What I did here is manually pair the second through fifth rounds as a round robin (not a round robin guided swiss) and then let a normal swiss pairing go through in the final round. I will grant that there is risk of a drop out or a very late add but doing the round robin let me avoid really weird colors, and I do admit that I made sure for round six rematch purposes that I could get good equalization reversing colors for the top two 1-0 players if they swept the rest of the field.
I’m curious
a) why not do the RR rounds in “Swiss-guided”
b) why revert to Swiss pairings in the last round (or, did they turn out to be the same either way - I didn’t check).
I think I would have been tempted to run this as a straight RR - except for the risk of dropouts (or, worse…additions!),
but my second choice is the RR-guided Swiss. This allows you to you to put off the really important game(s) until the last round(s). You do have to juggle the colors.
The way you did it, there were no color issues - but it probably felt more like a RR than a Swiss. Your 1st and last rounds may have had some Swiss considerations. My preference would have been for ALL rounds to be influenced by Swiss considerations.
How many TDs have used and know how to do cross-round pairings? Rather than give the1292 player a bye in round 2 pair him against the first player to lose in round 2. That first loser will be player his round 3 game, while the 1292 will be playing his round 2 game. You then will have 6 players to pair for round 3, other than the “first loser”. Going forward after round 3 you could continue with cross round pairings for 2 more rounds, or you could give full point byes for the last 3 rounds.
Larry s. Cohen
Thank you!
With 6 players, it is always possible to fit the Swiss to RR after two rounds. But the rounds may not be consecutive in the table, so the colors might look screwy at first.
Bill Smythe
When doing a RR-Hybrid, the colors in the RR table are irrelevant.
Yes, I was aware that that was the rule. That’s one reason I suggested it.
Yes, I was aware that that was the rule. That’s one reason I suggested it.
Using the colors from the table assures (if there is an odd number of players) that each player will receive equal whites and blacks. If you assign colors on the fly, it will be hard to avoid giving some player(s) two extra blacks and other player(s) two extra whites.
Bill Smythe
Please read 29L.
When using the RR-Hybrid feature in WinTD, that’s true, because that option in WinTD apparently tries to wing the colors to look the way they would in a Swiss. That has advantages and disadvantages.
If the TD wants to make sure each player’s colors come out even at the end, he might be better advised to use the colors in the table.
Well, yes, that rule places more emphasis on getting the colors to look right near the beginning, than to equalizing colors after all is said and done. And that’s a legitimate approach when using the RR tables to guide a small Swiss, if there is no intention of completing the round robin.
But if the organizer is trying to complete the round robin, equalizing the total colors seems to be a reasonable goal.
In the example that began this thread (not the later example presented by jwiewel), IMHO the organizer should have tried to complete the round robin.
Bill Smythe
Yes, I agree that the organizer would have been better off trying to complete the round robin – including giving a bye (instead of an unplayed game) to the late entrant. As it was, this player ended up with a U and a bye.
Running a “straight” round robin (without cross-round pairings) would involve having 7 rounds instead of 6, with each player getting a bye at some point in the tournament. Using cross-round pairings, as you suggest, with any luck, would still have allowed the tournament to finish in the same amount of time as a 6-rounder.
However, cross-round pairings in a 7-player round robin is quite a different beast, not at all like cross-round pairings in a standard (and reasonably large) Swiss.
First of all, once the decision is made to convert to a round robin, all the pairings (for the rest of the tournament) should be posted. In addition to the already-posted round 1 pairings, you would now post six more pairing sheets, one for each round 2 through 7. This would be done while round 1 is still playing. Round 7 would be a “ghost” round – you would want to get those games out of the way early if at all possible, one by one.
In round 1, you wouldn’t want to just pair the lowest player (who is temporarily sitting out) against the first loser. Ideally, you should pair the lowest player against whoever he is scheduled to play in round 7. If that isn’t possible (because that opponent is in a long round 1 game), find some other opponent (could be first loser, or first winner, or second loser, etc) and look on the rounds 2-7 pairing sheets to find out which player is supposed to have the white pieces in that game. Always honor the colors listed on the rounds 2-7 pairing sheets – that way colors will even out in the long run.
Keep hustling the players to play extra games (cross-round pairings) in this manner throughout the tournament. By the time you’re halfway through the event, it may not even be important anymore to play the games in a specific order. Just observe colors from the pairing sheets, and mark each result on the appropriate pairing sheet as soon as that game is over.
One possible problem is that cross-round pairings do not work well with higher-rated players, because those players tend to have longer games. This event was a scholastic state championship, so the skill level of the players might have been relatively high.
Still, IMHO cross-round pairings should have been tried, as outlined above, essentially converting the event to a 7-player round-robin.
