What about “Crosstable data is NOT SHOWN in tiebreak order and does not reflect any distribution of trophies or prize monies.” don’t you understand?
Giving them something which they think is useful but actually isn’t is much worse. If you want things listed in tiebreak order, then you’re going to need to have the tds supply that. In most cases that wouldn’t be hard. Just have them upload in standings order and then don’t reorder them. But don’t pretend to be able to calculate something useful when you don’t have sufficient data or defined default tiebreaking procedures to do it.
Yes, but why list the players in order of number of wins (counting draws as a half a win), when this does not necessarily tell you who “won” the tournament? There are plenty of other equally interesting orders.
If the organizer provides the entries in his standings order, then that would be the relevant and hence the most interesting order. Otherwise, maybe the best is to just use wallchart order rather than do a sort which may not correspond to the actual standings.
We don’t know what order the TD uses for upload files or when entering events online.
The chances are pretty good that the TD doesn’t know what order WinTD or SwisSys used when generating the upload files, either.
When MSA first came out (in 2003), we did not re-order crosstables into score group order, and that resulted in MANY complaints, far more than we get now.
Whatever order the TD supplied them in. That was before online submission, so some events would have been done from the three DBF files, sent to the office on diskette, others would have been entered by the USCF staff from printed crosstables. In 2003, less than half of the events were sent in on diskette, though that did include most of the large events.
For events with DBF files generated by WinTD or SwisSys, the TD may not be able to control (or even know) that order.
If the previous system was that the players were in no particular order (i.e. random order), then I can see why that might have led to complaints. I don’t think anybody is advocating that the list of players be unordered.
The current system is to sort by a computed point total followed by pre-event rating.
The point total is not being submitted by the TD’s, right? The program is computing it and using it as the primary sort. The discussion is about whether the program should compute a point total based on wins+draws/2 and use that point total as the primary sort.
The problem is that the point-total sort apparently creates the impression that the results are being presented in “standings” order. That impression is wrong, and necessitates a warning message, which some people may not read, or may not understand.
The order wasn’t necessarily random, it just wasn’t known or consistent from event to event.
It seems likely that if we took a vote among players, they’d probably opt for the current (computed) score group order over some other order, like pre or post event ratings, despite its limitations.
But the interesting question is what SwissSys would do if you were to change the “primary wall chart sort” to “Standings (with tiebrk)” in the formatting options dialog.
Close. Wall chart order would be the rating used by the tournament for the player. MSA doesn’t know which rating was used. It could make a very good guess most of the time, but sometimes there are rating adjustments made, CCA minimums applied, supplements for previous months used, most recently updated rating used, etc.
So if the wallchart pairing number is saved (sounds like it is) and you remember your final such number, then you are fine. Otherwise you may have to search for a little while.
Most of the people I’ve talked with are much more interested in seeing MSA in point score order than in wall chart order.
One advantage that hasn’t been mentioned for using rating-based wall chart order is that a re-entering person would have both entries together and it would be a bit easier to quickly see all of the opponent’s played.
Tom, just because you use current published ratings in that field, it doesn’t mean everyone does. (CCA minimum ratings could apply, for example, and not all events use the current published ratings in effect as of the start of the event.)
What it probably means is that it is in rating order as used in that event. What does it do if more than one player has the same pre-event rating for that event?
Wasn’t it Voltaire who said “The perfect is the enemy of the good”?
I’m not sure why this would matter. That’s what the software currently does. The optimal solution would be for the pairing programs to generate the USCF report based on the standings order (or whatever order the td specifies) - and then for the USCF to use that order when publishing to MSA.
Maybe the best listing would be in alphabetical order - easiest to find whoever you are looking for, minimizes the chances that someone will think they deserved a better trophy.
What does it do if there are 3 pairing #'s at a rating, but one of them is a re-entry? Are those two always going to be next to each other or could it be:
A (original entry)
B
A (re-entry)
Does that matter? Possibly not, but then does ANY of this discussion really matter?
Alphabetical order might not work well for non-westernized names.
Simpler might be to (as previously implied) have a place in the TD/A upload to put a web page address for the tournament results, and then have a link to that page added to the tournament report.
One caveat is that such links might become dead or hijacked (or the initial setup might intentionally or unintentionally link to an inappropriate site).
What’s wrong with how the crosstables are ordered now, other than that a few aliterate people who sail right past the disclaimer think they say something that they don’t?
That is what is wrong with the charts right now. It is all very well to speak disdainfully of people as “aliterate”, but that is just blaming the user for our own unclear and misleading presentation of data. The fact remains that the MSA is sorting by a field which it computes (namely, game point total), which was not submitted by the TD’s, and which people are rather reasonably assuming to be meaningful. If there is no problem doing this why is there a warning? Any time you are putting up a warning that says “THIS DOES NOT MEAN WHAT YOU THINK IT MEANS, STUPID” , it means that you aren’t displaying the data properly. Warnings like that should not be necessary.
What lesson do we want people to learn? Don’t pay any attention to the data which the USCF puts up on its website. It doesn’t mean what you think it means, and if you suppose it does, then you are obviously an aliterate moron who didn’t read all the caveats, warnings, and fine print. Is that our message?
I guess the same part you don’t understand when I also said to you that it should be disclaimed if sorted the other way?
When I said “tell them nothing” I mean there is basically no explanation of anything - I am advocating that changed. I assumed you’d get that from my sarcastic reference regarding tiebreak descriptions online.
Even so, I note that you still have yet to actually address any issues. You keep calling this worse, then presenting a non sequitur in defense of it, such as below:
Once again - argue to the point ACTUALLY MADE. We can sort the file in Standard Rulebook tiebreak order which in many (probably most, probably most by a large percentage) cases WILL BE USEFUL.
No, we don’t have to have the TD’s supply that.
We have sufficient data and we have sufficient procedures for the vast majority of cases.
Why are you avoiding this for what - 1%? 5% of cases? Maybe we should just take down MSA because 1% of the time there are going to be issues with it.
I think your issue is that you’re entrenched in a really bad argument and now you don’t know how to extricate yourself without saying “Oh, I was making a really bad argument.”
The idea that we shouldn’t undertake an improvement simply because it won’t be perfect is hogwash, Tom.