Dream Final Four in St. Louis

Should be interesting. I'm not a fan of the format, but it won't be boring for the fans.

Go Naka! You need the prize money to furnish your new apartment. Decent haircut is also expensive in St. Louis.

Just to note that there was no draw for colors, it was based on the fact that Nakamura and Onischuk had four blacks in the initial seven rounds, while Kamsky and Shulman had four whites.

If only Larry had seen Rxd5! But I will readily admit that Naka fully deserves his place in the final quad.

Question for Mr. Shahade: What would they have done if three of the four had a 4/3 colour allocation and the fourth had 3/4?

I forget there was some system in place that ended with random draw.

I think if Christiansen won he wouldn't have been eligible for the 6th white because he had white the final two rounds of the Swiss so would have been random draw between Kamsky and Shulman.

I have ideas to alleviate this possible problem
but I'm sure FIDE wouldn't allow them.

My feeling is that extra color should never just be granted to someone due to good luck, it's too big of an edge and results in annoying complaints. One way to alleviate is that in the final round, instead of just giving the color to one player, let them decide how much starting clock time thy would give up for color of their choice.

Because such a game couldn't be Fide rated I'm sure it wouldn't be allowed but it would certainly end the annoying tradition of final round colors being determined, or extra colors, simply due to fate/luck and aside from the fact that it's not traditional, all of the complaining about extra blacks just flies out the window.

There are other negatives to it as well in terms of how it would affect the quality of play of the game in question. It's just that it's basically proven that having white is a gigantic advantage and I don't like that players can do nothing to avoid this big disadvantage.

Five player final is annoying because a player receiving a bye every round looks really bad IMO.

take the top 3 and make it DOUBLE round robin.
No problem of any kind there.

take the top 3 and make it DOUBLE round robin.
No problem of any kind there.

Not a huuuuge fan of the format (though admittedly, you were very likely to see a lot of people at 6.5 if they just played 2 more rounds), but the right 4 are in the finals.

Not a huge fan of the format either, but I will admit that it's probably preferable to deciding the champion based on how well the leaders beat up on the non-contenders in the last couple of rounds.

I prefer a system where the tournament is used to pick the challenger against the champion. Then there is a match of the Challenger vs the Champion. That is the best method.

The tournament should have many fewer players. Top 4 or 8 is enough. A quick tournament and then the big match.

In the early history of the US Championship the Champion was determined by a Match between a Challenger and Champion. We should return to that best method.

Right now the challenger should be chosen to play Nakamura in a match. That would be truely exciting. It should also cost much less to stage because there would only be 2 players instead of 24.

For the love of all that is holy, please give a reason for not liking the format instead of just offhandedly saying you don't like it. I find it very bizzare that so many have done this. I can't imagine criticizing something without giving even one reason, and even some who say they don't like it and then not only don't give a reason, but instead write why it's good! So bizzare.

To Bobby Fiske

lol

Thanks for my first true laugh of the day, and keep it up !

As I already said in the other thread, I think that a Swiss is actually a pretty good way to reduce a large number of players down to a small number in only a few rounds, certainly better than a single-elimination knockout. The thing about colors is a pain in Swisses because if you have an odd number of rounds then some people get more white games than others, but an even number of rounds is even worse because some people might get two extra whites. I think that is why usually Swiss events are an odd number of rounds. I'm sure you could do some sort of forced pairings to avoid that "two extra Whites" thing from happening in the last round, but the effect of getting pushed over to play someone else, instead of your normal opponent, could even outweigh the color advantage of having an extra White or two, so I would think it is best to just go with an odd number of initial rounds and have the first tiebreak be color, and also to let the color breakdown from the first phase drive the color assignments in the second phase. Which I think is exactly how the tournament was indeed designed. Possible complaints? Well, Swiss events are pretty good at identifying a worthy #1 finisher but not so great at identifying a worthy "top N" group, especially as N gets bigger and bigger. So I applaud that it wasn't the top 6 or the top 8 with a subsequent 3-day knockout or anything like that. Could N have been lower? Having three players qualify would be lame because of byes in the final phase, so it could only really be N=2 as a potential improvement on N=4. And would that really have been better, especially if one player had a higher score and would just have to draw through a final two-game or four-game match in order to win? I think the actual format is pretty neat and it's kind of a bummer we didn't get to see anything unique with the "carry through the scores from the first phase" rule.

It's funny I feel like it's the opposite...the people who like the idea give their reasons why, and those who don't just say "ehh I don't like it the format so much".

