I just tried to submit my tournament online and got two errors back.
The first one was an expired membership. The member showed me his carbon copy renewal at the tournament.
The second was a poor performance (by me) with this error message:
Overridden: Possible Bad ID, Reasons::Q
Q - Player’s performance rating is more than 600 points below
his current rating.
Unfortunately, Q is me and I really had a bad tournament.
What do I do now?
My Affiliate is : A6018788
The membership issue will require an override, a mechanism that I’m hoping to have in place in the next few days. I’ll post more details on this at that time. (You will have to indicate who issued the membership receipt and when.)
The performance issue can be overridden using the check box in the crosstable edit form.
However, because of the membership issue, you won’t be able to submit the event until the membership override request form and the mechanism for responding to override requests is in place.
A hands-on involvement in the error checking and correction process (with occasional delays) is something TD’s will have to get used to and, yes, it is a radical change from the old ‘sign it and forget it’ days.
But as I have said on more than a few occasions, the USCF has always held the TD responsible for the accuracy of the events he or she submits. The difference is that now the TD actually has the ability to do something about errors before the event is rated.
Can you manually bypass this for now so I can get this tourney submitted? Or will I need to wait until this process is done online?
It won’t make this week’s rate batch either way, Nancy wants to cut off as of last night.
If you don’t mind, I’d rather keep it as an easy ‘test case’ for the automated override programming when it is ready for real data, probably over the weekend.
No problem, I figure keeping it in this holding pattern is going to be faster than mailing it in.
On a side note, I have a forfeit Draw last round. I was unable to get the result after repeated attempts to contact either player. I assume this is the correct way to figure it out?
I think it would depend on whether there were any prizes at stake. I’d be inclined to record it a double-forfeit loss myself.
However, I’m not sure if the program would consider that as an inconsistent result, I’ll have to try it.
A double-forfeit draw should pass muster.
BTW, Nancy and I just finalized the cutoff schedule. We will cut off online submissions at 8AM the morning before a ratings batch. Thus, if there is a batch on Tuesday (which is the day she prefers to run it), the cutoff will be 8AM on Monday morning.
There may be a batch run tomorrow, she finished the rating supplement last night and it is now at the printer. The event submitted through this morning should make that run.
The good news is that so far every event we’ve run from the online submission process has passed validation on the first try on the old system, which means that the data checking process on the new programming appears to be catching everything that the old system did.
It’s a major asset to have something like this on here. It’s a pain in the butt to go through and physically mail this off.
Happy to see it nolan, keep up the GREAT work!!!
Does this mean that anything not already in MSA as of last night won’t make it into the annual supplement? The batch being done tomorrow won’t make it either? Just curious because our club championship seeding hinges on the annual supplement.
USCF has generally used the end of October as the cutoff date for the Annual List. This usually means that the last rate run before creating the annual list has been the first rate run in November.
This year, that was the one run on November 5th. It included a few events that ended the weekend of October 31st but mostly ones that ended before then.
That rate run, as has been USCF’s normal procedure, did not include any events that ended on or after November 1st.
The whole process of ratings supplement cutoffs will have to be reviewed in the light of real-time event rating.
The data files that match the printed supplement are usually posted around the time that the supplement is sent out, so I would expect those files to go up on the website fairly soon.
But in answer to your question, any event that doesn’t show in MSA by now will not make the annual list either.
Certainly, a double forfeit loss should not be regarded as an inconsistency, as it happens all the time.
Probably, a forfeit loss for one player vs a forfeit draw for the other should not be regarded as inconsistent, either.
The only kinds of double forfeit that should be considered inconsistent should be those in which the sum of the scores exceeds 1 – i.e. a double forfeit win, or a forfeit win vs a forfeit draw.
In the meantime, there is of course a workaround – just replace the double forfeit loss by an unplayed game (U or blank, with no opponent listed) for each player.
A double forfeit draw is not a good way to handle an unreported result, as one of the players may then blunder into a prize as a result of a half-point awarded to him in a game which he may actually have lost.
Bill Smythe
When I asked TDCC Chair Tim Just, he thought that double forefeit wins and forfeit win + forfeit draw are possible results, though both are fairly rare.
I’ve changed the program so that it will permit any combination of forfeitedresults as long as both players have a forfeit as a result. (Most crosstable programs appear to change the opponent number to 0 for a forfeited result, which was already permitted.)
Tim also thought there might be occasions when non-forfeited results of Win/Draw might be reasonable, but I haven’t asked the Ratings Committee about that yet.
I have heard of this happening, but usually because the TD doesn’t have the guts to make a ruling and just wants to please everybody. I certainly hope that a (non-forfeit) win vs draw, or double win, will NOT be permitted by the new software. For one thing, it could catch on, with disastrously inflationary effects on the rating pool.
