In practice, this is a bit terse. The Solkoff tie break for individuals has adjustments for unplayed games/matches which aren’t included in this. A particularly serious problem is when a team wins a match by forfeit. Taken literally, this would give you 4.0 x whatever that team scored in the remainder of the tournament even though you never actually played them. (Solkoff gives you zero credit for an unplayed match). WinTD has (pretty much since day 1) computed this by taking points scored in the match x Solkoff-adjusted opponent’s match points. The Solkoff adjustments are to use zero if Team A won by forfeit and 1/2 point match credit (rather than the tournament match credit) for any unplayed match by B (i.e. if B had a full point bye and had a tournament match score of 3.0, that would get adjusted to 2.5 replacing the 1.0 bye with 0.5).
I was wondering whether anyone thought maybe we should try to be a bit more exacting with how these are to be done.
At least the computer will figure it out for you. I remember once doing the team tie-breaks by hand with a legal pad. I had to recalculate due to a team that withdrew after the first round.
Are you asking whether “we” (U.S. Chess) should be more explicit in spelling out the details, especially regarding forfeits, or whether “we” (WinTD) should try to match U.S. Chess policy more precisely? In the latter case, of course, the next question is “what exactly is U.S. Chess policy in this case, especially regarding forfeits?”.
i.e. should the US Chess rules be updated to have consistent treatments across (similar) tie breaks? There is also a similar problem with the description of S-B tie breaks, where the S-B for a Swiss defers to the S-B description for a RR, though that at least does include that unplayed games count as zero for player A.
An S-B score for each opponent has three possible values: 0, x/2, and x, where x is the opponent’s final game point score.
A USAT tiebreak score for each opponent has nine possible values: 0, x/2, x, 3x/2, 2x, 5x/2, 3x, 7x/2, and 4x, where x is the opponent’s final match point score.
Of course, the values of x may need to be adjusted for unplayed games. This is fraught with the possibility of unfairness, but the purpose of tiebreaks is not to be fair, it is to distribute prizes that cannot be divided.
There are five tie break systems have the opponent’s scores as a crucial input. Median, Modified Median and Solkoff all inherit the adjustments in 34E1 (apparently—the position of the paragraph describing the adjustment is a bit odd). OTOH, S-B and the USAT don’t refer back, though S-B does include the adjustment to zero for a game not played.
Sure, but individual S-B and the USAT system share the property that both each opponent’s total performance and the player’s performance against that particular opponent figure into the player’s tiebreak total. That property is not shared by the other systems (Solkoff, Median, Modified Median).