Of course it's probably not best to beg people to give all of the reasons why they don't like a format, but I also find it so unbelievably bizarre that on the Internet of all places, where people complain with such ferocity and depth, that all the detractors have really come up with so far is "not a fan of the system".

It's all a matter of priorities, and to me one of the absolute worst ways to end a sporting competition, is to have someone who is in the hunt for the Championship, play someone who isn't (This will happen a ton in RoundRobins, and reasonably office in small Swisses). Not only is this extremely anticlimactic, but opens the door to huge collusion, and even if no actual collusion ever takes place, it opens the door to suspicion, paranoia etc. etc. I cannot get behind any format where one player may be playing for $30,000, while his opponent is playing for a sum much lower. This will likely happen in all non-knockout formats, but it's by far worse when it might happen or give the appearance of happening in a game that decides the Championship.

So those are my reasons for disliking the RoundRobin format, and also my reasons for disliking the typical Swiss for a small field. I understand that not everything is perfect with this year's US Championship system, but to me any faults are minor, and they are by far superior to having two players play with such a massive inequity in the prizes that they are playing for. I'm not saying that is impossible to occur in this system, but it is much less likely.

Also don't forget just 3 years ago Friedel (who I think was 1.5 points behind Shulman going into the final round) gives Shulman a quick draw with white to clinch a GM norm, automatically giving Shulman the title, and robbing all of those who were a point out of any chance. Seeing a U.S. Championship end like that just makes me sick, and the format was what made it possible.

Wow my spell checker is on crack....office of course = often.....Iphone is not very smart sometimes.

I didn't literally mean there was a drawing of lots for colors. I was trying to imply they "came out on top" for the final, but my scare quotes weren't scary enough. But I suppose they would have needed a drawing of lots had there not been a convenient distribution among the four. What else to do?

I like the format very much - the leaders play each other for the championship, so the final rounds are exciting.

Playing a straight Swiss the whole way means the leaders play the also-rans in the final rounds; how dull.

Anything but that debacle with Freidel last year, and his debacle three years ago is fine. Good job of them making this tournament debacle proof.

I like the format in St. Louis. The winner of a Swiss often depends too much on luck of the draw. Having all the top players duke it out against each other in the final rounds is great. If the last rounds were still Swiss, the tops would be playing middles because they've already all (almost) played each other earlier, and we get back to "luck of the draw" (or winner is the player who best beats up on the middle).

Curious, but why is Friedel not in the event this time ?

Bizarre is S. Polgar's strong negativity about the format. She goes on and on and on about how terrible the format is and really doesn't have any compelling reason for her whining--"confusing tie-breaks" is about the extent of it. Mucho criticism for an event that is very well run and with high standards of professionalism, yet remarkably quiet (or even pro-Moscow) about the shenanigans in Russia.

I'm impressed by the commentary from Maurice Ashley and Jennifer Shahade. They work quite well together. Their commentary has increased my understanding and enjoyment of these games. This is by far the best internet presentation of a chess event that I've seen. Big fan here.

ciaj, don't believe everything that you read someone say on the Internet.

Wow I really enjoyed the live video today; both commentators were quite entertaining and it was fun to hear Nakamura and Kamsky analyze afterwards. My favorite part was when Maurice Ashley said that Kamsky's analysis was "crispy". Despite what the commentators said, I didn't think that Shulman was trash-talking when he pointed out that Nakamura almost lost today. I thought Shulman's point was that if the other finalists are just trying to draw with each other and beat him, then why did Nakamura risk losing today?

Greg,

I think you should consider the criticism of the format without reasons symptomatic of the usual resistance to change for the sake of it. Every system of choosing winners and losers has strengths and weaknesses and this system is a decent way of solving some of the problems with a usual Swiss. Changing how winners are selected often seems unfair just because it is a change.

Okay, some points as to why I don't like this format:

We need to figure out what axioms we are going to use in the US Championship. If our axiom was "We simply want to find the best player in the country", then this would be easy enough. It sounds like we have a lot of other things in play, such as making sure people can make norms and the like.

What I don't like about a quad is that your last round pairings can still be a bit "weird", in that one player might be out of it facing someone who is still in it, and so on and so forth.

Also, the cut line feels a bit random. This year nothing bad happened, but in principle, being a half point back shouldn't be a massive tragedy after 7 rounds in a normal event. In this case, it could have meant everything.

