My club does a 1 section, 4 round Swiss, G/60,d5 tournament once a month. It says that the TD “may” change it to 5 round Swiss, G/45,d5 if there are more than 25 players (to avoid multiple perfect scores). I was wondering, if the tournament gets a big turnout (and what is the magic number here?), is it better to add a round and make it a 5 round Swiss, G/45,d5 or to do accelerated pairings for the first two rounds? Some might not like the change to the faster time control.
Whichever way you go you should include the possible change or acceleration in your pre-tournament publicity. Players may not like the faster time control, but they like even less unexpected changes. I know 1 player who showed up for a Quad tournament and left when it was converted to a swiss event.
It is always better if possible to inform the players ahead of time if possible. Yes I know that it is not necessary to announce the use of accelerated pairings in advance, but advance notice is always preferable.
Personally I would tend towards the acceleration, but depending on the numbers maybe only accelerate the first round. remember that the whole point of either the acceleration or extra round is just to try to make sure there is only 1 perfect score at the end of the tournament. I think players understand that a tie is always possible for any place or prize in an event.
Larry S. Cohen
Personally, I think accelerated pairings are a minor abomination. Sure, it’ll cut the number of perfect scores in half twice as fast, but the effect on the middle 3/4 of the tournament (everybody except the top 8th and bottom 8th) is almost like playing the rounds of a non-accelerated Swiss out of order. The third round has a similar feel to the 1st round in a standard event.
I also don’t like the idea of having the number of rounds (or the time control) depend on the turnout. Players like to know what to expect before they enter.
Better to just add a round, unconditionally, when planning the tournament. Or divide the tournament into two or more sections – also unconditionally.
Bill Smythe
Accelerating only the first round simply switches the first and second rounds around. The key to acceleration is the second round pairing with, essentially, top half losses paired against bottom half wins.
I agree with Mr. Smythe. I’ve never found it necessary to accelerate pairings, and had many sections with no perfect scores. It also tends to really mess with lower prizes.
Alex Relyea
Accelerated pairings are quite common in scholastics with sections having players numbering at least 1.5 * (2 raised to the power of the number of rounds). Since such scholastics often have team trophies involved, and a pre-determined number of individual trophies, it is problematic to split them further with little notice. Particularly in the lower grades there are not that many draws that would otherwise reduce the number of perfect scores.
Here is an example this year of an accelerated section (7 rounds, 307 ploayers).
main.uschess.org/msa/XtblMain.ph … 3-12816147
Why is there a concern about multiple perfect scores in the first place? Trophy? Club championship?
In an open section, you’re not likely to get two perfect scores with 25 players. In fact, you may not even get one.
Adding a round is almost always problematic in a tournament of 20+ players. That’s unfair to players who show up expecting the primary advertised conditions - and have scheduled their time accordingly. Also, if it is a Grand Prix event, one cannot change the advertised conditions at all.
Accelerating pairings for a field of 25 in a four-round tournament may not do all that much to cut down on perfect scores, depending on the rating spread. In a 25-player, 4-round event where the top two or three seeds are rated 300-400 points above everyone else, there’s a good chance two players will go 4-0.
If an organizer doesn’t want multiple perfect scores, and knows they typically draw around (1.5 * 2^n) players, where n = number of rounds, the organizer could just go to n+1 rounds. Unless there is some indivisible prize at stake, though, it is unclear at first blush why multiple perfect scores are problematic.
Absolutely!! At nearly every major event I run, the question of when should one expect it to end comes up numerous times. A very valid question, as folks are doing their best to
program chess into busy schedules. To then add additional rounds which they will not be
able to play in due to these schedules, would create an imposition for them. I would imagine that such a practice over a few events would spell mud for the organizers/tournament directors as to future event attendance.
Rob Jones
I don’t see that restriction in the Grand Prix Rules as published in Chess Life, for example, see uschess.org/images/stories/g … pr2014.pdf
However it is true that if the organizer increases the prizes that does not increase the number of Grand Prix points available. And the Grand Prix prize amounts upon which the Grand Prix points was determined cannot be reduced. (If prizes are based-on, only the guaranteed portion, ie, half of the prize amount, would be used for Grand Prix point determination.)
