Fellow TD’s,
I’ve submitted an event where the 1st round is G/60 (dual rated), the 2nd round G/75, 3rd round G/90, 4th round G/105. It just occurred to me that when it comes time to select a G/ in SwissSys (for both section setup and rating report), I’m at a loss. With the G/60 round being dual rated, I can’t just submit all games as G/75, 90, or 105. As it stands now, I’m thinking of changing round 1 to G/61, but that’s like running away from the mojo. I want to run through it.
A true conundrum. Perhaps your solution is best. Prevention is the best medicine, next time don’t make the first round G/60.
The Frequently Asked Questions for the USCF TD/Affiliate Support Area
says "If there are multiple schedules for a round, or if not all rounds use the same time control, in general the slowest time control used for any game in any round determines whether the event can be regular rated only, dual-rated, or quick rated only.
Two important principles to remember are that games played at a primary time control faster than 30 minutes can never be regular-rated and games played at a time control that is slower than Game/60 or not a sudden death primary time control can never be quick-rated. "
I hope this helps
Cheers,
Terry
Terry has cited the correct source. This event would be regular-ratable only, because the slowest time control in effect was slower than Game/60.
The thing to avoid doing is having time controls faster than Game/30 and time controls slower than Game/60 in the same event. Such an event would have to be split up for ratings purposes.
BTW, your subject heading is a bit misleading. The fact that you used Swissys is irrelevant. 
Thanks Terry and Mike,
Mike, sorry about the misleading title. I was thinking that I’d have to configure some source code do-hickey in SwissSys to achieve the dual rating for the G/60 round, but since the slowest time control for the event is G/105, I’ll just put G/105 in the section setup and rating report. I’ll make a note to announce this before round 1, so as not to sustain a sprained neck during the ensuing melee.
Al,
I generally don’t input the time control into SwisSys. The print is too small for anyone to notice it anyway. Putting up signs with the round times and applicable time controls might be more effective. It is also easier than remembering to change the time control in SwisSys each round.
The time control in SwisSys doesn’t flow to the rating report anyway does it?
GrantPerks,
I thought it did, but I may be mistaken however, in the process of uploading the rating report files to the USCF, a time control is necessary to successfully complete this process, so whether it’s given by SwissSys or manually entered, I can’t use GAME/60 or the event will be dual rated.
Two points:
a) I believe the intended bias is towards the Regular rating system. The most important principle is that games faster than G/30 should not be rated as “Regular”. If some of the games in your event are faster than G/30, then it should not be “dual rated”.
b) currently, there appears to be a similar breakpoint at G/60 (“no games slower than G/60 should be Quick rated”). So, events which have some games faster than G/30 and some slower than G/60 cause problems. BUT…BUT…we just changed the rules! For complicated reasons (mostly, the Quick system is broken) we will soon be Quick rating ALL games. Repeat…all games. This probably makes the name “Quick” obsolete. To be rated as Regular, all games should be G/30 or slower.
Hmm. An interesting, perhaps radical, idea with a lot of good points.
Will anything be done, on a one-time basis, to align the quick ratings with the regular? Presently, almost everybody’s quick rating is about 100 points lower than their regular.
Bill Smythe
While there was some discussion about making the change to the ‘rating system formerly known as quick’ retroactive to 1/1/2004, I don’t know if there was any discussion of a one-time realignment. (However, I missed part of the ratings workshop due to another committment.)
Ken Sloan and I talked about a few possible new names for this rating system.
Perhaps it should be called the "Universally Significant Complete Formula’ rating system, though the abbreviation of that might be easier to say.
Hey guys,
What do you think about naming the new rating the “constant” rating?
Just an idea.
Re-rating solves everything.
We can probably just mark a fairly ancient block of events as “dirty” and let the re-rating system clean up.
Our historical crosstable data is IMHO not clean enough to rerate events that were initially rated prior to 1/1/2004. (That’s why we only went back that far in the first place.)
It would probably take a lot of work to clean things up enough to go back to rerate 2003 or earlier events, and we could only go back to late 1991 anyway. We’d probably have to rerate both types to get them in sync, and that would likely cause a lot of changes in people’s current regular rating.
I don’t know if rerating all the ‘formerly quick’ ratings from 1/1/2004 on would improve things, or just make things worse.
I may try that on the backup server this weekend.
FWIW, here’s a table showing how many regular and quick rated records we have:
[code]year reg quick
1991 30467 191
1992 160241 8392
1993 166351 12163
1994 183249 16835
1995 199402 19923
1996 205898 20792
1997 206858 20381
1998 204901 17207
1999 198779 17381
2000 196804 15250
2001 205591 25694
2002 225921 53418
2003 227874 71347
2004 221342 101385
2005 216371 141415
2006 142765 103370
[/code]
Even if re-rating is done all the way back to 1991, and if quick ratings were already deflated 100 points at that time, I don’t see how such re-rating would “solve everything”. It might, however, reduce the 100-point discrepancy quite a bit, since now a bunch of regular-rated games would be included too.
