Is that the system once known as “Team Sonnenborn-Berger” – the team’s game score times the opposing team’s match score, round by round?
Bill Smythe
Yes.
In that case, I disagree with your earlier sentiment:
Straight game points is not a good tie-break system in team tournaments. If team A defeats team X by only 2.5-1.5, while team B defeats team Y by 4-0, it is likely that team A played against tougher competition than team B. Straight game points punishes team A’s greater achievement.
Bill Smythe
Perhaps. I’m not sure that the gain in “accuracy” is enough to outweigh the extra trouble, though.
34G comes out clearly in favor of the USAT (Barry) tie-break system. 34G1 explicitly says that game points are of “questionable value” as a match-play tie-break. As far as extra trouble goes in using USAT, it involves merely clicking on the “use tie-break” option when doing the prize list (at least in WinTD). Based on 34G3, individual tie-breaks like Modified Median and Solkoff could also be used, but they make the remaining individual games in a match meaningless after the match has been decided.
Perhaps you haven’t read the earlier posts on this thread. The reason I am unhappy with the Barry System tiebreaks is that the SwissSys calculation of them is not reliable. Calculating them by hand is very tedious, especially since you can’t start it until every game is finished. Is the inconvenience to the players worth the marginal gain in “accuracy”? Opinions will differ. Mine is that all tiebreaks are inherently arbitrary, and arguing about which is “better” comes close to the “angels on the head of a pin” dialog.
Note to moderators: This is a worthwhile subject for discussion, but it doesn’t really belong on this thread. Perhaps you could split off a new one.
This tiebreaker discussion now has its own thread.
Is 34G a case of the existing practice establishing the rule or is it really ‘better’ than other possible tiebreaks in team events?
I’d say it was the personal opinion of whoever wrote that section. The trailing “and is preferable” was added in the 4th Edition.
“Existing practice” is a little slippery here. There aren’t that many fixed-roster team tournaments. If you go by number of players/games, the USATE is a lot bigger than all the others combined. If you figure it by tournament, well, that would take some research.
It’s certainly better than game points. I’m not convinced it’s better than Solkoff (computed at the match level). This is particularly true for tie breaks involving the teams with one loss each. I’ve seen too many examples where the Barry tie breaks give what I (and probably most people) would consider unnatural orderings. A good team that loses in a middle round will often be able to put up crushing wins against their next few opponents. 4-0 match result vs 4-2 team is better (in Barry tie breaks) than 3-1 against a 5-1. A 4-0 against a 3-3 is better than 2-2 against 5.5-0.5.
WinTD has an additional tie break which takes a weighted average of the Solkoff and game points, where the weights shift towards Solkoff as the number of rounds increases. This approximates the team “performance” ratings if all teams started with the same rating.
(5+N)(Solkoff/N)+8(Game Points/G)
where N is the number of rounds and G is the number of games per
round.