On Mon, 16 Oct 2006, Don Dailey wrote:
What I have decided on, based on several different simulations, is the
following:
1. Fix N at 3
2. Pick one unpaired player P at random.
3. Loop 3 times doing the following:
4. Select a potential unpaired opponent R as a candidate opponent
for P, randomly. Opponent R must not also be P.
5. Compare the rating difference of the current selection to the
best selection found so far.
6. If the difference is smaller, R is new best candidate.
7. Repeat whole procedure until all players are paired.
I like it ... it's simple.
The pairing-probability table (NP=14) looks like:
2 3 4 5 6 7 8 9 10 11 12 13 14
01: 3.5 2.2 1.6 1.2 0.9 0.7 0.5 0.5 0.4 0.4 0.4 0.3 0.4
02: 2.5 1.7 1.2 0.9 0.7 0.5 0.4 0.4 0.3 0.3 0.3 0.3
03: 2.3 1.5 1.1 0.8 0.6 0.4 0.4 0.3 0.3 0.3 0.3
04: 2.1 1.4 0.9 0.7 0.5 0.4 0.3 0.3 0.3 0.4
05: 2.0 1.4 1.0 0.6 0.4 0.4 0.4 0.4 0.4
06: 1.9 1.4 0.9 0.6 0.5 0.4 0.4 0.5
07: 1.9 1.4 1.0 0.7 0.6 0.5 0.5
08: 2.0 1.4 1.0 0.8 0.7 0.7
09: 2.0 1.4 1.0 0.9 0.9
10: 2.1 1.5 1.2 1.2
11: 2.3 1.7 1.6
12: 2.5 2.2
13: 3.5
Christoph
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
