What are the guidelines for TD’s, as to the required timing for the submission of tournament results (i.e., ratings updates)?
Also, are tournaments always rated in chronological order? My reason for asking is probably obvious, and perhaps already on another thread (although I tried via ‘search’, and couldn’t find it). Summed up by a quick hypothetical scenario -
“Jones” participates in tournament events X, Y, and Z. X begins on Jan. 3 and ends on Jan. 31. Y is an all-day event on Jan. 26. Z is an all-day event on Feb. 2. TD-Y submits results on Jan. 27; TD Z submits results on Feb. 3 (both are obviously pretty conscientious!). TD-X, however, dragged his feet a little - maybe got occupied with Super Bowl partying - and submitted results later… say: Feb. 8th.
Jones is impatiently waiting for his rating update calculations, speculating on which section of a big upcoming regional “Open” he will want to compete in. (He is not much of a football fan.)
So how are the rating adjustments sequenced? By end-date, by submission date, or… some other formula? Does “X” use the latest input ratings, that are based on "Y"s update? And does X ignore Z’s data, knowing that they are from a later event? Is the onus on Z to re-calculate, once X’s belated submissions have been tallied? Do these TD’s have to work together to get this sequence right, or does the USCF provide that oversight somehow?
This seems to happen fairly frequently in my area, and I was just curious as to how it is systematically handled. Thanks in advance for any feedback/ insights.
My understanding is that the tournaments will probably be rated in the order Y, Z, X, but then because X ended before Z, this will be noticed and Z will be rerated.
It avoids the disappointment I experienced when my first Expert rating, that I was eagerly awaiting, came to me on the mailing label (or maybe it was printed on the magazine itself) as … 1995. Two tournaments had clearly been rated out of order, and I knew which ones. I wrote at least two letters to USCF, confirmed that was the case, asked that it be fixed, and the response I got was basically to ignore my observation. Now I can see that rerating at that time would have been effectively impossible, and I couldn’t get a special dispensation.
Yes, ratings are now calculated in order of event end. There’s a more complicated formula when a player plays in more than one section of an event, which happens more often than you might think. If a tournament comes in that ended before one that has already been rated, all tournaments that ended after that one have to be rerated, as far back as the computer can handle (I think that’s some time in 2006, so most tournaments before that deadline should have been rated by now).
As far as not knowing what section you’ll be able to play in, official ratings are published anywhere from three to seven weeks before the tournament starts. For example, if you were looking to play in a tournament that started September 23, you can look and see on your MSA page what your official rating will be for September. These were finalized on 7 August (the first Friday of the month) and will be official all month. TDs can use more recent ratings if they want to, but are strongly discouraged from doing so. If your rating can change in the middle of a tournament, it’s a very bad thing.
TDs are still expected to submit their rating reports within 7 days of when the event ends. Around 93% of events are now being submitted online, over 60% of those are rated within 2 days of when the event ended.
We can rerate all events that were initially rated on or after January 1, 2004. Most of the time we don’t have any changes in events more than 6 months old, but occasionally there are changes from events older than that, usually due to corrections in an event, such as a result or a USCF ID.
As noted, the sort order is:
Section ending date
Section beginning date
12 digit event ID
Section number (as assigned by the submitting TD)
But it can happen. I’m playing in a Monday Night tournament that runs over the course of 4 Mondays. Right in the middle of that tournament I played in a Saturday tournament. So the Saturday tournament will be rated before the Monday night events. Turns out I had a lousy day rating point wise Saturday and lost about 59 points. So this means that the draw I had with a player on the second round of the Monday night events will mean that I will be about 50 points lower rated than 2 or 3 points higher rated when the Monday night event gets rated. (On the other hand it could have gone the other way if I had had a good day.)
Thanks, to all the responders, for the useful feedback. I knew this was probably a pretty common issue.
I am not knowledgable (obviously) in the myriad rules that TDs have to know. I acknowlege that openly. But from a pedestrian mindset, if you will, I think that a changing rating “mid-tournament” can be a tolerable situation… provided that the sequencing rules are well defined, and consistently applied. I didn’t know the mechanism for calculating ratings (I had thought the TDs did some or all of that); it sounds like it gets done by USCF. Thus in my original scenario, TD ‘Z’ does not bear any burden from TD 'X’s tardiness - it is corrected by USCF, in some re-rating process. Thanks for clearing that up.
But now the situation can be much more complicated, as of course there are other players than ‘Jones’ - and each might be in different tournaments, during the same time frame, that don’t overlap (or minimally overlap) with each other. How does USCF track this? If ‘Smith’ travels across country mid-tournament, say for work, and happens upon a USCF event ‘W’ that she (EO here) enters, and TD ‘W’ also beats the laggard ‘X’ with the results entry - does USCF’s algorithm pick this up?
I apologize for my piciune focus here - but I somehow find this quite fascinating, having recently reentered USCF after a 33-yr hiatus. (Instant rating updates, digital-delay clocks… too much for me!)
From what Mike has said before, when tournament A is submitted and rated all tournaments with a later standard rating sequence that were rated ahead of tournament A (and which have players from tournament A) are flagged to be rerated. Then those tournaments can cause a further cascade. If the 2005 US Open or 2005 Supernationals or 2005 World Open end up getting rerated (possibly triggered by an error in a March 2005 tournament that a PA scholastic player went to before going to those other three tournaments) then the cascading effect might end up causing the majority of the tournaments for the last four years to get rerated.
I can’t remember if every player in a rerated tournament triggers the cascade, or only the players that had a difference in the pre/post re-rating results.
One thing that helps minimize the cascading is on-line tournament submission. With that I’d expect that few tournaments cause more than a month of cascading.
I played around with trying to flag players/events that needed to be rerated, but it turned out to be easier (and just about as fast) to rerate the entire month, and then subsequent months, of course. Of course, if there are no updates in an event when I rerate it, I can skip that part and move on to the next event.
Two years ago I played in two events the same day, both started and ended that day. The first event was rated in about seven days. The second event was rated that evening. After the re-rating process the two events actually show up in my history in the correct order. Somehow the event that was submitted later has a lower tournament ID number. I assume the ID is what drives the rating order of events that end on the same day and this is how it was ultimately rated first.
Actually, this is my point. Everyone’s unofficial rating can change during a tournament; this is a good example. This is why TDs should use official published ratings. Should the TD pair you differently now that your rating has gone down 59 points?
Event IDs are based on the ending date of the event and a sequential number (and are checked for uniqueness).
The sequential number rolls over periodically, but most of the time if two events with identical starting and ending dates are uploaded the one that gets uploaded first will get the lower event ID and thus be rated before the other one.
I assume the sequential number is a fixed number of digits (say, 4) and that rollover occurs when the number reaches all 9’s (say, 9999), at which time it rolls over to all 0’s (say, 0000).
So, to determine which sequential number was assigned first, it would be logical to simply look at the difference between the two. If the two numbers differ by less than 5000 (say, 3639 and 4201), assume the smaller number came first. But if the two numbers differ by more than 5000 (say, 9781 and 0346), assume the larger number came first.
It’s actually a 3 digit sequential number, the 12th digit is used for other purposes.
I don’t see much point in trying to order the events more precisely than we currently do.
How often does someone play in two INDEPENDENT events that have the same beginning and ending dates? And if the event is part of a group of events by the same organizer and most likely at the same site (such as morning and afternoon quads at a chess camp), the submitting TD has full control over which events get rated first via the section number, assuming they’re submitted as one multi-section event.
