Found this interesting item in the Tournamant Organizing forum’s thread on Tiebreakers. Didn’t recall seeing this in the rule book and went back to be sure. Thought it might be better to pose this question in the TD forum.
Background: I’ve had a few unfortunate tie-break results at some scholastic events which kids, parents and/or coaches just couldn’t understand. At one event in particular (040124239) there were 4-way and 3-way ties after 5 rounds. Cumulative, Solkoff, then Kashdan were used for that one and posted as such during the tournament. After trying out many possible tie-break combinations following the event, I had to agree with the rule book statement, “there is no perfect tie-break system; each has it’s faults”. However, if I choose one parents and coaches don’t agree with, they become quite generous about pointing out my faults too.
At scholastic events since then, I’ve used the USCF recommended order of Modified Median, Solkoff, Cumulative then Cumulative of Opposition. If anyone has a problem with this, they’re told that it’s the USCF recommended order of tie-break systems in the rule book, and any concerns or other recommendations should be submitted to the USCF. Yes, it’s a dodge. But, some of the parents can be a bit scary and I want to get out of the tournament hall with all of my fur intact.
If indeed the Modified Median isn’t recommended for less than 6 rounds, is this stated in any USCF official document that could be quickly whipped out to use as a shield when parents and coaches become riled at trophy time?
Mike Swatek
ps: Still in search of a perfect tie-break for scholastics, and hoping to take some heat off the USCF, I’m planning to try blitz playoffs for trophy tie-breaks at scholastic events in February, March and April. Should I be planning for more fur replacement?
I prefer to use common game as the first tiebreak. This usually doesn’t work if there are more than two players in the tied group, but I find that parents seldom disagree with me when I tell them that if Jimmy beat Joey head-to-head then Jimmy gets the (better) trophy.
Since in many scholastic events I consider the ratings to be imprecise, I prefer to avoid any tiebreak that depends upon the rating of the individuals in any form. (Cumulative does, because it rewards winning early and that tends to favor the higher rated player.)
Perhaps it’s just because I’ve been working on the ratings programming for the last several days, but I’ve been wondering if a tiebreak could be devised that assumes the players are all unrated and calculates a steady-state adjusted performance rating for them, something like the following:
Step 1. Assume all players are rated 1000. Calculate performance rating using the traditional win+400 loss-400 draw+0 approximation.
Step 2. Recompute the performance ratings using the output from step 1.
Step 3. Recompute the performance rating using the output from step 2.
Repeat until the ratings no longer change after an iteration.
I don’t know if this would converge in every case. It also might tend to favor players with unplayed games.
I’ve searched a bit in the rule book, and I must admit I don’t see this mentioned either. It is, however, a direct quote from what I’ve read somewhere in either previous rule book editions, or in one of the TD Corners of the Rating Supplements. I will have to do some more hunting and get back to you.
The idea is that since the exteremes are discarded, this tiebreak is somewhat questionable for 5 round tournaments, as then only 3 of the 5 games would be used to break the tie. It would be an even less reliable break in 4 round tournaments.
My tiebreaks (for tournament of less than 6 rounds, AND in scholastics) are: Solkoff, Cumulative, Opp Cumulative, Median. Note that in scholastic tournaments, the TD often makes transpositions in the score groups in order to avoid pairing family members, players from the same club, etc., and therefore, the cumulative can sometimes be devalued.
In adult Opens (which are usually 4 rounds in my events), I use Cumulative, followed by Solkoff.
One of the reasons directors do not have scholastic events, is the problems with tie-breaking. Number of directors with the problems of a tie, would just give half the prize money (if a tie with two players) in a case of a tie. The problem with a scholastic event, there is no good way to break a trophy in half.
The only way to be fair for the scholastic students (and parents and coaches), would be having back up trophies ready in a case of a tie. It would be the best way to be fair, for the all parties … except the director.
Still not perfect: You open yourself up to the excuse that “my kid isn’t very good at blitz”. Many, if not most scholastic players are atrocious at blitz chess, and a player who’s even a little to moderately better at blitz may still get beat by an unintentional illegal move (which, btw, occurs most frequently in the regular time controls of a lower grade scholastic tournament).
As imperfect as it is (btw: there ain’t no such thing as perfect), the systems described in the rulebook are far better than blitz playoffs, IMO.
Probably the best that you can do is to decide upon which system sounds fairest for your events (and all events are not the same, btw), and be consistent with it. Prepare a “Parent’s Guide To Scholastic Chess” detailing your ideas (and ideals!!), and be prepared to spend time with newcomers, both parents and players. Hopefully, they will see that there is no perfect system (though I doubt it, at least at first), and will enjoy your events to come. A willingness to explain in understanding tones, although very time consuming during a tournament, can go along way.
