Join Bridge Winners
All comments by Max Schireson
1 2 3 4 ... 33 34 35 36
You are ignoring the author of this comment. Click to temporarily show the comment.
Phil,
I would take the weakest 1-16 in a heartbeat. Seeding, however well done, is not that accurate. In the top 16 for example a sponsor may be a much bigger liability than indicated by their seeding points.
That said I think the problem is not symmetrrical and the best of the bottom 16 is a much clearer favorite over 48 than is 1u over the worst of 1-16. At the bottom of the field, it is very realistic to have a whole team of players umderrated by seeding points; the converse is in my opinion less likely at the top.
8 hours ago
You are ignoring the author of this comment. Click to temporarily show the comment.
Justin,
I don't think this is fair to 16 (or probably 15 either) but if you allow them to pick from a wider pool than 49-64 (say bottom 24) then it is at least somewhat better. Of course then 17 or 18 or 19 will have a rough draw - whomever is stuck with the strong team nobody wanted.
The irony is that in the current system a higher seed gets stuck with that opponent, but it was an unknown high seed. If we adjust the system so that that high seed gets to avoid the tougb match, it feels unfair because a lower but known - or nearly known - seed faces the same problem.
To an extent it comes down to the definition of fairness. Is it more fair for a random near top seed to play the badly inderseeded team, 16 to play them, or for it to be randomized a bit in the 17-plus-a-little zone.
March 27
You are ignoring the author of this comment. Click to temporarily show the comment.
Well, if the 5H bid is natural then there is no problem. If it is otherwise then there should be whatever penalty is appropriate for playing an illegal undisclosed convention.
March 25
You are ignoring the author of this comment. Click to temporarily show the comment.
John,
I would argue that this approach reduces the impact of a small difference in seeding points at the top. Under the current system Nickell (2) played 63 (Irish Juniors) and won by 300. Because of a few less seeding points, Fleisher (3) played NRK (62) and lost when there were plenty of teams available that they would have comfortably beaten by 100 imps or more.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Geoff,
The problem with that is that seeding points are also given for masterpoints. A team of players with say 3000 masterpoints each but no success at the national level is seeded higher than NRK. Their 3000 masterpoints each do not really represent sustained high level performance.
I would argue that even a close loss against a top seed means more than 3000 masterpoints.
I agree with you that sustained high level performance should be required to get a high seed, but I would argue a day of competitiveneas should go a long way towards shedding a very low seed; there were plenty of teams with no chance at all (mine among them) and those with even a small history of success could have been objectively differentiated from those without, which imo would have resulted in a fairer event.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Its not perfect but it is something of an improvement imo.
Assuming multiple teams are mis-seeded, n+1 will get the lowest seeded team among the mis-seeded teams, vs getting the most mis-seeded team. Assuming the distribution of mis-seeded teams among the bottom seeds is pretty random, it is hard to predict exactly what seed you want to avoid. Someone still gets bad luck, but I think someone in 17-20 getting bad luck is better than someone in 1-8.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
The top n should be picking from more than the bottom n, since some of those will be mis-seeded. Top 16 pick from bottom 24 seems reasonable.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Brian,

Another approach that might help avoid NRK vs Fleisher is to continue to seed objecitvely, but to allow the top N seeds to pick their opponents. Speaking as captain of a team that plausibly might have been picked quite early, I can assure you I would not have been offended. I suspect that most of the other “top pick” candidates would have felt similarly.

This would also help with the problem Adam described; if you let the top 16 choose their opponents, a strong team like NRK that is seeded very low is likely to wind up bubbling up to play a team in 17-20, which is at least a somewhat more reasonable path.

This preserves objectivity in selecting the top seeds, but makes those high seeds more valuable and avoids the negative impact of objectively seeding the very bottom of the field where errors are structurally most likely.

Thanks again for both doing the work on the committee and amswering these questions.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
In substance it looks reasonable.

I know this isn't what you asked, but I would definitely avoid calling something used in these games a “novice” chart. Players in Gold Rush events might be novices by BW standards, but many of them have played regularly and successfully in limited games for many years.

I think “intermediate” is a better description; I think a separate chart for 0-20 games for example could be called a “novice” chart. Surprising as it may sound, I think “intermediate” might even offend some gold rush players who are life masters and are no longer playing in events labeled “intermediate/newcomer”.