Mr. Nolan is correct that it isn’t against USCF GP rules to change (most) advertised conditions. I was thinking of possible cases where it was illegal. That said, it’s generally not a good idea to change GP event conditions unless absolutely necessary, as that may subject the event to ED review.
That’s true for any advertised event. Player consent to any changes goes a long way in both preventing and resolving complaints.
The reality is that if you (truthfully) advertise the event as 4@G/60 or 5@G/45 at the discretion of the TD, you’ll probably lose enough potential entrants that you don’t have to worry. There are people who will skip something quicker than G/60, but I’m not sure anyone has passed up a tournament solely because there might be more than one perfect score.
As long as we are talking about acceleration, one thing to remember is that it needs a broad, semi-smooth range of ratings to work well.
Quartile acceleration is:
Round 1
Top quarter versus second quarter, third quarter versus fourth quarter
Round 2 (somewhat oversimplified)
Top half 1-0 against each other
Top half 0-1 against bottom half 1-0
Bottom half 0-1 against each other
The logic behind acceleration is that the majority of the top half 0-1 players will beat their bottom half 1-0 opponents. Acceleration needs at least 50% of the 0-1 vs 1-0 games to not be won by the 1-0 players or it actually increases the number of perfect scores.
If the rating range is not broad (think sections of a class tournament) then you risk more than half of the bottom half 1-0 players winning their second round with the worst case scenario being a 50% increase of 2-0 scores over standard (non-accelerated) pairings. If the rating range is broad but not smooth then there may be a clumping of ratings in the middle that makes those 0-1 vs 1-0 pairings have players close in ratings (think of a 4-round 32 player tournament with 4 experts, 4 A players, 16 C players, 4 E players and 4 U1000 players).
If you do have a significant average rating difference between the second quarter and the third quarter and at least 1.5*2**(#of rounds) then acceleration is an option to reduce the number of perfect scores (useful when first place is a title or a trophy).
I’ve looked at results of accelerated tournaments and have done dry runs, and the average field faced in accelerated and standard pairings come out to be very similar for the broad middle of the field (the top of the field has better opponents when accelerating and the bottom of the field has weaker opponents). The round by round distribution of those opponents often comes across as strange.
Because of the pairing method, when accelerating is possible then the tournament’s tie-break system may have Sonnenborn-Berger replacing cumulative in the tie-break sequence.
P.S. I have had people ask for acceleration even when it is not needed to get a single perfect score. That request is usually because they think that acceleration will more quickly get to the final rounds where the pairings do not have the huge rating differences. Once I explain that accelerating doesn’t really change the overall strength of players that the broad middle faces, but rather kind of shifts them around, they then drop the request.
This is IMPORTANT! If you are going to accelerate pairings, you have do it for at least TWO rounds.
Cumulative ought to be OK with accelerated pairings, except that the accumulation should begin with round 2. In other words, a player’s cumulative tie-break in an accelerated tournament should be defined as the sum of the player’s scores after each round beginning with round 2.
How do you even define accelerated pairings in round 3? Remember that in round 1 you do quarter pairings, then in round 2 you take advantage of it. It seems the accelerated rounds must come in pairs. Something like this:
- Rounds 1 and 2: same as “normal” accelerated pairings.
- Round 3: 2-pointers vs each other. 1-pointers vs each other, by quarters. 0-pointers vs each other.
- Round 4: 3-pointers vs each other. Top half 2-pointers vs each other. Bottom half 2-pointers vs top half 1-pointers. Bottom half 1-pointers vs each other. 0-pointers vs each other.
Bill Smythe
Pair the bottom half of the 2-0’s with the highest rated 1.5’s and 1’s. You could pair everyone else normally. It would probably be seen as an act of desperation, but theoretically would work if the some high seeds lost or drew in the first two rounds.
Any TD who doesn’t understand this concept ought not to be trying accelerated pairings.
I think it would be better simply to break it into two sections. I would not want to come to a tournament only to find that it was going to be played at a different time control than the one I’d mentally prepared for, nor would I want to be made to either stay for an extra round or forfeit any crack at a prize. Such a “solution” is likely to “solve” the problem by driving away the excess players, not just in the current tournament but in future ones as well.
If you refuse to divide the section, then accelerated pairings are an adequate solution, far preferable to changing the time control and/or the number of rounds.
But it would be far better to play under the announced conditions and if there are multiple perfect scores, live with it. The problem may resolve itself anyway through draws.