TieBreaks and Checkin

A number of you have asked us why we always run a check-in at all our tournaments. Why don't we pair it the night before, like many of the elementary tournaments do? There are several reasons but I just figured out one of them this past week. Of course the obvious reason is all the first-round forfeits from "no-show" kids. Forfeits are bad because kids pay for 5 games and they should get 5 games, unless we have an odd number then one player has to get a bye. But in addition to this, Ollie always told us to run a check-in because the first round forfeits mess up the tiebreaks. I didn't understand that until this week when, while rating one of the elementary tournaments, the subject of tiebreaks came up.

We all use the same tiebreak scheme, which includes 3 levels, Solkoff, Cumulative score and Opposition cumulative score (SCO). I know I'm not alone in thinking that Solkoff was simply the sum of a player's opponent's scores. Not so. It's the sum of the player's adjusted scores. The adjustment involves converting all the unplayed games to a 0.5 score. So for example, if Kirk (that's me) played Paul, David, Randy, Murlin and Jay in a tournament and won all his games, his score would be 5.0 (fat chance!). But if he was tied with another player who also went 5-0, the tie would be broken by the tiebreaks. Kirk's first-level tiebreak score (Solkoff) would depend on the scores of his opponents. If for example, his opponent's tournament scores were Paul 1, David 3, Randy 4, Murlin 3 and Jay 3, the total of his opponent's scores would be 1+3+4+3+3 = 14. But that's not Kirk's Solkoff total. If Paul's only point was the result of a bye and David had a forfeit win in the first round, then Paul's score for purposes of Solkoff tiebreaks would be adjusted to 0.5 and David's would be adjusted to 2.5, so Kirk's Solkoff tiebreak would be 13, not 14.

The point of all this discussion is that Kirk's tiebreak total was hurt because Paul had a bye and because David's opponent didn't show up. Although this seems unfair (all tiebreaks are unfair), there is some logic as to why they do it that way. The point of Solkoff is to get at a player's strength of opposition. Paul's bye and David's forfeit tell us nothing about their strength as players. The use of the 0.5 point neutralizes the effect of these unplayed games. Moreover, the use of 1.0 point if the adjustment wasn't made, would overstate the strength of these opponent's over-the-board performance in the tournament. But none of this changes the fact that Kirk got the shaft here because of these unplayed games. The unfairness comes in because Paul and David were paired on the basis of their tournament score, not their Solkoff score. So, for example, if I'm in a group of players at the end of round 3 who all have 3 points, and I just happen to be unlucky enough to have been paired against David who got one of his 3 points in a first-round forfeit, I start that round at one-half Solkoff point down from the others in that pairing group who's opponents all had played games. Although one could argue about the benefits or detriments of different tiebreak systems, the real problem is not with Solkoff here, but rather with the unplayed games.

The facts are in, and the jury has voted: first-round forfeits mess up tiebreaks and make an inherently unfair system, even more unfair. Ollie understood all this and implored everyone to run a check-in. I will not go that far. I understand the difficulties and manpower requirements to run a check-in. We all go above and beyond the call of duty to run these tournaments in the first place. My point here is to make sure everyone understands that when you make a decision not to run a check-in, you must accept that your Solkoff tiebreaks will be unfair, and you may have to explain all this to a parent who wants to know why his kid didn't get a trophy.

Hey — just send them to this website.

Kirk Winges
NWSRS Rating Statistician