Simplest of all would be to define a new standard output data file for developers to switch to in the future, that has all the data that we want.
Of course, it might take 10-20 years for TDs to switch over to programs that use the new format. (We know for a fact we still have TDs submitting events with versions of pairing programs from the 1990’s.)
And about 5% of events are entered online, so the online data form would have to change as well.
That seems to make this not all that simple.
That makes an assumption that we already know just what type of data we will ever want.
That makes an assumption that we already know just what type of data we will ever want.
[/quote]
Nope. Chess players seem to have a funny way of looking at things. All it assumes is that we’re capable of making significant progress.
Systems evolve, they are never done; the idea that we would ever know everything that we would ever want it be doesn’t make sense.
Time for us to start leading that process then.
There is nothing wrong with the charts as they are now. In fact despite what some may say they are useful in the sense that one can see right away if someone has tied for a place. “Look there are three people with 3 points and one with 4 points. Wow the lower rated player actually tied for second.”
You don’t actually know that. There could have been unrated playoffs. It could have been a team tournament. Wins might have counted for 3 points and draws for 1. It might have been an Australian Draw, a King-of-the-Hill tournament, or a McMahon tournament. Players may have played varying schedules and different numbers of rounds, with some other system for determining the winners than simply totalling up won games.
Basically, you are assuming that the tournament was a Swiss, though the USCF does not require this for the tournament to be rated.
And all of these estimates imply that there will be restrictions on what type of tiebreaks are used. I’d rather not see such restrictions ever implemented.
This speaks to part of my point, as stated earlier in the thread. The increase in complexity won’t actually accomplish much for the MSA-browsing end user, especially since tiebreaks aren’t totally standardized.
I don’t know how an organizer who runs a tournament such as this one could ever report the final prizes in MSA, unless they could just upload their own prize report as part of the tournament submission. This seems like it would be, by far, the preferred solution, both for ease of implementation and completeness of information.
But I do know it because I am looking at tournaments that I played in. And for those that maybe I’m looking at for some other reason (can’t think of any off the top of my head) I don’t really care. What is there is close enough.
It doesn’t have to be perfect. If I need more perfection then I would go to the source. I imagine that there are others like me that feel the same way. The system we have has been working for years and now all of a sudden because someone thinks that explaining to someone how the system works is too much effort and that there should be a better solution than explaining how the system works the system needs changing. I think your time would be better spent explaining the grammer of that previous sentence.
Although the most common tie-break sequence is Modified Median, Solkoff, Cumulative, Opponents’ Cumulative, there are other sequences used. The scholastic nationals replace the third in the sequence with Sonnenborn-Berger (primarily because the national scholastics often have accelerated pairings, which negates part of the logic behind cumulative). Some state scholastics with accelerated pairings have mirrored the national scholastic sequence. Round Robins generally start with Sonnenborn-Berger (all of the logic behind cumulative is negated and the identical opponents’ scores - for tied players - eliminates the usefulness of median, modified median and Solkoff).
Also, fixed board team events (with or without weighting the boards) would require a very different entry structure if prizes and tie-breaks were expected to be shown in MSA.
The MSA does show full and half point byes, does it not? These are irrelevant for rating purposes. So, it would seem that at some point the MSA was trying to capture information that would only be interesting if it was trying to report final tournament standings, at least for regular Swiss tournaments.
However, it would also seem that this goal was only half-implemented, with the result that you cannot get final standings out of the MSA. Why not try to implement that fully? Why was it implemented only partially the first time around? It does not seem that it would be so hard to do a proper job of it, and it would make MSA even more useful.
But Wayne, the funny thing is that I’m one of the people that would like to continue to upgrade MSA and I feel EXACTLY the same way as the above. The fact that its never going to be perfect and that it doesn’t have to be isn’t a reason not to improve it.
If it WERE a reason not to improve the system of reporting and recording rated events, then we’d never have MSA to begin with!
