I’ve noticed some data being thrown around in various threads (particularly by @nolan). Is this data made publicly available? Or is it available on request? For now, I’m specifically interested in 1) # of USCF rating submissions by pairing program over the past 12 months, 2) % of FIDE rated USCF tournaments over the past 12 months. Can I self serve to get this?
Enjoy the data I currently post while you can still get it, I do not know if I will have access to that information once the new ratings system is operational in early 2025.
You can probably get better data from the FIDE website about events submitted by US Chess than from the US Chess website. BTW, FIDE Academies can submit rating reports directly to FIDE without going through US Chess.
Here’s what I show for FIDE rated sections since June 1, 2023:
fide_rated sections events
---------- -------- ------
N 40342 16409
Y 1022 507
There’s a field for the pairing program in the 2C version of the upload file format but it is not used during validation, so I don’t know if we’re getting valid data in it, but here’s what I show for events uploaded in calendar 2023. Whether the N/A entries are because they used the original upload format, which doesn’t have that field, or because the pairing program didn’t supply a value is unknown.
program count
---------- -----
CAISSA 744
CHESS NUT 8
CHESSMNGR 6
SHRTCSTLE 25
SWISSSYS 1086
SWISSSYS10 4797
SWISSSYS11 4347
TOM2166 1
TOM2194 203
WINTD4.10 23
WINTD4.11 74
WINTD4.20 851
WINTD5.00 1746
N/A 4260
1 Like
I guess the N/A is for manually entered results. Is that correct?
If by ‘manually entered’ you mean entered online using the online editor, the answer is not necessarily, the original version of the upload format didn’t have that field and some pairing programs may not supply that data.
It appears that around 3384 of the not available entries were for events entered online, though that count probably includes around 526 events from chess.com.
I would anticipate #1 being either hard or impossible to get. What US Chess receives is a dbf format that (as far as I know) does not include the name of the program that created the data. An approximation may be possible if US Chess can correlate the submitting TDs with the pairing programs owned by those TDs (assuming the pairing program sellers would give that information) but then they’d still need to query those TDs that either do not own a pairing program (using one owned the an organizer they work for) or own more than one.
#2 would need specifications on how to handle tournaments with only some sections being FIDE rated.
Being something of a data hound, I thought the breakdown of sections coded by FIDE rated status and by ratings system might be interesting:
by
rs fide_rated sections events
-- ---------- -------- ------
B Y 74 64
D Y 276 70
Q Y 33 28
R Y 639 367
B N 2346 1608
D N 27086 8924
G N 255 255 (Foreign FIDE adjustment events)
Q N 3311 1868
R N 4298 2009
S N 879 471
4 N 1429 1301
5 N 750 738
Note my recent post. The 2C upload format (which also includes color information) has a ‘pairing program name’ field, but I cannot vouch for its accuracy or completeness.
@nolan one more data related question if you have a minute:
In this file format page:
There is a field called S_SCH_LVL with options:
‘N’ => ‘Non-scholastic event’,
‘P’ => ‘Primary event’,
‘J’ => ‘In-school JTP event’,
‘S’ => ‘Scholastic event’,
Do you know if that field is still being used? It seems like the import isn’t taking it into account. Thanks!
I don’t think either WinTD or SwisSys includes those fields, so the upload program wasn’t importing them, though as noted they are used during event validation, especially to denote JTP events, where membership requirements are not enforced.
The distinction between scholastic and non-scholastic (sometimes called adult) events is very subjective and is left to the discretion of the submitting TD to code, we included those codes because people keep asking us how many ‘scholastic’ or ‘adult’ events are being held. Is a quad where all 4 players just happen to be kids even though other sections have a mix of kids and older players a scholastic event? Is a 50 player kids event with one adult house player a scholastic event? (Reminder - house players who are not age-appropriate to the type of JTP event are not permitted in JTP events.)
I’ve added code that should hopefully upload that field, but I don’t have a way to test it. If you can generate 4 test events (one for each value) and send me the sets of DBF files, I will test them.
A post was merged into an existing topic: Safe Play Policy ADM
@nolan here is a zip with an event with 4 sections each with a different setting for that field.
Interesting. When I wrote my own personal tournament software, I just had S_SCH_LVL hardcoded to ‘N’, S_GR_PRIX to ‘N’, and S_GP_PTS to ‘0’, but I don’t remember ever noticing that I still had to select Non-Scholastic later.
That’s because the validation program tries to guess whether it is ‘adult/non-scholastic’ or ‘scholastic’ based on the ages of the players in the section.
But the TD must specifically select either of the JTP values in order for the event to be exempt from membership requirements, though players must still meet age requirements.
@nolan - did those test files I sent do the trick? Let me know if I can help. It would be great to remove this alert and be able to submit the field directly
I don’t see any test files in my email. (Nevermind, I found them.)
Looks like section 4 is incorrectly coded as N not S.
Mike Nolan
ah and chalk that up to human error on my part. So is this available going forward then? if so thats amazing. Thank you!
Yes, it should be in production now. This field can still be updated using the online editor. I also added a line showing the information in the ‘Program Name’ field, which is supplied by the pairing program author. For your test file it says ‘CHESS NUT’. (You might want to add a version number, the field length is 10 characters.)
1 Like
Since it’s a SaaS app, there really isn’t a version number. Everyday we are pushing improvements. Is there an advantage to having a version number that I should take into consideration? If we use less than 10 char I’m assuming (maybe incorrectly) that it should be ok