On KOTH Pairings

Dec 23, 2011 12:02

I've been thinking about Scrabble pairings recently, and how best to do them. Here are some initial thoughts ( Read more... )

pairings

Leave a comment

Comments 33

cuzzakenji December 23 2011, 20:43:34 UTC
the MAIN goal of pairings is to match players (in what SHOULD be a SENSIBLE manner) to determine eventual outcome. ALL pairing systems should be adjusted AT SOME POINT to provide the opportunity to those players who can POSSIBLY win and cash.

As if the use of spread as tiebreaker is not idiotic enough, the spread rule of 500 is wickedly stupid. While it does not OFTEN effect pairings, it CLEARLY should NEVER.

In the absence of STANDARDIZED pairings (dat do whatdafuck dey SHOULD), all pairing system are flawed and hamstrung by idiotic rules.

The Fontes pairing system (dat lookback shit) is probly da best of a bad lot. but again, all pairing system should be adjustable as necessary (ask Rick n Jeff what dey used 2 do).

Reply

wantonhalo December 23 2011, 21:43:20 UTC
Okay, this seems reasonable enough, and I agree with you about the rule of 500. Could you be more specific about your notion of standardized pairings?

Reply

cuzzakenji December 24 2011, 07:17:01 UTC
quad the field to start (first 3 games) or snake it, fontes swiss thereafter UNTIL backstretch--the point where whomever got a shot 2 win (or cash) GET IT (STANDARD according to tourney length or ALWAYS wif 3 rds to go).

bandaid a gunshot wound (yeah, a GOOD paining system--i SEE da typo--n moving money is juss what da dr ordered)...

Reply


magratheazaphod December 23 2011, 21:15:00 UTC
the pairing system you're describing is exactly what was used this year at causeway. it was fun but there are some issues with it...I had a ton of back-to-back games against the same player because with X-2 in place, if you guys both won your last game before playing then you will most likely play again. it was quite an odd experience.

maybe X-2 but not looking at spread and pairing so as to minimize total repeats?

Reply

wantonhalo December 23 2011, 21:41:41 UTC
Disregarding spread and trying to minimize repeats sounds quite reasonable. If we can get fewer repeats at pretty much no extra cost, we should. Nice point!

As for your point about Jakkrit playing Nigel 5 times in a row, that doesn't seem so awful, unless they could've paired at differently as you pointed out.

Reply


magratheazaphod December 23 2011, 21:17:33 UTC
also it seems worse for long tournaments. jakkrit had to play nigel richards 5 times in a row on day 2. it ends up being a matter of timing your winning streak to the right moment to end up in the money for places 6-10 (1-5 you more or less are going to have to have gutted it out the hard way for a lot of the tournament).

Reply


spherulitic December 23 2011, 23:09:05 UTC
It would disadvantage a person in 3rd place, and make it harder for them to control their own destiny, if 1 and 2 are always playing each other.

Reply

wantonhalo December 24 2011, 00:55:23 UTC
Well, Jesse's idea of least repeats within similar records should avoid this for the most part. If 1 and 2 are significantly ahead of the field though, this situation is still feasible: other players had a chance to catch them earlier in the tournament. Plus, if they are in such a situation, they will beat up on each other and lower-ranked players have a chance to catch up.

Reply

spherulitic December 24 2011, 01:02:18 UTC
What you describe isn't that different from Swiss pairings or Chew pairings.

Early in the tournament, wouldn't it be better to treat all players with the same record as equals and pair them in a large group (rather than rating them strictly by spread as in a KOTH), and then towards the end of the tournament make it gradually more like a strict KOTH?

Reply

wantonhalo December 24 2011, 03:14:58 UTC
Treating all players with the same record as equals is "Jesse's idea" as aforementioned. Should we even have a strict KOTH until the last round? How much of a role do we want spread to play in the tournament at all?

Chew pairings and Swiss pairings seem okay, but I think they could do with fewer repeats, at least for people who are in contention. They also seem to give way too many people too much of a chance near the end of the tournament. I also agree with Marlon that the 500 rule is silly.

Reply


bbstenniz December 24 2011, 05:27:52 UTC
Tournament Scrabble isn't an ideal world. I personally think it's kind of silly, but people want to play as many opponents as they can. If they're spending $500 to go to a tournament, and can only afford to do it a couple of times a year, they don't want to play the same person 5 times in a 20 game tournament.

the KOTH-ish x-1 pairings at Causeway (I say -ish because every now and then they threw in groups-of-4-mini-round-robins) were interesting, and seemed to work fine, but I still think you'd have to deal with the PR aspects of pleasing people.

Maybe that's not what your after with this post; still something to consider, though.

Reply


Leave a comment

Up