You need to spend one tournament following Robert Tanner or Robert Singletary. “win-draw” is one of their favorite methods of resolving difficult cases at National events. I finally saw enough of these to absorb the general idea, and once awarded a “win-win”.
My first choice in these cases is that the inconsistent result should be used for tournament purposes, but not for rating purposes. In the old days, I would always win these arguments by saying “the softward can’t handle it”. Now it does. On balance, I’d say this is one area where the difference can make a big difference to one player (or his Mom) - but makes no real, lasting difference to the rating system as a whole.
We get enough e-mail about events where the player (or more likely a parent/coach) computes the rating manually or with the online rating calculator on the website and complains about any differences that I’m very reluctant to have situations where the rating and crosstable use different results.
Thus, while agreeing that the long term impact is minimal, I think it is best to be consistent about how we handle inconsistent results.
BTW, the latest interesting request is from someone whose latest published rating is 666 and the person wants it changed to 665.
The problem is that many of the “inconsistent results” are assigned because of “off the board” administrative problems. That is, they are not necessarily tied to any ratable activity at the board.
for example, it might be correct to report a ratable game as 1-0, but award 0.5 (or even 1.0!) to Black for tournament prize purposes.
Or, it might be correct to report NO ratable result, but 1.0-1.0 to the players for tournament purposes.
The latter case is easy enough (simply nuke the unplayed - unratable - game and assign forfeit wins to both players). The former is more difficult.
And, believe it or not (it took me about 5 years to “get it”) it’s occasionally necessary to do this. “Necessary” in the sense that it’s the best choice from many (mostly bad) options.
Now that I know that these incocnsistent results are being rated, I suspect that (when I’m in charge…) these games will be reported as “extra rated games” to get the correct rating result, with “forfeit win/draw/loss” in the main section.
The bad news is that this will mean that the appeal hearing will last 3 hours
instead of the usual 2 hours. It will take that long to explain the final disposition to the players/coaches.
I am probably going to submit a Delegate motion for next year either banning the rating of such results, or attaching very stiff “TD justification” requirements. (In theory, such requirements exist now, but clearly are not being enforced.) Frankly, I can’t think of any circumstances under which an inconsistent result should be rated (as opposed to used on the wallchart to resolve some dispute). A game is won, lost, or drawn. Period. This should not be subject to political influence.
The problem with the circumstance of an inconsistent result, is a director not at the board when the problem did happen. When the players, with the players as the witness are in two different groups, both groups cannot be correct. If the director was not a witness, any final win, lose, draw from the director can be in error. The inconsistent result, can settle the issue, if it is accepted with the parties.
The inconsistent results should only be granted, if the director was not at the board during the early stage of the problem. The longer it takes to get a director, the board or the witness statements can be corrected from both groups. When the director does come to question, the players and witness statements can be false from both groups.
The inconsistent results not ideal, but if the director does more could find both parties should both be forfeited.
A player comes to you in the middle of the 4th round complaining that his 2nd round result is wrong on the wallchart.
It shows him as losing, he has a scoresheet that shows he won the game as white.
The other player also has a scoresheet showing that black won the game.
Both scoresheets have numerous errors on them, though they do appear to be the same game.
What do you do for pairing/prize purposes, what should the USCF do for rating purposes?
If you’re going to introduce a motion to deal with this, please take into account the following:
Whatever is done for rating purposes should be clear from the online crosstables. However, if the results for pairing and prize purposes are different from those for rating purposes, that should also be clear from the online crosstable. (We get several inquiries every week about results not being sorted in prize/tiebreak order, which is nearly impossible to achieve because A: We don’t know the tiebreaks that were used, and B: some tiebreaks cannot be recomputed from the rating report.)
There may not be a way in WinTD or SwisSys to report inconsistent results, as I think both programs check the crosstables for consistency. How should this information be conveyed to the USCF?
If there is to be a manual review of any kind for inconsistent results, this will require a reporting mechanism as well as staff time to review those reports.
Please allow reasonable time for any programming changes to be made and for any retroactive changes (if mandated) to be made to our crosstable records. My two least favorite words in Delegate motions are: 'Effective immediately".
Mike, when I do cross-sectional pairings (bye players in different sections, paired so that they have a rated game to play even though they have the bye for individual and team purposes) I do the following:
I put the loser of the game in the winner’s section and enter the result there (inactivating the loser) while the loser still shows as a bye in the original section. If the game is a draw then it goes into an extra games section so that the byes show up in the original sections.
If there are going to be a noticeable number of such pairings (such as the 30+ games in the most recent local 27-section scholastic I did) I’ll simply leave the byes in the original section and put all of the extra games in an extra games section.
I hadn’t considered the bonus points aspect. However, since the players with such pairings are the ones getting the byes and are thus not doing well, there probably aren’t many bonus points that might be affected.
What would you consider the best way of handling this?
Assuming no other information was available, I would probably give both players a full point for the round and submit it the same way (unplayed forfeit win for both players). Of course, if prize money were involved, I would make a pretty strong effort to find someone who saw the game.
It shouldn’t. Inconsistent results should not be rated, except in very rare circumstances by manual intervention. (You can force SwissSys to do this, by the way, by ignoring the error warnings, though I’m not sure what happens when the rating report is processed.)
As well it should. TDs should not be able to give out rating points like party favors. Doing something like this should be difficult.
I checked with the TDCC chair during the development of the online rating report system, he had no recollection of an event having ever been referred to TDCC for approval of inconsistent results.
I can’t find it at the moment (it is difficult to find anything on the USCF web page), but one of your “explanation of the rating system” brochures states explicitly that inconsistent results shall be rated only if accompanied by a written explanation from the TD. I believe it then says this “may” (rather than “must”) be referred to the Rules Committee for approval, but if, as you say, such requests are being approved routinely, it’s an abuse of discretion.
Let’s try it another way: Where in the rating regulations does it say that inconsistent results should ever be rated? It’s up to you to justify so irregular a procedure. If there is no written rule (just a demand for special treatment by a few NTDs), then shouldn’t it be stopped now?
Note that I am not advancing this as a legalistic argument – even if such a rule can be documented, it is obviously a dead letter. Nor am I claiming that you and the office are acting in violation of the rules, though I’d say the burden of proof is on your side.
What I am saying is that rating inconsistent results is a very bad idea.
Now, there are two possibilites here. The first is that the office backed into this without anyone thinking it through. (It would never have occurred to me to pass a rule against rating inconsistent results, since it would never have occurred to me that anyone would try anything that silly). The second is that this procedure was, in fact, approved somewhere along the line, without anyone really noticing it. (I doubt if many people read closely all those Ratings Committee reports, for example.) In either case, now that this aberration has become visible, either the EB or the Delegates should squash it.
