Here are the results, along with the official seed and this year's RRI:
Rank Team PMR Seed RRI
1 Sockeye 3009 2 2838
2 Furious 2997 1 2774
3 DoG 2870 3 2728
4 Bravo 2803 5 2722
5 Ring 2750 4 2731
6 SubZero 2737 7 2687
7 Revlvr 2728 6 2710
8 Chain 2719 10 2670
9 Rhino 2704 8 2690
10 Condors 2674 9 2665
11 Vicious 2673 11 2715
12 Metal 2634 12 2591
13 Machine 2608 14 2599
14 BAT 2577 13 2554
15 TrkStop 2556 15 2565
16 Monster 2365 16 2373
Here's how:
1. Enter in the tournament RRI for every team. Give a 40 point bonus for winning. Give a 20 point bonus for finishing 2nd. (Possible enhancements: make additional changes (5 or 10 points) for each win or loss in the tournament. I think RRI underweights wins, since it relies on point differential to make accurate predictions.)
2. Assign a weight to each tournament, since some are more important than others. I gave a weight of 2 to the major tournaments held in July or later (Colorado Cup, ECC, Chesapeake, Labor Day, Tuneup, each Regionals), a weight of 1 to lesser tournaments later on (including Sectionals) and major tournamentes earlier (Boston Invite, Solstice, Live Logic), and a weight of 0.5 to the lesser early tournaments. (Enhancements: tweak the weights further).
3. Take a weighted average for each team. If a team has fewer than 8 weight points, add tournaments with the RRI of a low-level Nationals team. For instance, a team has only Regionals (2 points), Sectionals (1 point), and one lesser tournament (1 point) for an average RRI of 2700. Add 4 tournaments worth of 2500 RRI for a modified RRI of 2600. This is to encourage play, make sure one fluke performance doesn't put someone too high, whatever.
4. Add points for last year's Nationals. I chose 200 points for the winner, 150 for 2nd, 100 for the semifinalists, 50 for quarters, 25 for 9-16. I decided to give partial credit to teams who didn't make it but whose regional equivalent did well. For the first iteration, the only team affected was Revolver, getting 25% credit for Jam's semis appearance. I didn't do anything to any of the Mid-Atlantic teams, though, which might not be right. (Enhancement: finer gradations; incorporate entire season.)
Comparison to my previous ranking:
Subzero down 3 to 7th
Ring up 2 to 5th
Chain up 2 to 8th
Machine up 2 to 13th
Others within 1
I can live with either, or the official.
Subscribe to:
Post Comments (Atom)
15 comments:
All this seedings talk is boring the f out of me.
Could Nationals-bound teams maybe talk about what their final days of practice are like?
we usually sit around and discuss how the seedings will pan out.
then we discuss why the seedings are wrong once they come out.
$
I have never seen so many pools drawn up on napkins in my life... I thought regionals was bad.
yours, -CTW
back in the day you could be out after the first game, a la double versus chicago in sarasota in '97, with few to no ways to get back into the championship bracket. but with the newer format it seems that any seeding within reason is fine b/c everyone gets ample chances - the system resulting in efficient separation of wheat from chaff.
Summary of post:
"I wanted to present empirical evidence of what I think. So I applied arbitrary numbers to algorithms in an attempt to appear analytical.*"
Get back to writing about topics you know about. Mathematical theory isn't your strong suit.
*I attemtped to appear alliterative.
*I attemtped to appear alliterative.
Alternative version from a different anonymous poster (not necessarily sharing the same view): "I wanted to present empirical evidence of what I think. So, attempting to appear analytical, I arrogantly applied arbitrary asinine arithmetic and algorithms."
whew. that was fun.
I can't believe the UPA is taking so long in getting these seedings/national pools set. This is worst than the NCAA BB guys stuck in a room for a week in late April...Let's go..
Stebbins,
Double could have beaten DoG to make it, but as Corey would probably say, we cheated them out of it and laughed about it afterwards. The flip side to the new format is that strong play on Days 1 and 2 doesn't help you that much. You could find yourself undefeated and matched up against Sockeye, as Ring did a couple years ago.
Alliterative Anonymouses,
You forgot to say ", asshole." There is always a cry for specifics on how to seed. Forgive me for writing about topics I know a lot about (seeding, analysis) because I quantified it without rigor. The numbers are arbitrary but approximately represent the weight I give to events, and were in no way chosen to get desired results. I didn't even look at the results until everything was done, in fact, so the fact that this algorithm so closely matched the actual seeding, better than RRI itself did and just about as well as any individual's seeding stands as credit to the arbitrary assignments. Seeding is imperfect, anyway.
"The flip side to the new format is that strong play on Days 1 and 2 doesn't help you that much."
... Which is to do with the format, but more to do with the greater depth and parity in the Open division. Back in the good old days, drawing the 6, 7, or 8 seed as your Quarters matchup actually meant an significantly easier game. Now, you may get an 8 or 9 seed that is capable of taking down anybody (Rhino and Condors, for example), OR you could get an higher ranked team that has already run into trouble from those lower seeds (like Ring-Sockeye in '04). The biggest trap of the current format is thinking you've won something with strong play on Thursday and Friday. At least on those days, if you lose you get another chance. Not so on Saturday.
- MC
JP-
It isn't that Double was mathematically eliminated, just that their road to the championship bracket was a really tough/virtually impossible haul. (Didn't they have to not just beat you, but beat you by more than Chicago beat them making it more " virtually impossible" rather than just "really tough"? And wasn't there some travel snafu that resulted in some of their ptp'ers not being there for that round 1 game on the first day?)
In any case, I think of the newer format as the nine-lives format.
stebs
Double's travel snafu was that they got knocked out at Regionals but Pack of Lies decided not to go to Nationals, and some Double guys decided not to go as well.
Since DoG beat Z by a bunch, Double would have come out 2nd in a tiebreaker if they won by one.
In the old, old days that Stebbins referred to, there were no quarters and a single pool play loss could knock you out (although it never happened in club, but did in college). But you're right, too, the lack of distinction between teams 3-8 is the primary reason compared to when quarters were first introduced in 1999.
I can't believe the UPA is taking so long in getting these seedings/national pools set. This is worst than the NCAA BB guys stuck in a room for a week in late April...Let's go..
Dude, relax. The last Regionals was Oct. 6-7. Then all the teams that qualify get to weigh in on seeding (which I think is important or at least worthwhile); they are given 5 days, and then seeds come out that Monday. Sound reasonable? It's been that way for years.... why your panties got bunched up this year is anyone's guess.
Hmm, so Parinella says that quarters was introduced in '99. DoG hasn't won a championship since '99.
Elaborate, please.
Quite simple, Mr. I Hate Sarasota Because They Make Me Buy A Ticket To Attend. They introduced it that year to try to break up the dynasty, but we won it anyway, then decided we had nothing left to prove so we started letting other teams win. We were very clever with it, too, sometimes losing in semis, sometimes in quarters, sometimes in poorly played games, sometimes in games for the ages. It's all part of the master plan, culminating in 2006, exactly 12 years (12 being one more than 11, which is one more than 10) after the first. Coincidence? I think not.
nationals has gone over the cliff since 2003. your thoughts as to why.
i am standing by w/ a skilled north korean id-falsifier to join a team...
Post a Comment