Keeping scores from the previous round could also do some damage - in this case everyone entered level, imagine if Hikaru got some win because Christiansen had to go bonkers and try to win the game and gets a huge edge? Somehow that doesn't seem so equitable.

Color distribution in the quad also feels iffy - we're talking about 5 blacks vs. 4 whites, now make it 2 blacks vs. 1 white? Eeek.

Frankly, you don't get a good method for determining an overall winner without some type of KO (real match) at the end. That has some logic and PR appeal.

Want to make it a true challenge? 9 round swiss to determine a challenger to last year's US Champ. Then you get everything.

I admire the folks for trying, and it looks like this format worked so far (unless they just go ==, == over the next 2 days at which point weirdness sets in), and either way the US Champion won't be someone random, so it's all good.

Just feels like Christiansen said it best, for the other 20, they're really just going through the motions. The possibility that someone is going to make a norm (when you have only 3 non-GMs in a field of 24?!) isn't really *that* important to keep having everyone there.

Ok get ready for the longest post ever....

LarryC never said that everyone is just going through the motions. Do you think Shabalov and Stripunsky aren't going to go buck wild trying to win 10k tomorrow, which is practically the highest prize fund of ALMOST any chess event in the country? Do you think that in a typical year someone with 3/6 with 3 rounds to go somehow plays with more energy because they think they can somehow win the tournament (when they obviously can't), whereas in this year they will just shrug their shoulders and say "oh well"?

Also of course the quad could end up with someone out of it playing someone who is in the running in the final round. The chances are much lower than the typical Swiss of course. Why point out a flaw that will happen rarely, when it happens relatively often in the normal Swisses that have been happening? And in a round robin it's completely commonplace and normal. You act like it's a detriment that it might happen occasionally when instead it would happen often!

Last year Nakamura played someone in the final round with zero chance to win the title, and the previous year Shulman played someone in the final round who had zero chance to win the title. I don't understand the logic of how it's a problem that it might rarely/occasionally happen in this format. (and I think it'll be pretty rare...because please note that if two people are ahead of the field, they would automatically be slated to play in round 10, but since they were even this year there was no such priority).

In regards to the cut line being random....you have to set it somewhere. I mean doesn't it suck after a 9 round Swiss that all the people just a half point out don't have a chance to win? Can they complain that the tournament should be 1 more round in order to be fair to them?

Also in regards to LarryC going bonkers trying to beat Hikaru.....is it so bad to see people trying so hard to win? This is like in the 9th round of a Swiss, one player has 6 points playing a 5.5, and the other board is 2 6's, and then the two 6's draw and yet when the 5.5 goes nuts trying to beat the 6 and ends up losing, you claim the result is inequitable. I mean you are making an argument that can happen in literally every possible format aside from a knockout. It has nothing to do with our format, and in this case it doesn't even immediately determine the champion at the end of the event, it simply just gives them an extra half point, so it's much less annoying.

Also if the players go == == over the next two days who cares. That is completely results oriented thinking and means absolutely nothing. It could happen, just like there could very easily be a 3-6 way tie at the end of a regular Swiss (or Round Robin of course). Sure it hasn't happened the last few years, but to act as though it's not a very distinct possibility, is just results oriented madness. I mean do you not realize that we were one draw away from like a 8 way tie for first in 2003?

I appreciate that you've taken the time to write out a bunch of reasons for your objections to this format, but I find a lot of them to be lacking greatly in logical consistency, at least when compared with the typical format used to determine the US Champion in the past 10 years or so. I'm just sick and tired of a format where none of the leaders play head to head at the end of the tournament. I am a chess fan and just don't find such a thing to be interesting or accessible and the possibilities for collusion, corruption and suspicion are endless. When you see that person who is behind 1.5 points against their close friend, and they just happen to lose.....the questions in everyone's mind will be endless, regardless of what actually happened. This type of scenario is far more likely in a system other than this one (although of course the best to curb such possibilities is a pure knockout, which has many problems as well)

Also while I think the idea of a match versus the U.S. Champion makes sense, I would be strongly opposed to doing this without holding a tournament to determine the U.S. Champion first. Whoever just happens to be the U.S. Champion when such a system begins should not just automatically get this ridiculously sweet deal (and we all know that if someone who was considered a less-deserving champion was holding the title, that this idea would never be implemented at that time, the reason it sounds fun now is because our champion is Hikaru Nakamura).

