Our club likes the ease of pairing events at night with Quads and Hexes; but sometimes the number of players does not fit that format, so I wondered if anyone had tried what I will call “Pseudo Swiss” pairings which might work this way:
The first round is paired the same as a true Swiss; the second round assumes that all games were won by the higher rated player; the third round assumes some upsets or draws; the fourth & final round then pairs those players that should have the same score play each player that they have not already played and tries to have equalization and alternation in so far as possible. The essential purpose is to avoid the downtime in pairing between rounds and use our limited 3 hour time as effectively as possible. We usually play Quick at G/15 d3. Comments please.
How many rounds are pre-paired? When are pairings done in the middle of the tournament? Are the results of the “upsets” considered? Are many players within < 200 points of the top player?
It might work, but I see many possibilites for some train wrecks.
All rounds are pre-paired. The situation occurs when we have more than 6 players (6 players gives us an easy way to have a Hex, but only play 4 rounds of it). We rarely have more than 10 players, so the Pseudo Swiss would seem to me to be no more random than a round robin. We could have several ways to pre-pair rounds 3 and 4 that might be reasonable.
How long does it take to pair a small section using a computer? I bet I could do it in less than a minute.
Alex Relyea
Yes, but I thought the point was that players would be able to start as soon as their opponent was free instead of waiting for all the games to be finished and the pairings done.
Even if you could do the pairings in zero time, it might slow things down a bit waiting for the last game in the round to finish before starting the next round.
If part of your issue is lack of a computer then one option is to simply use the RR tables for a RR-directed Swiss. That ensures that you don’t accidentally lock yourself into having to have two players have a rematch in round four.
5 or 6 players (sequence by rating)
Tournament Rd 1: table B (p295) round 2 (flip for colors on board one - if the top player gets black then use the RR table colors for all rounds and otherwise reverse them)
Tournament Rd 2: table B round 3 (if the higher-rated player won all games in round 1)
Tournament Rd 3: table B for whichever round (1, 4 or 5) works.
Tournament Rd 4: table B for whichever remaining round works.
Tournament Rd 5: last table B line left if it turns out you are going to have the time available after all.
7 or 8 players (sequence by rating)
1: table C (p 295) round 1 (flip for colors on board one to determine whether to use or reverse them in the table for the rounds used
2: table C round 4
3 and on: table C for whichever remaining round works.
We tried that this Tuesday night; the TD with the computer and software cancelled, so we did 2 sets of Quads.
Four rounds G/15 d2… Absolute max game time in the neighborhood of 34 minutes, four rounds of that = 2 hours 16 minutes. 44 minutes for pairings doesn’t sound bad to me. But I know that’s not the answer you’re looking for, and that of course assumes you’re not waiting 15-20 minutes for people to show up.
I would probably live enter all results at game finish and then do my best with adjudicated pairings for any games unfinished after 20-25 minutes (or maybe institute an ‘all remaining will be paired as draws’ rule.) Pair at 25 minutes and post, start next round on time but allow any late finishers 5 minutes restroom/etc. before starting their clock. But then again, I’m really trying to commit to not play while I direct anymore and would have the luxury of entering results as they finish.
I think with the system you propose any round 1 upsets will have an easier time, facing their 0-point opponent for round 2, a paried draw for round 3? You could get a horribly mismatched round 4. And I wouldn’t as a player be too thrilled at a cold-decked rounds 2 and 3 only to open it up in Round 4 again… But OTOH without knowing your local club environment, it might not be a bad arrangement either.
ETA: Wrote the above this morning before the above post… sigh…
How many players do you usually have (minimum and maximum) for the tournament?
We may have from 6 to 12 players, rated from 1950 to 850 strength. About half are USCF rated; several were, but are no longer. When we do Quads and Hexes we have some mismatched pairings which we find good for helping the weaker players’ learning experience.
How does someone become no longer rated?
Alex Relyea
A somewhat simpler version of the same idea might be as follows:
- Pair round 1 as usual and start the round.
- As round 1 is starting, immediately pair round 2, ignoring round 1 results (which aren’t in yet) but observing round 1 colors and round 1 already-played opponents.
- Immediately after round 2 has started (or, for that matter, any time after all round 1 results are in), pair round 3, ignoring round 2 results (which aren’t in yet) but observing round 1 results, rounds 1-2 colors, and rounds 1-2 already-played opponents.
- Any time after all round 2 results are in, pair round 4, ignoring round 3 results (which aren’t in yet) but observing rounds 1-2 results, rounds 1-3 colors, and rounds 1-3 already-played opponents.
- Continue in this way for the entire event.
This would probably work better for a large event than a small one. The idea is to pair each round based on information available at the start of the previous round (i.e. colors and opponents from all previous rounds, and results from all previous rounds except the immediately previous round).
That way, some games could start before all previous games finish (one of your goals), and there would be no downtime between rounds while pairings are being made (your other goal). And it’s simple – no assuming there will be “some” upsets, etc.
The 2^n rule would become the 2^(n-1) rule. For example, with 16 players you’d need 5 rounds, not 4, to guarantee at most one perfect score.
Bill Smythe
Sorry ,no longer USCF members.
I’d say then that, your other players concurring, either your method or Bill’s above would be fine. Though Bill’s idea is very nice indeed. Maybe a hybrid of those two, where the round 1/2/3 pairings are as Bill suggests, but you wait for the end of Round 3 and pair your final Round 4 with all results in.
If it stays under 15 players and I were the TD I’d use pairing cards to play it out if I didn’t have a computer. (And, if you get 15 players or more… a ten dollar donation from each would buy an old used laptop and your choice of pairing software for the club. 
)
One thing that might be considered here is that some players actually might enjoy the downtime between rounds.
By removing it you may also be removing the social aspects that actually make the club meeting enjoyable.
(this comes under the heading of you can’t please everybody)