For the vast majority of participants, the number of pairs doesn't really matter... it only affects those right at the top who may feel disadvantaged if they are the only undefeated team but still have to play out 1 or 2 more rounds. With that in mind, my preference would be a minimum of 2^(rounds-1) and preferably 2^rounds. So with 7 rounds, that equates to a minimum of 64 and preferably at least 128; For the more common county 6 round matches, you would be looking at 32/64, and for club swiss pairs evenings of 5x5 then 16/32 doesn't seem unreasonable. Of course, many EBU events have lots of additional rounds to these figures, but those competitions focus more on consistency (compared to a 7 match swiss pairs where the "last pair standing" would expect to win at around that point).
I haven't found any official recommendations but would be interested to find out if anyone knows...
I was told many moons ago that a swiss should have two more rounds than that suggested by log2(n). otherwise you may as well play a knockout and consolation. The point being that the best players get to play more of the other good players to prove their worth.
The problem with that is the risk of over-swissing (as I'm sure the thread starter was referring to). With only 32 pairs in a 7-round event, that could definitely lead to problems (though it would depend on the exact results on the day).
With reference to your point about "proving their worth," the benefit of using log2(n) is that the leading pair does, in effect, play a knock-out, and will roughly play the 2nd best pair in the last round. If you add on extra rounds beyond that, they will then play the 3rd, then 4th, then 5th best pairs and may well win those 12-8 or 13-7; That could lead to pairs further down the ranking list making a late surge against easier opposition, while the leading pair had no opportunity to do so. I would have to try to create an example, but that would be my intuitive thought on that.
I may be biased here admittedly - I would love to see more (or indeed any) events with a knockout-style competition (probably with a repechage) and losers joining into a swiss pairs event - with a normal swiss pairs, you'd need to win each match by around 15-5 to be victorious, which often requires a bit more fortune than securing at least a narrow win in each match.
The converse also applies- if you lose one match heavily (no matter how - no doubt it was your partner or teammates) then you don't have any chance to recover - so what's the point of playing on - other than for the 0.25 greens?
My rule of thumb for Swiss Pairs is not to play more rounds than a third of the number of possible opponents. So for 7 rounds, I would want 11 (full) tables.
Comments
For the vast majority of participants, the number of pairs doesn't really matter... it only affects those right at the top who may feel disadvantaged if they are the only undefeated team but still have to play out 1 or 2 more rounds. With that in mind, my preference would be a minimum of 2^(rounds-1) and preferably 2^rounds. So with 7 rounds, that equates to a minimum of 64 and preferably at least 128; For the more common county 6 round matches, you would be looking at 32/64, and for club swiss pairs evenings of 5x5 then 16/32 doesn't seem unreasonable. Of course, many EBU events have lots of additional rounds to these figures, but those competitions focus more on consistency (compared to a 7 match swiss pairs where the "last pair standing" would expect to win at around that point).
I haven't found any official recommendations but would be interested to find out if anyone knows...
I was told many moons ago that a swiss should have two more rounds than that suggested by log2(n). otherwise you may as well play a knockout and consolation. The point being that the best players get to play more of the other good players to prove their worth.
The problem with that is the risk of over-swissing (as I'm sure the thread starter was referring to). With only 32 pairs in a 7-round event, that could definitely lead to problems (though it would depend on the exact results on the day).
With reference to your point about "proving their worth," the benefit of using log2(n) is that the leading pair does, in effect, play a knock-out, and will roughly play the 2nd best pair in the last round. If you add on extra rounds beyond that, they will then play the 3rd, then 4th, then 5th best pairs and may well win those 12-8 or 13-7; That could lead to pairs further down the ranking list making a late surge against easier opposition, while the leading pair had no opportunity to do so. I would have to try to create an example, but that would be my intuitive thought on that.
I may be biased here admittedly - I would love to see more (or indeed any) events with a knockout-style competition (probably with a repechage) and losers joining into a swiss pairs event - with a normal swiss pairs, you'd need to win each match by around 15-5 to be victorious, which often requires a bit more fortune than securing at least a narrow win in each match.
The converse also applies- if you lose one match heavily (no matter how - no doubt it was your partner or teammates) then you don't have any chance to recover - so what's the point of playing on - other than for the 0.25 greens?
My rule of thumb for Swiss Pairs is not to play more rounds than a third of the number of possible opponents. So for 7 rounds, I would want 11 (full) tables.