feel applied.

The cheating in this case was running Pocket Fritz on a Dell PDA instead of E-Notate.

I have no reason to suspect the first game, as I saw the player using a paper scoresheet. (and there was almost a 1,000 point rating difference between the players.) I did not observe the 2nd round game. The 3rd round game was against a significantly higher rated opponent. The 4th round game was against a comparably rated opponent (however, looking at the expelled player’s history I feel his rating is inflated by around 400 points.)

8

Since three of the four players are local, and the fourth is about 2 hours away, I’m willing to front the fourth the travel costs to come back here and play a quad tie-breaker! :smiley:

I’ll be running a K-6 tournament on the 14th, and I’d be more than happy to be TD for that quad, too!

9

What about for tie-break purposes only treating the cheater as if he had withdrawn from the tournament. You treat all unplayed games as 1/2 point byes. Thus, for tie-break purposes the cheater would have a final score of 3 points. This may still put tie-breaks in the dumpster for those players with the cheater as an opponent, but it would not be as bad as a 0 score. Would that have actually made a difference in the tie-breaks?

Larry S. Cohen

10

That’s why I didn’t opt for this procedure. My reasoning was that Rule 28I was intended to deal with the situation where games were validly played against opponents whom the player shouldn’t have been facing, and it is therefore reasonable to assume that the opponents’ performance, had they been facing the correct player, might have been half a point better than they actually did. But when a player cheats, we have no idea how his opponents would have done if he hadn’t cheated. Would he really, e.g., have beaten a higher rated player? So it seemed fairer to just treat the games as though they had never been played.

Bob

11

I don’t have a rulebook with me so perhaps there is a direct text contradicting my instinct. I’m not going to touch the question of how to deal with final standings for prizes without reading up.

For the ratings report, I have no problem with forfeiting the cheating player and giving the opponent a rated win. If the fool wants to cheat to the level of tournament expulsion then I don’t mind them losing every rating point (s)he is entitled to lose in addition to losing an entry fee, losing time, and future sanctions.

12

There is currently no way to give one player a win for ratings purposes and the other player a forfeit loss (eg, non-rated) in that game.

It probably/hopefully/thankfully isn’t a situation that comes up often enough to justify reprogramming to permit that.

13

The expelled played played two of the four people who tied for 1st place. Assuming he lost to both, he could get a maximum of 4 points. (However, he would not have been paired with the one he played in the 5th round.)

Looking at the results, assuming the expelled played had 4 points (there were six rounds) for the tie-breaks of his opponents would not have changed the top dog. Even assuming 5 points for tie-break purposes would not have changed the winner.

There is only ONE scenario that would change the winner: the expelled player loses to his third-round opponent AND still gets a total of 4.5 or 5.0, AND that third-round opponent then beats players with one more point in his 4th and 5th rounds (because he would be in the higher score groups for those rounds). In that scenario only, the expelled player’s third-round opponent would have won on the first tie-break (Mmed).

14

I’m just glad that so far while there have been a few comments about alternatives, no one has said, “What an idiot! Of course you should do …”

It wasn’t a situation I had been in before, and hopefully I’ll never be in the same situation again in the future!!

15

I have a problem with that, because – as I tell my students – rating points are not M&Ms. They’re not a reward you get for winning a game. Rather, they’re a measure of the likelihood of your winning or losing against another rated player. If Player A cheats, you have no basis for knowing whether Player B could have beaten him in a fair game. You can’t say with any certainty that Player B would have won; the data are insufficient. Awarding Player B a rated win has the potential to undermine the accuracy of his rating, reducing its predictive value and potentially affecting pairings in his next event. It’s faulty feedback.

16

@sysadmin Good point on the fine line and rules of submitting an forfeit win.

@Keith, overall I agree with your comment on a statistical and academic sense. If I had students I’d say the same thing.

For practical sanctions in this outlier case, there are few avenues for hitting the cheater where it hurts outside of suspension from the USCF. Despite the academic definitions, ratings matter emotionally to many people. I’ve been trying to find an analogous case from the sports world about how cheating is handled, but unfortunately the theory of vacated wins just doesn’t match up to handling chess ratings.

I hope it’s an outlier. But it’s also a good discussion case for what can be done in the procedural and practical sense.

17

The cheating player gets no credit and no prize, he forfeits his entry fee, and he jeopardizes his freedom to play in future events. These aren’t sanction enough?

18

Agreed. It’s always tempting to want to use the rating system to punish the cheater. But it’s a serious mistake, because it compromises the integrity of the rating system. The cheater will now be underrated, which means that his next several opponents (in his next tournament or two) are the ones being cheated, until his rating climbs back to its proper level.

Besides, don’t cheaters want to be underrated? It’s called sandbagging.

Bill Smythe