Is there a database somewhere that has pre-1991 USCF ratings? It would be nice if the USCF started posting them just like they do now with crosstables (as a historical project).
This is the kind of project that frequently gets talked about, but seldom gets done. And the reason is simple – such projects are virtually impossible, or at least prohibitively difficult and time-consuming.
Even assuming USCF still has a complete set of paper copies of old crosstables, these crosstables would have to be scanned and processed with OCR (optical character recognition) technology, an iffy technology at best. I seem to recall Mike Nolan telling us that, when OCR was attempted on an old rating supplement, there was a 5% error rate in reading the ratings.
And then there’s the matter of USCF ID numbers of players (deceased or inactive) who were removed from the database prior to 1991. They don’t do that anymore, but what was done is done. Heaven help us even more, if any of those deleted ID numbers were ever re-assigned to other players.
For the above and other reasons, this project would be huge and uncertain, to say the least.
The latest revision to the data entry software made earlier this year has a ‘historical archive’ option, which would permit staff or volunteers to enter old crosstables for events initially rated prior to December 8, 1991. There would be additional fields activated for information on missing USCF IDs (name, state, expiration date) and for the pre and post event ratings for all players.
We have a number of old events, probably most of the events rated from 1983 through 1991. We have estimated that it could take 5000 hours to enter them all.
We believe the office has a complete set of rating supplements from before 1992.
Entering rating supplements manually would also be very tedious and time-consuming. We did look into having a document archival service scan them, but there was no funding available at the time so that project did not go any further. The estimate to scan all the annual ratings lists was around $9,000. Whether the scanned documents would be available just as PDFs or whether it would be digitized using OCR software was unclear.
It would also be possible to scan the old crosstables, but we do not have an estimate on what that might cost.
Well maybe as a first step the Crosstables that exist could slowly be scanned in and saved in jpg format, forget OCRing them. Just save the jpg with the title of the crosstable with the year and date included. Start with the newest first and gradually work backwards. Over time, even if it took years they would become available. The trick is to get started at some point.
Maybe as a first step just focus on getting 1991 or 1990 done.
Just to clarify, were you asking just about rating supplements, or were you asking for tournament cross-tables. Some of the other posters immediately started talking about cross-tables and that may be more than you were looking for.
It would really be nice if USCF started working back, perhapas one year at a time, and posted player information for regular ratings similar to what is on the website now.
BTW, a long time ago the USCF would sell a cross table of a tournament for $2. Does anyone know if that is still the case? If so how much?
When uploading an event the TD has an option of paying an additional $2 for a cross-table to be sent. It shows the pre-tournament ratings (or pre-section for players listed in multiple sections), not the supplement ratings used to pair.
The option seems of questionable value to anybody that can just go to MSA and get the exact same cross-table for free.
I have doubts that pre-MSA cross-tables would be available for purchase.
It is possible to order crosstables of events not in the rating system, but the person ordering it must provide enough information for staff to easily locate the event. Saying “I want that event I played in back in 1982 or 1983” would not be adequate.
There is also no guarantee that the staff will be able to find a specific event, even with a complete description of it, either because they don’t spot it in the huge stack of printouts or because it just isn’t there.
I played rated chess from 1968-1972, then 1976-85, then not again until around 1998. I’d pay for the Xtables from my first dozen tournaments. Not much, but I’d pay something just to see if my sieve-like memory is anywhere near accurate.
For this task, it makes more sense to volunteer $$, and scan them straight to PDF. Then if in a searchable DB volunteers can help with the OCR and proofreeding.
I hope I am wrong, but I will bet if they do post it, it will expose some of the cheating/fake results that was exposed when MSA first came out. The Forums will be filled with more stories I am sure.
Now, this is 21+ years ago. By the time such links from 1992 are up,
(if ever), it will be closer to 25 or so years. Memories dim, players and
organizers move on to other endeavors, including Heaven. Thus “filled”
is probably an exaggeration.
Robby, you’re probably not wrong, and I think it’s always good when the light of day clears things up. MSA has been very useful and helpful in this way.
My concern has come when the rumor mill create its own “truths” about an accusation without due process and taking the time to research facts. If there are such issues found, they should be dealt with appropriately with due process.
I thought that when MSA first came out it was a bit of a surprise and it covered a time period from mere weeks earlier to 12-13 years earlier. That relative immediacy caused results to be scrutinized closely while memories were still reasonably fresh with any anomalies still having an effect.
Adding the 1983-1991 period would involve tournaments 20-30 years after the fact and a player who played even one tournament a year would have had erased any rating blip effect caused by a problematic event.
There may be problems found, but filling the forums seems likely to be a bit overstated.
Adding the pre 1991 events to MSA will have no impact on anyone’s current rating, because we only rerate back to 2004.
It may, in a few cases, provide us with the most recent rating for someone who has been inactive for 20 years or longer, saving the office from having to post that old rating manually, and it may help in computing norms and Original Life Master games standings.