Monday, February 14, 2011

Bracketology following BU's loss to Harvard

By Sam Dykstra/DFP Staff

The immediate gut reaction to BU’s 5-4 loss to Harvard in Monday night’s Beanpot consolation game is “Just how long has it been since the Terriers finished fourth in what some have called the BU Invitational?” (It’s been 31 years since that’s happened, but there’s plenty about that in the game recap and grades on this site.)

Once that wears off – if it hasn’t already for some BU fans – the second reaction should concern what the loss to a team that’s ranked lower than 50 on the RPI rankings means for the Terriers NCAA tournament hopes.

BU coach Jack Parker said it himself, “I think it is a big loss for us. I think it is a loss to an unranked team. It is a loss that we have to get many more games over 500, and that just made it difficult to do that. As far as the NCAA Tournament selection process is concerned this is a huge loss for us.”

Before the loss, the Terriers were squarely on the bubble for the 16-team tournament, and they remain there despite the loss to Harvard. But the road to Manchester, Bridgeport, St. Louis or Green Bay (the locations of the four NCAA regionals this spring) may be a little bumpier now than it was at 4:30 Monday afternoon.

Before we lay out BU’s NCAA resume, here’s a couple of notes you should know. Unlike our usual stories, all rankings used here will be from the PairWise Rankings, a system that is very similar to the one used by the NCAA selection committee. That system ranks teams based on Ratings Point Index (RPI), record vs. teams under consideration (teams with RPIs of higher than .500), head-to-head games and common opponents.

With that in mind, let’s look at BU’s NCAA resume.

No. 17(T) Boston University (14-9-7)
RPI: .5326
Individual Records vs. TUC:
No. 2 BC 0-4-0
No. 4 Merrimack 0-1-2
No. 9UNH 2-1-0
No. 12 Notre Dame 1-0-1
No. 14 RPI 0-1-0
No. 15 Wisconsin 1-0-0
No. 19 Maine 1-0-2
Overall Record vs. TUC 5-7-5
Bad losses:
Brown, 6-1 on Jan. 1 (Shillelagh Tournament)
Harvard, 5-4 on Feb. 14 (Beanpot consolation)

So with that in mind, let’s move onto teams that will be battling out for the final at-large spots in the tournament as the 14th and 15th seeds (assuming the winner of the Atlantic Hockey Conference takes the 16th spot). It looks like if the season ended BU would battle Rensselaer Polytechnic Institute, University of Wisconsin, Western Michigan University and Colorado College for those positions. Here’s how those teams have performed in the crucial PairWise statistics and how exactly that would stack up against the Terriers.

No. 14 RPI (18-8-4)
RPI: .5422
Record vs. TUC: 7-5-2
Head-to-head with BU: 1-0-0
COP (Northeastern, Harvard, Brown): 3-1-1, BU 2-2-1

How they match up with the Terriers: The Engineers comfortably beat the Terriers in each category. With only one regular-season game against a TUC left (Princeton) and none against common opponents, it’s safe to say that the Engineers would take a spot over the Terriers with ease. Had the Terriers defeated Harvard Monday night, the two teams’ records against common opponents would have been equal, and that could have given the Terriers at least something to cling to against the Engineers.

No. 15(T) Wisconsin (19-10-3)
RPI: .5407
Record vs. TUC: 5-10-3
Head-to-head with BU: 0-1-0
COP (UMass): 2-0-0, BU 2-0-1

How they match up with the Terriers: On the surface, the BU vs. Wisconsin debate could go either way. The Badgers lead in RPI and records against the two teams’ only common opponent (two wins in two games looks a lot better than two wins in three games.) The Terriers have the advantage in records against TUC and won the only game between the two teams. However, the tie-breaker here is RPI, according to the PairWise rules, and that gives the nod to the Badgers.

