We submitted the USAT North last night, but it is not showing up as rated, or even in the queue. The office is closed today. Anyone else having issues?
Glenn
Records show that the event was submitted for rating at 11:48 last night, which is just after the final ratings run of the day, so it was not rated until the first run on Monday, just after 6 AM. It was posted to the website around 9:45 AM this morning, which is the first daily update of MSA crosstables.
Isn’t it amazing how expectations have changed?
I remember once getting an event crosstable back from US Chess “only” three weeks after the event. No one could believe how quickly it got rated. And it wasn’t really all that many years ago.
Perhaps after the move to newer/faster equipment is complete it will be possible to revise the schedule so that events are rated between midnight and 6AM. Right now that’s when we run daily reports as well as do backups and other maintenance tasks. More frequent updates of tournaments on MSA (currently more or less hourly between about 9:45 AM and 12:45 AM) might also be possible.
At the 2004 Delegates Meeting, where the online submission process was previewed, I made the prediction that it wouldn’t take too long before a 1-2 day delay in rating and posting an event would be considered slow. It took maybe 2 years before people started complaining.
Messrs. Priest and Parker have a good point, but please remember that probably close to 90% of the membership can’t remember a time before events were submitted online, and likely a majority wasn’t even born!
As an aside, a player at the last tournament I did requested a print out of the final wallchart. I told him that I had already packed up my printer, but I’d be happy to email him one. I then mentioned that I had already submitted it for rating and that it should be rated by the time he got home. That was good enough for him!
Alex Relyea
Eventually it will reach the point where people want to see their updated rating before the tournament has finished. 
Oh wait, that already happens. :mrgreen:
I’ve seen events posted to the web within an hour of when they ended, those TDs must do that before they pack up to go home. I never had that much energy left after running an event.
It would not surprise me if at some point there was a demand to go to weekly official ratings lists rather than monthly ones. I’m glad I’m retired!
If I have a wi-fi connection I’d rather do it while I still have everything at my fingertips.
Alex already said that he does it that way too.
PS Sometimes we can even submit a scholastic national for rating before the awards ceremony finishes.
We’ve had national scholastics that were posted on MSA before the awards ceremony ended!
It is my personal opinion that we’ve already gone too far in that direction. For players unlike me, it can be very difficult to plan far out now since they don’t know what section(s) they’ll be allowed to play in.
Alex Relyea
Organizers would have control over that by specifying what list they will use. We already see events in which the most recent list isn’t the official list for an event, that happens in some of the spring nationals and in the USATE. Organizers would have to decide based upon their understanding of their potential audience how much lead time to give players when they set a ratings cutoff date.
The problem I have always had with using unofficial ratings (eg, MSA) is that they can and do change, sometimes several times in a single day. So if an organizer uses the ratings he sees at 6PM on Friday by 9AM Saturday those ratings might be nowhere to be found on MSA. An official list would need to remain available. (We have official lists posted back many years. We track what’s been downloaded, and lists from two or three years ago do get downloaded from time to time.)
But Mike when US Chess does the actual ratings it has no problem using these “unofficial ratings”.
I don’t know of any events organized by US Chess that use unofficial US Chess ratings. The national scholastics and the US Open all specify what official ratings list will be used. (That’s a separate kettle of fish from using some other ratings entity’s ratings.)
Wayne is saying that the “unofficial ratings” are used to calculate the new rating. That kind of overlooks that official ratings are “only” official for sectioning, pairing, and awarding prizes. Those are very different from ratings calculations.
What’s the alternative to using someone’s post-event rating from their most recent event when computing a rating in their next event? Using the same rating for weeks on end to compute someone’s new rating, like FIDE does?
Thanks, Mike. I was the one who submitted this. Having never submitted an event for rating that late at night, I didn’t realize there was a schedule, or a “final run for the day.” Usually I submit these things between 6 and 9 p.m., and it shows up online a couple of hours later. When I couldn’t find evidence of it the next morning at 8:15 a.m., I got concerned. Now I know, and maybe other people know, too.
(Perhaps that information is already somewhere in the TD/A FAQ, but I failed to read the fantastic manual. 
)
And again and again until the ratings are stable. Surely we can see a difference.
Alex Relyea
I ran some tests a few years back, most of the change due to rerating occurred during the first rerate. (That was why we decided not to compute norms from an event until after that event had been rerated at least once.) At that time about 25% of events were being submitted out of chronological order, I suspect that’s gone down since then, since most events are now rated by the end of the 2nd day after they end.
However, it only takes one large event submitted a week or two late to cause significant changes in several other events, and that can cascade to hundreds of other events.
The other thing that causes major changes in rerating is an event correction, especially a change in IDs. Changing a single result (eg, from win to loss) doesn’t seem to have as big an impact as changing a single ID.
One of the unintended consequences of the change to storing post-event ratings in floating point is that the number of minor (some might even call them insignificant) changes has skyrocketed. Back in the days when there had to be a full point or greater change in someone’s post-event rating to cause an update to the database the number of changes even after a major disruption (large event submitted late or an event with multiple corrections) was in the 1-2% range, now it is not unusual for a weekly rerate to update the ratings for 90% of the players in a tournament, usually by a very small amount, like .01 or less. This has had an impact on the amount of time it takes to do a rerate, the amount of data storage a rerate takes, and the number of changes that need to be posted to MSA.
It might be worth revisiting the study referenced in my first sentence.