It's all relative. I was not at all offended when Bart Bramley said on day 3 of the Blues that I was doing well “for a beginner” because I was a beginner relative to him and almost all of the other contestants, but I suspect many in gold rush events consider themselves “advanced”, and they are relative to some of the 299er players, who have played for years and consider themselves “intermediate” relaitve to the newer players in that game and those in 99ers.

Sorry if you consider this “off topic”.
March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Brian,

First thanks for doing this.

I think you are almost right, but your parenthetical contains the key.

If the committee seeded the event I agree 100% it would generate a ton of all those complaints and would, even if it produced fewer upsets, be a disaster.

However if the field seeded the event, and a team of foreigners/women/juniors/Martians pulled off some upsets, I think the response would be “wow, next time the Martians will be seeded higher!”.
March 23
You are ignoring the author of this comment. Click to temporarily show the comment.
Brian,
Great question.
In this specific case I mostly mean that Fliesher would not have had such a difficult first round match. They earned their high seed and didn't receive much value for it; I view that as unfair.
I think a system like Eugene and others have proposed is not vulnerable to gaming but would not produce matches like Fleisher/NRK. Thus I prefer it.
Is your preference for an objective system practical (for example, fear that this other type of system could be gamed) or inherent (for example, belief that rewarding past performance is itself a good thing)?
March 23
You are ignoring the author of this comment. Click to temporarily show the comment.
Brian,

Eugene's suggestion or some variant of it addresses all the challenges with objective methods, and I think while subjective is “fair” in my opinion.

Do you think that type of system would be seriously considered?

I would encourage the committee to not focus too much on objectivity when an alternative is avaable that addresses so many concerns.

I think my team would likely have been seeded about the same as our actual #60 under this system, so I have no axe to grind personally, but I think overall the seedings would have been much fairer.
March 23
You are ignoring the author of this comment. Click to temporarily show the comment.
Brian,
Thanks for your effort to make the process more open.

In general I think a system like Eugene suggested is the best approach to seeding, and could be used for things like regional GNTs as well as the big events.

If you are going to use objective measures, it seems that platinum masterpoints are the obvious missing element. Giving them more weight and other masterpoints less seems like a clear improvement. A bracket 3 knockout win says nearly nothing about my spingold chances, but some overalls in other NABC events mean a little more (though not as much as the events that earn real SP).

To keep it simple, you could do (platinum * 10) + (other / 2) or some such, and otherwise use the same log scale.
March 22
You are ignoring the author of this comment. Click to temporarily show the comment.
Nat,

You seem very knowledgeable about bridge pros and their motivation for someone who doesn't seem to play with or against them face to face. Do you have some old friends that have gone pro? I would love to hear more of your perspective and where you got it.

I hope you are making progress towards your LM. Maybe if I am at the Fairfield regional again this summer or New York this winter we can meet up for a beer.
March 20
You are ignoring the author of this comment. Click to temporarily show the comment.
Gotta root for Aaron. Winning by picking yourself gets style points. Has it ever happened before?
March 17
You are ignoring the author of this comment. Click to temporarily show the comment.
Sucks to be 3rd and not have bragging rights in the sibling rivalry.

Congrats to all 3 pairs at the top on a strong first day.
March 17
You are ignoring the author of this comment. Click to temporarily show the comment.
When the UI is not from partner it is treated differently. See law 16C1, which says to summon the director. The director can award an adjusted score. You do not have the same obligation to take the LA wbich is not suggeated that you would if the UI was from partner, but under law 73 you still have to avoid taking advantage of the UI.
March 13
You are ignoring the author of this comment. Click to temporarily show the comment.
Congrats to Zach and Adam for leading a strong field, and good luck tomorrow!
March 11
You are ignoring the author of this comment. Click to temporarily show the comment.
Oleg,
I share that concern, though it is somewhat mitigated by the official investigation focusing on different allegations about the same pair.
March 11
You are ignoring the author of this comment. Click to temporarily show the comment.
I agree that the process was broken and public accusation and investigation was needed to get it back on track.

I hope that we won't need public investigations in the future. In this case the pair is already publicly accused so there is less damage in investigating them publicly, but it would be a shame if this became the normal process.

Kieran,


In any system of justice there is a balancing of the rights of the accused with the ability to get convictions. Yes, it would be a shame if private investigations failed to convict the guilty. But public investigations of the innocent are also a big problem - especially if not guilty verdicts are doubted.

I hope that this public investigation - which is really a continuation of the last batch - is the last one we see for quite some time.
March 10
1 2 3 4 ... 33 34 35 36
.

Bottom Home Top