However in principle I'm not at all opposed to such a format with a match vs the champion (although I'd still prefer this year's format to determine the challenger over a Swiss by leaps and bounds). I do admit however that thinking about the idea makes me like it a lot, and I could totally get behind that and pushing it starting with next year's U.S. Championship winner.

The only design that I can think of offhand, that might address some of the deficiencies of this one, would be a double-elimination knockout. I don't like the fact that in a single-elimination knockout it is so easy to be eliminated, and I don't like that the field is so stratified - people in one quarter of the bracket have almost no chance of playing people in another quarter of the bracket. So a double-elimination knockout mixes people around more as they fall into the losers' bracket, and there is a lot more chess played since you get to keep going even after losing one match.

The knockout format has the big advantage over Swiss and round-robin, as Greg says, that there is no possible hint of collusion, and every game matters toward tournament victory. I think this is a big deal to the players, and to the fans, and to people who hate draws. There is no arbitrary cut-off line or tiebreak rules needed; it is decided cleanly over the board, with the rapid/blitz tiebreaks. However, there is the huge question of whether each round is one day long, two days long, or three days long (i.e. when do you play the tiebreaks?) and whether it's okay to decide something of this magnitude with rapid/blitz games. I don't know what's worse on the players; having to stretch the event out to many days so individual days aren't so bad, or compressing the games into high-stress days. I think that maybe it would have to be a 16-player event rather than 32-player. The math works out well if it's a power of two, as I have written elsewhere - see
http://www.chess.co.uk/twic/sonas010704.html

Would get pretty brutal in the losers' bracket, since starting on Day 5 the elimination matches only last one day instead of two days, but you could get the whole thing done in 11 days:

Day 1: 8 first round matches (game 1)
Day 2: 8 first round matches (game 2 + tiebreaks)
Day 3: 4 winners bracket matches (game 1), 8 elimination matches (game 1)
Day 4: 4 winners bracket matches (game 2 + tiebreaks), 8 elimination matches (game 2 + tiebreaks)
Day 5: 2 winners bracket matches (game 1), 4 elimination matches (game 1, 2 + tiebreaks)
Day 6: 2 winners bracket matches (game 2 + tiebreaks), 2 elimination matches (game 1, 2 + tiebreaks)
Day 7: 1 winners bracket match (game 1), 2 elimination matches (game 1, 2 + tiebreaks)
Day 8: 1 winners bracket match (game 2 + tiebreaks), 1 elimination match (game 1, 2 + tiebreaks)
Day 9: (bye for unbeaten player), 1 elimination match (game 1, 2 + tiebreaks)
Day 10: Final match (game 1)
Day 11: Final match (game 2, unbeaten player has draw odds)

Seems pretty interesting Jeff....I definitely like that you have to lose twice. Verry interesting, although I'm sure people won't love playing two games in one day, but just speed up the time control or whatever, and at least it means that you aren't out the first time you lose. This idea definitely deserves a further look IMO.

However please note that I also really like the current format :)

Btw there are a few problems with the format as I understand it. Also please note that while I like it, it does not mean that I somehow don't like the current format, I just find the idea greatly superior to the typical single elimination knockout. I do think the format being used this year, with the time constraints we are under, is quite a good one.

Problems:

1. Have to invite a smaller field
2. Have to play a bunch of games at a rapid time control (no way anyone is playing 2 games of standard time control in a day in the U.S. Champs)
3. If the unbeaten player wins game 1 in the final match, the match is over.

I am still waiting for my drawless form of chess to come to fruition (as soon as game ends you reverse colors and keep playing with time you ended the previous game with). Ok lots of big problems with this though, but I like the idea :)

My ideal however is to continue using the format from this year (with perhaps some small modifications of course), but starting next year that whoever wins the Championship, then the winner of the following year's Championship will play a title match against them and that this method will continue on forever.

The real constraint is how to keep it within a reasonable number of days. You could increase it from 16 players to 32 players... if you can add two more days to the schedule. You could avoid having so many games in one day... if you can play tiebreaks on a separate day. As I originally designed the format, the winners' bracket matches would be four days long and the compressed losers' bracket matches would be two days long (not one)... but that makes the tournament a lot longer. It may just be too hard to compress it while still being inclusive enough.

Of course if you didn't have to have tiebreak games and people could just move on to different levels of bracket based on their score... +2 players face other +2 players, +1 players face other +1 players, and so on. What do you get? A Swiss, really, except it's a Swiss where you play the same opponent in two straight rounds, alternating colors, except as soon as you have -2 or worse after an even-numbered round, you're out of the tournament. I wonder if anyone's ever tried that twist! Probably everyone would be too conservative... I don't think it would work.

