(Knockouts)?
I directed one for Mikhail Korenman once. What is the question?
Alex Relyea
The 1990 US Championships, held concurrently with the 1990 US Open in Jacksonville FL) used a knockout format, but I don’t recall the details.
My slightly questionable 20-year-old memory is that there were 16 players, two-game matches during the first three rounds, and a four-game match at the end. Players that were knocked out could be transferred to the US Open with a bye-point score equal to the score they attained in the closed (obviously less than perfect because otherwise they wouldn’t have been knocked out). Alternatively, players could opt to play in both events so that they had a better score in the Open if they were knocked out of the Closed. I’m not sure if that was the event where a player in the Closed came to the TD room in the open asking why, in an early round, he would be paired with two blacks in a row (he’d overlooked that it was one black in the open and one in the closed).
As I recall, some of the players in the US Championship played in the US Open right from the start.
I can fill in some of the details from the last post.
My biggest memory of that event (aside from losing the PB race to John Donaldson) was when GM Dzindzichashvili came into the TD room to protest his pairing for one round of the US Open, claiming that he had gotten two blacks in a row.
Ira Riddle (the chief TD for both events, I believe) looked at his pairing card and told him that no, he had received White in yesterday’s game and Black in today’s game.
Roman responded: No, I received Black in this afternoon’s game, and Black again in tonight’s game.
(The afternoon game was in the US Championship.)
I set up a knockout tournament as an experiment for a club during the 1990’s. We modeled it on the NCAA basketball brackets, but drew lots for pairing #'s for the first round. That meant it would be possible for the first and second highest rated player to meet in round one. We had 28 entries, so #1 - 4 on the bracket received round 1 byes. It ran okay, but after two rounds, 3/4 of the players are knocked out and you have to find something else for them to do. Chess players are poor spectators; too fidgety, they want to play. We ran a quick chess G15 tournament for them and dropped in losers from the knockout into the tournament with an adjusted score. We had a prize for the winner of the quick tournament, and prizes for all who reached the 1/4 final stage and beyond.
Some of the players weren’t too keen on running another such event and preferred a swiss, but it may have been the novelty of the tournament that affected their opinion. With more experience they may have liked it better. BTW, there were few upsets; the two highest rated met in the semifinals. The top rated defeated the third highest rated in the finals. All of the players were under 1400 USCF. After the event the club decided not to send it in for rating at the request of the players.
A knockout makes sense for games/sports with a lot of spectator interest, limited facilities, and/or many officials needed for each match(up). At (almost) any point in a tennis tournament, if a given player wins the next four points and 55% of the remaining points after that, he/she can win the tournament. In chess we have plenty of equipment, one TD can handle quite a few games at one time, and there is practically no spectator interest. There’s no reason to kick anyone out for losing.
The first international chess tournament ever held, London 1851, was a knockout event.
Adolph Anderssen, generally considered the strongest player in the field, won, though there were some concerns that the random pairings used may have matched some of the stronger players against each other in the first round.
Chess as a double-elimination seeded event might make it more marketable.
You can’t have big open tournaments with elimination or double elimination, because no one will make a trip to one of them when they will probably be eliminated in an early round. That’s a serious strike against it.
Sadly, there’s not much that can make chess marketable in the way football or poker is, because you have to be good at it – very, very good compared to most people who play – to even start to appreciate GM-level play. Speeding up the game, to try to satisfy short attention spans, makes it even more difficult to follow.
Or maybe I’m wrong. Go and shogi have professional leagues and a reasonably big following in Japan (China and South Korea as well for go) and have the same obstacles to popularity that chess has. But you have to have that popularity in the first place to even think about being marketable – marketability requires a market.
Single elimination doesn’t seem to keep the big names out of professional tennis competitions, so my guess is that particular objection can be overcome with money, though it would probably need something other than a one game per round format.
I agree that speeding up the game can be overdone, but the time the game takes is something that can be overcome in post production with good editing, they don’t show every hand in most televised poker tournaments, after all, and they’re seldom shown live.
