Thanks for doing this - please keep it up!
First, a little history. Thanks to the Build Your Own Rankings Calculator we now have 11 years of Pairwise data to study. In that time there have been 20 teams that qualified thanks to the autobid and 156 who qualified by being ranked high enough (autobid or not) to make the NCAA tournament. Of the 156 teams, 112 (71.8%) that qualified as of the January 1 PWR would have made the eventual field. For teams ranked in the top 8 that percentage gets better. 38 of the 44 (86%) teams that were ranked 1-4 in the January 1 PWR have made the tournament. 33 of the 44 (75%) of teams ranked 5-8 made it. 27 of the 44 (61%) of teams ranked 9-12 qualified. In 2011 all of the top 12 teams in the January 1 PWR qualified. Ohio State is the only #1 team (2012) to fall all the way out of the tournament. If your team doesn't appear below, take heart. In five of the last seven years teams that were not in the top 25 of the PWR on January 1 earned at-large berths in the tournament. Last year, Wisconsin was ranked #42 in the RPI (unranked by pairwise) and rose to the #14 overall seed. Two years ago Union was ranked #25 and came all the way back to win the ECAC, get a #1 seed and advance to the Frozen Four.
34 teams fell out of the NCAA tournament that were ranked 1-12 in the New Year PWR. Colorado College and Dartmouth have the distinction, not only of falling out of the tournament four times but, being #1 seeds that fell out of the tournament. Last year the Big Green were ranked #4 and finished the season ranked 23.
This year has seen more changes to the pairwise system which have been discussed elsewhere. We may see some odd movement since all 59 teams are now ranked for pairwise purposes, but there is no more "TUC cliff" and the quality win bonus is on a sliding scale now, so hopefully we won't see wild fluctuations come March. I am using the pairwise from Reilly Hamilton as they appear to be the most accurate out there. The same formula was derived (independently) at Sioux Sports so I believe they are correct. Jim Dahl has a lengthy history with the pairwise and I trust his calculations. If you're interested in how the sausage is made, I highly recommend Reilly's site as it goes into great depth on how each calculation is made.
Here is the first official bracketology of the New Year
2 St. Cloud State
3 Boston College
4 Ferris State
12 New Hampshire
15 Minnesota State
34 Air Force
18 Notre Dame
20 North Dakota
First we make sure all teams that qualify for an at-large berth have a .500 record or better. They do so we can proceed.
The only host to make the field is Minnesota, so they go to St Paul. The remaaining #1 seeds are placed closest to home. So St Cloud goes to Cincinnati, Boston College to Worcester and Ferris State to Bridgeport.
The #2 seeds are then distributed according to the #1 seeds, in a 1v8 2v7 etc fashion. Providence to Bridgeport, Quinnipiac to Worcester, Union to Cincinnati and Lowell to St Paul, Cornell to Cincinnati, Clarkson to Worcester and New Hampshire to Bridgeport.
The same pattern goes for the #3 seeds: Michigan to St Paul, Cornell to Cincinnati, Clarkson to Worcester and New Hampshire to Bridgeport.
Air Force is the automatic qualifier from Atlantic Hockey, so they are seeded 16 and play in St Paul. Mankato goes to Cincinnati, Wisconsin to Worcester and Northeastern to Bridgeport.
Code:Bridgeport (Yale) Worcester (Holy Cross) Cincinnati (Miami) St Paul (Minnesota) Ferris State Boston College St Cloud Minnesota Providence Quinnipiac Union Lowell New Hampshire Clarkson Cornell Michigan Northeastern Wisconsin Mankato Air Force
That will not work as we have three intraconference games in the first round. It can be fixed easily enough be swapping the three seeds within regions: Michigan to Cincinnati and Cornell to St Paul; New Hampshire to Worcester and Clarkson to Bridgeport. That gives us a final bracket of:
The only thing left is attendance concerns. St Paul is all set with the Gophers there, Bridgeport has three eastern teams, as does Worcester. Cincinnati has Michigan. We might be able to bolster attendance by switching Mankato and Wisconsin, but IMO not enough to warrant te change. That remains our final bracket.Code:Bridgeport (Yale) Worcester (Holy Cross) Cincinnati (Miami) St Paul (Minnesota) Ferris State Boston College St Cloud Minnesota Providence Quinnipiac Union Lowell Clarkson New Hampshire Michigan Cornell Northeastern Wisconsin Mankato Air Force
About this thread: I have been doing this thread for several years but by no means do I consider myself an expert. Questions, suggestions, corrections all are welcome. As you can see above, bracketology often involves judgement calls and we all have our own opinions. I try to mimic what I think the committee would do, even if I don't personally agree. With that, let the countdown to Selection Sunday begin!
Thanks for doing this - please keep it up!
Thanks for the new thread Priceless.
In general, I believe that we will see less volatility this year because of the absence of the TUC provision. I say this because having 3 criteria rather than 4 should make things more stable.
Also, in general, I believe the RPI numbers will be less stable this year than before, because the Quality Wins Bonus has more room to jump around (If you had beaten a school that previously was #1 in the raw RPI, and then they are swept, your QWB goes down because their placement in the order goes down, too). However, this will be partially mitigated by the fact that the QWB is averaged over the whole season.
In all, I think this way is more fair than last year's way.
It will be interesting to see what happens.
Fascinating that so far, the power resides quite strongly in the east, whereas it has been more common lately for the power to be in the west.
Priceless, I also appreciate all the work you do on this annual thread, one of my favorites. Is there a possibility to post a longer list of teams in ranked order each week? Considering the changes, it might be nice to see the top 30 or 40 teams. If it is more work for you, please do not bother, thanks.
How close are your numbers to those on Sioux Sports? Do their calculations also include the changes this year?
What should be the most authoritative source for the pairwise factoring the new changes?
DANGER - MEN AT WORK
Or, maybe my east-coast bias filter is not working today, probably frozen
"I'm not surprised by the results you didn't get from the work you didn't do." (seen on a T-shirt)
Great annual thread...especially the last three years.
LET'S GO UNION DA DA DADADA
Went to take a look at the rankings, and it looks like USCHO and CHN are doing something different in their calculations and both are different than the list here from Sunday (could easily be the last handful of games moves stuff around there, so I don't know).
Maybe this question belongs in the thread that's out there where a few people have been running the calculations themselves all year to get a feel for how things work, but does anyone have a good idea which one of them is right? My guess is that one of the two isn't including the QWB that was introduced this year.
As of 1/8:
1 Minnesota 1 Minnesota
2 Boston College 2 St. Cloud State
3 St. Cloud State 3 Ferris State
4 Ferris State 4 Boston College
5t Union 5 Union
5t Quinnipiac 6t Providence
7t Providence 6t Quinnipiac
7t Mass.-Lowell 8 Mass.-Lowell
9t Michigan 9t Michigan
9t Northeastern 9t Cornell
11 Clarkson 11 Northeastern
12 Cornell 12 Vermont
13t New Hampshire 13 Clarkson
13t Vermont 14 Wisconsin
15 Wisconsin 15t New Hampshire
16 Minnesota State 15t Notre Dame
The Sioux Sports and Reilly Hamilton pairwise are the same. The USCHO pairwise also appears to use that formula. CHN is using the formula from last year I believe, and has not accounted for the changes.
Unless I am missing something,10 of the top 15 are from the east, with only Minnesota, Ferris, St Cloud, Michigan, Wisconsin from the Big 10, NCHC, or WCHA. That seems like lots of eastern power, especially compared to past years.
There are currently 1 users browsing this thread. (0 members and 1 guests)