Yeah, didn't think of the tournament ending after game 1 of the finals, although that's what normally happens in the final stage of a double-elimination event, where the once-beaten survivor has to win two in a row from the unbeaten player. Maybe you make it a three-game final match where the unbeaten player gets White twice... kind of a stretch! Or you have a sudden death kind of thing where the first decisive game ends it, and the unbeaten player gets White first.

I like the fact that in the current format, you get the top players facing a mix of other top players, with the bonus of lots of rematches at the top. You wouldn't get that in the long match - the defending champ would just face the same person over and over again. Honestly I agree that the current format is nice!

The current format may have benefits, lower probabilities of problematic constellations, but it's unusual and complicated. A normal Swiss is complicated enough, but at least we are used to its pecularities. A Swiss with a mini-round-robin has all those pecularities, too, and a few new ones on top of that.
The theoretically best way to determine a championship is all-play-all. But in reality collusion and other non-fought games are a problem, so use a match system.
Like John Fernandez said: Use any way to determine the challenger, and let him play a match against the defending champion. A clash of personalities, a great show. Make sure the title (money) is attractive enough to keep the defending champion interested.

How about 4 six-player RRs:

Group 1: players 1, 8, 9, 16, 17, 24

Group 2: 2, 7, 10, 15, 18, 23

Group 3: 3, 6, 11, 14, 19, 22

Group 4: 4, 5, 12, 13, 20, 21

The top two in each section advance to the knockouts. The highest scoring winner of the groups gets to pick his opponent for the first knockout round, second highest scorer gets to pick from the remaining five opponents, etc. In the next round the "most decisive" winner gets to pick first. Instant animosity! ;-)

As to the format of the US championship,

The first question to answer: Does the defending champion have any special priviledge? If so, what kind of priviledge.

Should the championship be a match between the current champion and the #1 contender?

Should the championship be a closed tournament with the champion (and maybe a select few) automatically included?

Should the championship be an open format where anyone can enter?

I favor a blend of the open and closed format, though it is probably too cumbersome for many.

Round 1 - US Open tournament, anybody can enter, top 24 advance along with 8 automatic entries (the 8 autos could be top 2 from last year's closed tourny, reigning woman's, senior's, junior's, and maybe top 3 by rating). All of the autos could play and cash in the open but they would not count towards the qualifiers and the top 24 not counting the autos would advance. Tiebreaks would be needed but since you're deciding the bottom of the field any tiebreak could be used. Time frame = 9 days (assuming 9 rounds with rebuys and accelerated rounds).

Round 2 - US Invitational tournament, field of 32 (8 autos, 24 qualifiers), at least 5 rounds, maybe more, top 4 advance. Tiebreak games might be needed. Time frame = 1 week.

Round 3 - US Closed Championship. 4 players double round robin. Time frame = 1 week.

It would be nice if serious money was up for grabs at each level. You would think a corporate sponsor could be found for the final 4.

However in principle I'm not at all opposed to such a format with a match vs the champion (although I'd still prefer this year's format to determine the challenger over a Swiss by leaps and bounds). I do admit however that thinking about the idea makes me like it a lot, and I could totally get behind that and pushing it starting with next year's U.S. Championship winner.

***
The irony of course is that FIDE went down this road (challenger tourney with finals match vs champion) in 1998 (Anand v Karpov)

Everyone bitched and moaned -- for two reasons 1) for logistical reasons, they had to play the finals close to the KO, so everyone thought that was unfair - instead of taking the long view of "let's get this format started and then over years it will work out" and 2) people hated Karpov

Now, various formats sound great when the champ is someone people like and then when it is someone they don't like they start tinkering with formats

Pick a format and stick with it please!

Nakamura made a horrible blunder against Shulman - 23. Rh2?? His position is totally lost now...

btw I changed my mind, a match vs the champion doesn't appeal to me that much anymore at all. Totally destroys the prestige of the US Championship (or US Championship match qualifier or whatever you'd want to call it in the future) and will make sponsors much less interested in ever running it.

It sounds good in theory, but there are a host of problems that will come with it (such as when the champion becomes too busy to play their scheduled match, which could happen a good amount of the time). Just can imagine a ton of nightmares unfolding from such an idea. Meanwhile when running a tourney, the world doesn't end if one player is suddenly too busy to play, or whatever.

Yes , very easy to change your mind, you are soooo great Mr IM Genious, LMAO.