Why let the system continue to fall behind like we have in the past - so that every upgrade is a an even larger undertaking. Let’s start planning for next steps now, and start implementing things now that will help us make a better system over time. System improvement is a process, not an endpoint.
Take the above statement and apply it to the system of submitting events on paper and publishing the ratings only in a magazine.
Do you notice the parallel?
LOL.
Look, I’m all for better uses of the resources - maybe a better use would be integrating pre-1991 norms.
But the point is that we can make a system that is likely to resolve some questions and that can be easy to use. That’s not a be all and end all - no one claims it is. It’s an improvement step and THAT step shouldn’t be that hard.
Another improvement step would be to start thinking about what data we actually want and need and how we are going to change reporting definitions and get people to move forward.
Just because something served the needs of the past, is no reason to ignore continuous improvement.
These are good points, and a great example why the steps we’ve suggested are JUST STEPS and show that even more needs to be done over time.
Brian noted that MSA doesn’t have to show the crosstables to show ratings; in fact, the magazines did not. Not only was the MSA a step up in the information delivery system, it improved the information delivered.
There’s nothing wrong with continuous improvement, so that as people ask “why is USCF relevant?” we can say “here is why” rather than “I dunno, they haven’t changed their systems for 10 years.”
Kevin I’m not against improving the MSA. I just have yet to see any suggestions here that I view as improving anything.
Then I don’t see why.
By providing a (or some) limited sort options some questions that now go to the office get answered online. By including an explanation and a link there is no downside (certainly none that exceeds the current downside) and it is simple to reach the organizer - who would have had the question eventually anyway. So, there doesn’t appear to be any downside to the organizer.
In addition, it starts a process and a path toward a continued improvement to eventually make MSA full tournament reporting. This of course provides an upside for the organizer who will no longer need to pay for a space or for the time to post these results online.
Of course, the idea is an upside for players who have more information and explanation available.
Another upside is that we work toward further standardization of data.
I’m struggling with the issue here and the negative response this is getting. Bewildered in fact.
I actually did see one posted here, but it’s back a little ways.
Why Brian M’s proposal, to display crosstables in anything other than score order, is still being discussed, is a huge mystery to me. Almost everybody wants it in score order, almost all the time.
If you’re looking at a McMahon event, or a team event, or a draws-count-one-third event, just go to the organizer’s website. Surely any organizer who runs such things would put them on their website.
Brian seems to want to punish anybody who wants even more, by taking away what they already have.
Not if you do it correctly. When using cumulative with accelerated, you should ignore the first round. For example, a player whose wallchart reads 1,2,3,4,5 would have a cumulative tiebreak of 2+3+4+5. (I think that’s in the rulebook, though I don’t have time to look right now.)
All very true. S-B, however, tends to reward inconsistency. The player who beats the masters and loses to the fish will have better tiebreaks than a player who does just the opposite.
Bill Smythe
Actually, my “proposal” if you want to call it that was:
(1) Let the TD’s upload the final results order and display by that; or
(2) Let the TD’s tell you that the event was “standard”, and compute scores
and tie-breaks according to the USCF rules, and display by the resulting order.
(3) Failing either (1) or (2), or when you have implemented (2) but the TD’s don’t tell you the event was “standard”, do NOT compute and display a “score” (since you don’t know how to compute it) and sort by something else reasonable, such as pre- or post-event rating, which you do know. Pre-event rating would be the wall-chart order at the event, more or less.
“Scores” as currently displayed by the MSA do not come from the TD. They are computed by the program. The problem is that the program does not know how to compute a tournament “score” that is meaningful in terms of tournament results, unless it assumes that the event was a standard Swiss. Assumption is the mother of all f-ups.
Mike Nolan has answered this numerous times. The more data sought,
the more TD work in entering the data. We are loosing TDs because they lament the time it takes now. There are many fields, that from
what I gather (Mike when you get bored you can help me here–lol),
such as school, grade, graduation year, etc, that for the most part,
do not get filled out. Adding more data fields, thus, will not guarantee
a fountain of new data.
rob jones