THE GHU FORMAT: A NEW PLAY OF PLAYING HEADS-UP
16:29, October 1st, 2011Currently, the Swedish Online Championships of Poker are underway at Unibet; four events played so far, with Omaha to follow this evening.
Now I am particularly fond of heads-up poker; as writer Bob Ciaffone once put it, ”poker in its purest form of all”. Once I get heads-up in a sit-and-go, the records I keep show that I will win it twice as often as I will finish second.
What happened day before yesterday was however that I had, as per my contract with Unibet, been given a ”ticket” to the HU Online Championship… which I took to be ”a reserved seat”. When I sat down before my computer with half an hour to go before the tournament, I duly exchanged my ticket for spot number 150 or so; and then, when the tournament started, no table popped up. Confusion, and a telephone call; whereupon I was informed that no, the ”ticket” was not equal to a reserved seat, and that only the 128 first to register (a total of 179 had registered by 8 p.m.) were allowed into the tournament. Had the number of registrations reached 256 (= the next level) or above, then 256 would have taken part.
Understandable in a way, as the players in the current HU format knock each other out according to the reverse progression 256>128>64>32>16>8>4>2>1… but it led me to thinking that it must be possible to arrange a heads-up tournament in such a format that it can accept any number of players.
Also, there would be advantages to this. With the limit set at 128 players in the HU Online Swedish Championship, there was a total of 3712 euros going to the victor; had all 179 registered players been allowed in, first prize would have been over 5000 euros! And, as I said, all registered players would have taken part, causing no disappointment – as it did now – to those 51 who had their money refunded while staring dumbly at their screens.
Now here is my suggestion for a new HU format, for future tournaments:
The basis of my format is the wellknown ”Swiss system”, which since 1895 has been used not only in chess tournaments throughout the world, but also in checkers, oware, Othello and dozens of other games where the players by necessity face off two at a time.
In short, adapted to HU poker for any number of players (but a large starting field), the software (or the tournament director and her/his crew, when playing live) would randomly choose which opponent you would face in the first round. If there is an odd number of entrants, the player who receives a ”bye” (= does not play an opponent in that round) is also chosen randomly.
After this first round of play, half of the players will have won their match (win = 1 point, loss = 0 points), while half of them will have lost. Now the Swiss system kicks in: if you won your first match, you will in your second match face another player who also won his/her first match; and if you lost your first match, you will instead face another ”loser” in the second round.
After the second round, ¼ of the field will have won two matches; ½ will have won one match; and ¼ will have lost two games in a row.
In the third and subsequent rounds, the same principle applies: you will as far as possible meet an opponent who has won the same number of matches that you have. If you have 2 points so far, you will meet an opponent who also has 2 points; if you have 1 point, you will meet an opponent who also has 1 point; etcetera.
And so it goes on, for another 4 or 5 or 6 or 7 rounds (depending on the number of entrants): players with a certain number of points so far will in the next round meet another player with the same number of points. In this way, the Swiss system works like an old-fashioned separator for milk and cream; you climb upwards or descend downwards until you ”have found your level”.
Four basic rules within the Swiss system are that (A) you may never meet the same opponent twice, but must instead as far as possible in each round meet a new opponent with the same number of points that you have; (B) a player who is given a ”bye” (when there is an odd number of entrants) must always be randomly chosen from the group with the lowest number of points so far; (C) that player who receives a ”bye” is given 1 point, quite in accordance with the Swiss system; and (D) a player can only receive one ”bye” in the entire tournament.
After let us say 7 or 8 or 9 matches in total (again depending on the number of entrants), the best players will have risen to the top points-wise. Now the best 16 players (or possibly 32, if the starting field is sufficiently large) are in the money and then play off for the respective prizes in the classical fashion: 16>8>4>2>1 winner.
What happens if you need a tiebreaker to select the 16, because there are several players with the same number of points? In chess and other games it can get a tad complicated here, as you start counting ”sub-points”; basically, if you defeat player A who in turn in his/her matches has defeated better opponents than player B has, then defeating player A is worth more ”sub-points” than defeating player B. While this can of course be calculated at lightning speed by the software, I instead suggest a much simpler and more crystal-clear rule for my format; that in case of equal points, the tiebreaker be the total accumulated time you took to defeat your opponents in the matches that you won (the ones you lost are ignored here). Thus, aggressive play will be rewarded.
The obvious question is of course, how many rounds of play are needed? The rule of thumb within chess etc is that to produce a clear winner with reasonable accuracy, you need the same number of rounds as if it had been a knockout tournament. Thus, with let’s say 193 entrants, you round up to the next multiple which is 256, and find that nine rounds are needed in order to produce a clear winner.
However, we do not need to produce a clear winner in my new HU format (call it the Glimne Heads-Up Format or GHU format for short); rather, we only need to produce the top 16 players, for which seven rounds should be sufficient here. Add a further four rounds to that for the final 16>8>4>2>1 progression, and you have a total of eleven rounds with such a large starting field as 193 in the example here. (Smaller starting fields will of course need fewer rounds.)
It should be pointed out that the proven Swiss system has been used in hundreds of thousands of tournaments since Julius Müller first suggested it and it was put to practice in 1895 in Zürich; and in the GHU format I have adapted it to ”select” those players which will wind up in the money, and then have them fight it out for the top spots – possibly with a ”seeding” built in, so that the players at the very top will start in separate halves of the field.
The advantages of the GHU format are several:
(A) It will adapt to any number of players, instead of stopping at 128 or 256 and shutting out disappointed entrants.
(B)The prize money will be bigger.
(C) Even if you lose a game or possibly two, you still have a chance of winding up in the top 16.
(D) It will be much better at actually selecting the best players, unlike the classic knockout system where all it takes against a worse opponent is a ridiculous two-outer on the river and you will be history – the luck factor is simply lower with the GHU format, since everyone gets to play several rounds.
(E) Each participant will get more play value for his or her money.
(F) The excitement and the hope will last longer, before the top 16 (or whatever number) are eventually chosen.
As for the disadvantages, there is pretty much only one: the GHU format requires another two or three rounds in total, compared to the knockout format. But is that not a reasonable price to pay for a format with more play value, more money (also for the arranging site) and less luck?
Over to you, dear readers…