For a new name, how about “quick-inclusive”?
Bill Smythe
I know a lot of people whose correspondence ratings are vastly different from their regular ratings. Should we rate games slower than G/2 under the correspondence system? Of course not. They are different games.
So to is quick chess different from slow chess. The difference isn’t as great, of course, but it’s still there. The reason quick chess ratings lag behind regular ratings is because most people haven’t played enough quick rated games. In Atlanta we have weekly G/10 tournaments. The regulars in those events tend to have quick rating similar to their regular ratings. Players whose only quick rated events are the dual rated games are more likely to have a gap.
Including games less than G/1 in the quick rating pool was an attempt to get more games into the mix. This was logical, but of course the line drawn was arbitrary. I personally think G/30 was a better cutoff. But I can’t come up with a rational and validatible criteria for that value judgment. Perhaps the best shot would be the probability that a game will reach the point where neither player has to keep score (i.e. one player has < 5 minutes left). Most of my G/45s don’t reach that point, but about half of my G/30s do.
To me, including regular rated games in the quick pool is nuts. We’ve already gone further than I would down that slippery slope. “Quick-inclusive” is a good description of this, but it also sounds rather pointless. So what if there is a rating gap, as long as the standard in major quick events still use regular rating for pairings and prizes so people gain no serious advantage by that gap. I remember how thrilled one of our local high school players was when he broke 2200 in quick, though he is barely an expert in regular chess. He’s that good at G/10. If quick ratings serve any purpose at all, and I believe they do, let them continue to serve that purpose. They are supposed to reflect your performance in that chess discipline, just as correspondence ratings do.
Agree 100% with Mulfish. We should just recognize quick chess is just a differnt “game” than chess.
Just for the fun of it I looked at the relative ratings for myself and two other players who regularly play in both types of events. We’re all around 50 and all long term players. They are experts, former masters while I’m life A player who dipped into B a few years ago. Three data points are hardly sufficient for proof, and the patterns for rapidly improving players could be different, but it’s still interesting.
My pattern:
When life was “normal”, my ratings fluctuated, sometimes in opposite directions. In Dec 03 my regular rating was 113 points higher. Two supplements later my quick was 87 points higher. Then they ran fairly close for a while. Later, when personal issues affected my play, my quick rating plummeted to its floor and my regular rating dropped, but less. I believe that’s because the slower time controls gave me the chance to compensate for my mental malfunctions. But another player might have had the reverse situation, because their reflex play at G/10 might have been less impacted than their slow play. When my mind recuperated, both ratings went up a bit and the spread has remaines small.
Expert A
He’s had a spread as big as +261, narrowed to 21, back up to 150, down to 6, back up, then to -6, back up, etc. His quick rating has never been much higher, but the spread fluctuates wildly. His last 4 supplements its been -6, +103, +63, +6
Expert B
Four years ago his regular was 115 points higher, but that gap dropped steadily and 2 years ago his quick rating took the lead. That spread has stayed mostly in the 50-90 point range, though it’s dropped to the teens. Last supplement regular surged back on top by 13 points.
Draw your own conclusions. Mine as that for an established player, the rating spread is unpredictable and will fluctuate because at times a player is better able to play on instincts and at other times they need the thinking time to be successful. YMMV.
Your first sentence contradicts the second.
Mulfish approves of using Regular ratings to control Quick events. How does this “recognize quick chess is just a different ‘game’ than chess”?
If there is a contradiction or confusion, it is in my explanation of my position. I oppose having regular rated games factored in to Quick ratings, and I believe G/30 should be the cutoff. For major quick rated events, however, I believe we should use regular ratings for prizes and pairings because there is a high probability that a number of entrants will have unreliable quick ratings because they haven’t played in many quick rated events.
I don’t see that as a contradiction. If quick ratings should exist at all, and I believe they should, they should be kept pure. Their reason for existence, however, is really a bone to people who like quick chess and want to have a rating recognizing that brand of chess. For the local G/10 events we have, it’s great to use quick ratings. But for an event with serious prizes, I don’t want to have some A player whose quick rating is 1200 based on 5 games 10 years ago come in and steal the prize. Using regular ratings is more appropriate. It’s analagous to me going up to play in the Canadian Open. They should, and would, use my current USCF rating rather than my provisional rating from the 1977 Keres Memorial. That doesn’t mean they should factor my USCF rated games into their rating system.