SuperNationals III & Accelerated Pairing

Running Chess Tournaments
1

I apologized, not sure which topic to post this one.

Someone told me that the SuperNationals III will use Accelerated Pairing.
Is this true?

Thanks.

2

I didn’t see anything specific in uschess.org/scholastic/schol … ns.php#170
but considering the large number of players vs the low number of rounds, I would think Accelerated Pairings would be used in the first few rounds. Don’t take this as gold, though: Al Losoff may have other ideas.

3

According to Al Losoff, accelerated pairings will be used in all the championship sections except the K-1. Accelerated pairings will not be used in the reserve sections, because Al says they don’t work in a reserve event.

4

Someone expand on this please: why WOULDN’T accelerated pairings work in a reserve section?

5

I am not trying to speak for Al, but can give my best guess as to why he might have made that statement.

The goal of accelerated pairings is to reduce the number of perfect scores.

Standard pairings have at most roughly 25% of the players at 2-0 after two rounds (actually 25% of the number that is the number of players rounded up to a multiple of 4). Accelerated pairings with the no upsets can have no more than roughly 12.5% of the players at 2-0, while accelerated pairings with many upsets can have up to roughly 37.5% at 2-0.
Thus accelerated pairings work well when the ratings are solid enough, stable enough, and far enough apart that there is a good expectation that the top half players with round one losses will not lose against bottom half players with round one wins. If that expectation is not at least 50%, and preferably much higher, then using accelerated pairings is questionable and risks having more perfect scores than standard pairings.

A partial list of when accelerated pairings have problems is:

  1. there is a large number of unrated players that might be able to win in the lower half round one pairings and again in round two against top half players that lost round one.
  2. it is a class tournament where the range of the classes is fairly small (often when <300 rating points and definitely when <200 rating points).
  3. the ratings are volatile enough that it is uncertain whether or not most of the players in the bottom half by rating are really in the bottom half by strength. Over the decades I’ve seen a greater chance of this occuring with scholastic players at the lower rating levels, as players can often significantly increase their actual playing strength by learning just a few techniques while their rating can take a while to catch up to their new strength.
  4. the number of players is no greater than 2 raised to a power equal to the number of rounds (128 for a 7-round tournament), and a case against accelerated pairings can still be ventured at 1.5 and even 2 times that number.

1 and 3 above are noticeably higher risks in the reserve sections. In addition to 1 and 3, 4 is an additional risk in K-1, though that risk of number 4 is likely to decline as time goes on.