Bill Smythe
A correspondence of pairing numbers on the Swiss wallchart and the numbers on the guiding pairing table is not necessary, and is likely not desirable.
When I follow Rule 29L, I assign each player a separate round robin number for my use only, not for publication, such that the round one pairings fit on row 1 of the pairing table. The purpose of the pairing table is not to pair a round robin, but to guide a Swiss to a conclusion without a repairing. That’s why I’m in the Ken Sloan camp on the color assignment issue.
That’ll work, as long as you’re able to keep track of the two numbering systems without getting mixed up.
That’ll work too, if it doesn’t bother you that some players may get two more blacks than whites.
If the intention is to actually complete the round robin, then I would think the argument in favor of following the table colors is stronger than it would be if the plan is to fall 1 or 2 rounds short of a round robin. In the latter case, I’ll agree that it’s not worth the trouble, and might not work anyway.
Bill Smythe
In practice, after the first two rounds (at which point the top score group will almost always have three or fewer players), you pair the highest ranked player with the highest ranked player she has not yet played.
That may be true with six, but I’m not sure that it’s true with more than that.
Indeed, I’m not sure either, and I don’t have the slightest idea how to work out the combinatorics.
If N is the number of players, and C is the first round number where you must be careful, and U is the first round number that might be unpairable if you’re not careful, then I suspect the table would go something like this:
[size=150][code]
N C P
6 3 4
8 4 5
10 5 6
12 6 7
[/code][/size]
but as I said, I can’t prove any of this, except I know that it’s correct for N=6.
Bill Smythe
And, there’s no need. It’s bad practice to switch back and forth between pairing systems during the event.
With a 29L event, the tournament is always a “Swiss”. The difference is that the pairings are guided by the RR
tables. You should know you are doing this before R1, but it makes no difference to WinTD until you are about to pair R2.
In particular, if you get a few late entries, you can abandon the table and depend on the fact that normal Swiss pairings will work. Question for Tom - can WinTD handle the situation where you get 4 late entries in R3, after starting off as
a Hybrid? [perhaps unusual, but definitely not “rare”!] Can you turn it back to a normal Swiss and keep the R1 and R2 pairings and results?
Cleaning up all the loose ends: the entire history of the “6-player trap” reminds me of the old joke “Dr! Dr! it hurts when I do that”. It has a certain appeal for those more interested in solving puzzles than in actually running real-life events. It’s a great topic for the kind of discussion that used to be common in backrooms 15 years ago. In my opinion, it’s easier, and better, to simply avoid situations where such “traps” occur - whenever possible. A 6-player event with 4 rounds qualifies, in my opinion. Especially in an event where late entries, or withdrawals, are common. Similarly, in my opinion, the “6-player trap” is not really suitable for a Local or Senior TD test. It might be appropriate for the NTD test, but there the appropriate answer might be “Why did the TD allow this situation to arise in the first place?”
You can always switch to a Swiss no matter how you started. It’s switching from a Swiss to an RR (can’t do it after round one), or RR/Swiss hybrid (can’t do it after round two) that are restricted.
You’re confusing two issues. The fact that there’s a way clear to pair the rest of the tournament means that there is some round robin system in some form that will work. It does not mean that there’s a numbering that would make that a Crenshaw-Berger table or any other specific RR schedule fit the first x rounds. For instance, the Crenshaw-Berger six player table relies heavily on repeated colors (for instance, rounds 1 and 2 have 3 playing WW and 6 playing BB). If you pair the first two rounds with everyone alternating, unless I’m missing something, I don’t believe there is any pair of rounds out of the RR table that will match that, as any pair of rounds will have one player with two W’s and one with two B’s.
Ah…but, as previously noted - if you are switching to a 29L Hybrid, colors don’t matter.
In my opinion, it is improper to switch from Swiss to RR once the first round is paired. But, IF POSSIBLE, I think it would
be OK to switch from Swiss to a 29L Hybrid. Here, “IF POSSIBLE” means that:
a) you can find rows in the appropriate table that match the pairings, so far, ignoring colors
b) can’t think of another restriction!
On the third hand, I don’t think this represents anything other than a “neat puzzle”. The number of times this would actually occur (given that the TD was reasonably competent up to that point) approaches zero.
Cross round pairings were suggested as another alternative to the bye problem, they were not suggested in addition to a substituting/converting to a round robin format!
Larry S. Cohen
Cross-round pairings are certainly useful as an alternative to byes in general, but in this case your suggestion came in a thread about small events which are close to being round-robins. If cross-round pairings are to be involved in a case like this, the details are a bit different from what they would be with a normal “large” Swiss.
Bill Smythe