You’re answering an argument I didn’t make. Big open tournaments were what I referred to. Of course you can have professional competitions with knockout systems. They do that in shogi and go as well, I believe. But you wouldn’t have hundreds of people playing in, say, the World Open if half of them were going to be out after round one (or even a quarter out after round two, as with double elimination.) Then you have no more World Open, Chicago Open, U.S. Open, National Chess Congress, etc., as we know them. And I don’t see what you’re thinking about when you suggest that this would somehow make chess more marketable.
Or suppose you were just talking about top-level events, which now that I think about it you probably were. You could make them knockout events, and some have been, but I still don’t see why you think this would increase marketability. I don’t know of any empirical evidence for it in the history of (rare) knockout events. Who cared about FIDE’s fake world championship? (Though that may have had to do with it being a clearly fake world championship.)
I have trouble thinking of the World Open as a ‘big open’ tournament, when the Open Section has less than 10% of the participants. (98 out of 1222 in 2009.)
I do agree that a knockout format doesn’t work for large tournaments and is probably best suited for invitational events.
Note that after the 1851 event that the knockout was criticised and the chess world turned to round robin tournaments for the professionals and and amateurs to play. That format remained the standard for many decades until the Swiss System was developed to accomodate large numbers of amateur enthusiasts when large roubin events became too expensive and unwieldy to run. Look on the Swiss System as a compromise to popularise chess. In the end it is all about the expense.
I’ve always read that the main criticism of the knockout format used for London 1851 was that they did random pairings for the first round, so some of the stronger players met in the first round.
Seeded pairings are the norm in professional tennis tournaments, and they wouldn’t be hard to do with USCF or FIDE ratings,
In that era, level of strength was base on reputation rather than on any concrete measure. Many of the players had big egos (a term to be coined decades later) and thought they were better than their peers. The tournament was a way to determine who really was the best at that time. I am not sure how the committee that organized the 1851 event chose seedings; perhaps they drew lots as a fair method. NCAA bracketologists would likely be stunned at their naive methodology.
I haven’t looked at a copy of the book on the 1851 tournament in a while, but as I recall there was a ceremonial drawing of pairings with the tournament backer taking part.
I hesitate to call 1851 London a ‘knockout event’, at least compared to the recent World Cups.
The first round (round of 16) was played first to 2 wins, but the quarter-finals and subsequent rounds were played first to 4 wins. It seems more similar to the candidates matches that used to be held to determine the challenger to the world champion (1965-1993) then the ‘one loss and you’re in DEEP trouble’ knock-out.
Pairings for the candidate matches were not completely random, at least not until the 1984 cycle.
In the 1966, 1969, and 1972 cycles, the top 2 players from the Interzonal, plus the two qualifiers from the previous cycle were guaranteed not to play each other in the quarter-finals. This gave the top player in the interzonal a (relatively) easy pairing in the quarters, probably why Bent Larsen was able to break through to the semi-finals three times for he finished 1st or 2nd in each interzonal. In cycles 1975, 1978, and 1981 there were two interzonals rather than one so the ‘top half’ was changed to the two qualifiers from the previous cycle and each interzonal winner.
Again I ask Mr. Lawson what his interest is here. It seems that members of this forum have a lot of experience with these events, and could possibly be helpful, but he hasn’t posted after his original post.
Alex Relyea
While Knockout format would not work for big open events, as people would not make a trip out of town with a large chance of being eliminated in the first round it might be a valid system for a weeknight events in local chess clubs. Say you have a 16 people, 4 round single elimination G/30 tournament (draws result in 1 game G/5 or G/3 tiebreak). If you are not playing well, and get eliminated early its an extra advantage of getting a solid night’s sleep!
Bullfeathers. If you are not playing well and want to go home and sleep, you can always drop out of a Swiss (just tell the TD in time). There is absolutely no advantage in forcing a player out.
Bill Smythe