Bill Smythe
Unless the Ratings Committee gives its imprimatur on such a result, all RATABLE games must have consistent results, e.g., when A plays B if if player A lost, Player B has to have won, etc…
It is capable of finding cross-round pairing in adjacent rounds, but not beyond that.
For those who haven’t signed up for access to TD/A yet, it will color code the online crosstable to make certain types of results easier to find and review.
For example, items in RED are inconsistent results that must be corrected.
Items in orange are where two players meet more than once during the event.
Byes are shown in light blue.
Forfeited results are shown in yellow.
Cross round pairing are shown in green.
Why that limitation? From a programming standpoint, it would seem just as easy to check for cross-rounders in non-adjacent rounds as in adjacent rounds.
I once directed a smallish CCA tournament in Florida where a player had been given a full-point bye in round 1. I had asked her several times during the tournament whether she would like to replace her bye with a played game, but she kept saying no. Finally, after she went 0-4 in rounds 2-5, and another low-scoring player had waited an hour for a no-show opponent in round 5, she and he agreed to a cross-round pairing – round 1 vs round 5.
Bill Smythe
I would argue that someone playing their 1st round game against someone else playing their 5th round game is not really a cross round pairing, it is an extra game and should be submitted as such. You’ve been carrying the first player with a bye or unpaired game for several rounds, that’s not the purpose of cross-round pairing.
Programmatically, I would have to write a separate check for potential cross round pairings for each round beyond the adjacent ones, because there are things other than cross-round pairing that are being checked for at the same time that I would not be checking for in that new code segment.
I tested several months worth of events. I think I found a grand total of 1 with what appeared to a cross round pairing that didn’t involve adjacent rounds.
True, but it might be LISTED as a cross-round pairing on the rating report, just as a matter of convenience for the TD or organizer.
I have done something similar, in the crosstable-checking programs I used to use as ICA Tour Statistician.
You can adopt a two-pass approach or a one-pass approach.
In the two-pass approach, go through each round for each player (so that for a 5-round 10-player tournament you’d have 50 cells to check). At each cell X, check for a matching result in the same round for the player’s listed opponent (cell Y). If the result matches, mark both results in red (so to speak – or change a flag from 0 to 1 for each cell X and Y). As you proceed, any results already marked in red need not be checked again, of course.
Then, go through the crosstable again, looking only at cells not yet marked. Whenever an unmarked cell X is found, go through ALL the rounds for the listed opponent, looking for an unmarked cell Y with matching results. If one is found, mark both cells X and Y in blue (or change the flag from 0 to 2 for each), and end the search for X’s mate.
At the end, any unmarked cells represent unmatched results.
The almost-equivalent one-pass approach would work as follows. Go through each round for each player, as before. At each cell, check for a matching result in the same round for the player’s listed opponent. If the result matches, mark both results, as before. If the result does not match, check ALL other rounds for the same opponent, looking for a (not yet marked) matching result, which would then represent a cross-round pairing.
The two-pass approach might be slightly more accurate, as the one-pass method could stumble if a player is mistakenly listed as having played the same opponent twice, but the one-pass approach might run faster and would ultimately produce the same result once all errors are corrected.
Just some old-fashioned advice from an old TRS-80 programmer.
Bill Smythe
Mike,
I just submitted a tournament online, but it is not showing up in the “Edit Previous Tournaments” field. I think there are a total of 5 previous test tournaments there. Is the a max that the field will hold, and are some overwritten?
Once you submit an event it isn’t on the ‘edit’ list any more. That’s to keep you from submitting it more than once.
After it’s rated, you will be able to look at it using the ‘history’ program, which will eventually be able to do corrections, too.
Hi - just tried to enter my first tmt online … Glad it was a small one! It is great gettnig immediate feedback.
One note - we ran a kind of tmt that I think we old fogeys used to call an ‘insanity’. THe time control was different for every round - which wasn’t a big problem (I divided each section into dual-rated games only, and quick-chess rated games only). however the ‘Time Control’ field wanted (it seemed) just one G/nn figure and so I lied. OK with you?
Thanks - Bob
The Ratings Committee has concerns about rating only SOME of the games in a section under one rating system. Their opinion is that either ALL of the games in a section should be rated under a rating system or NONE of them should be.
How to translate this into which ratings system choice to indicate (dual/regular/quick) is sometimes not all that clear. In general their recommendation is to use the SLOWEST of the time controls in effect, especially if those are the latter rounds, which is usually the case.
Suppose you have rounds 1-2 at G/15 and rounds 3-4 at G/45.
If I understand the RC correctly, this should be reported as if the entire event was at G/45, which would make it dual-ratable.
Splitting a section up and reporting it as if it were two different sections is another way of handling it, but one that will almost certainly impact the post-event ratings a bit, probably by decreasing the amount of change in most player’s ratings. (I think this is because the bonus formulas are less likely to come into play.)
Out of curiosity, what time controls did you use?