No. 15(T) Western Michigan (15-7-10)
RPI: .5359
Record vs. TUC: 4-4-4
No head-to-head games with BU
COP (Notre Dame): 0-1-1, BU 1-0-1

How they match up against the Terriers: The Broncos win the battles for RPI and record against TUC with BU winning records against the team’s only common opponent in Notre Dame. That would seemingly give Western Michigan the edge. But luckily for BU, there is still plenty of hockey left to be played. The Broncos have four games remaining, and they’re all against TUCs (two on the road against Michigan, a home-and-home with Notre Dame). If the Broncos don’t win any of those contests, the Terriers could take the advantage both in that category and overall. For what it’s worth, Western Michigan has yet to play Michigan this season and has not defeated the Fighting Irish in two tries.

No. 17(T) Colorado College (17-14-1)
RPI: .5293
Record vs. TUC: 11-14-1
No head-to-head games with BU
COP (RPI): 1-0-1, BU 0-1-0

How they match up against the Terriers: The Tigers (record vs. TUC, record vs. common opponents) beat out the Terriers (RPI) 2-1 in the PairWise categories. This is proof that the loss to RPI not only hurt BU against the Engineers but also against Colorado College. Had the Terriers defeated the Engineers on the road, they would have taken the lead over Colorado College, who has held its own more or less despite playing the fifth-toughest schedule in the nation this season. Luckily for BU though, the Tigers loses to No. 19 Ferris State, a team BU beats out, according to the PairWise parameters (Ferris State has better records against TUCs and common opponents). That leaves both teams tied at No. 17 in the PairWise, meaning BU wins out because of the RPI tie-breaker.

So to sum up, Colorado College wins a one-on-one PairWise matchup with BU but loses out in the big picture because of its loss to Ferris State.

Overall analysis: Things aren’t looking too good as of Monday.

It’s easy to say after the fact that if BU had done this or that, then it would be in the tournament. Heck under that discussion, you could argue if the Terriers had gone undefeated we wouldn’t be even need to write a post about their NCAA chances.

But yet, here we are. So if you want to look at a couple of games that could have helped the Terriers’ cause, the four losses to BC wouldn’t be a bad place to start. Had BU split the season series with its rival, it would have a 7-5-5 record against teams under consideration, much better than its current mark of 5-7-5. Or if the Terriers had scored just one more goal during its three-game stretch of ties against Merrimack and Maine, perhaps that could have given them a slightly stronger resume as well.

The fact still remains though that Monday night’s loss to Harvard, a team with only five wins after the Beanpot, did nothing but weaken BU’s case for a chance at the national title.

Hope remains in that there are still six games remaining on the schedule in which the Terriers can build their resume as well as potential games in the Hockey East tournament against TUCs BC, Merrimack, UNH or Maine. Or they could win Hockey East altogether. For now though, BU’s tournament chances look pretty bleak.

For sticking with me throughout this massive post, here’s my thoughts on how the NCAA seeds would shake out if the season ended today.

1. Yale
2. BC
3. North Dakota
4. Denver
5. Merrimack
6. Minnesota-Duluth
7. Union
8. Nebraska-Omaha
9. UNH
10. Michigan
11. Miami
12. Notre Dame
13. Dartmouth
14. RPI
15. Wisconsin
16. RIT

On the bubble (in order): Western Michigan, BU, Colorado College


  1. You've got way too much conjecture and arbitrary decision. This is a strictly formulaic decision by the committee. They don't make any "decision" other than seeding teams and placing them in regionals based on travel, host sites, etc. So your end seeding is correct, but because it's the top 15 plus the AHA winner (provided favorites get the auto bids). Your bubble teams are out of order, BU has the tie breaker over CC with the better RPI. I applaud you for taking this on, it's really tough to predict this stuff, especially with the new TUC rules this season, but in order for this to be most accurate, make sure you're talking more in absolutes with this and less open-ended speculation about who the committee may or may not like, this is not like basketball. There's no gray area with the PWR, the formula is what it is, and you can predict with absolute certainty after all tournament championships are played who gets in. Only gray area is who plays who and where.

  2. I've edited the post to include the RPI tie-breaker for both the Wisconsin and CC sections. Hopefully, that makes it all less arbitrary.

    Side note: this is not the be-all, end-all of how the NCAAs will shake out for BU. Just a look at how it stands right now. This is all subject to change on a game-by-game basis as the season comes to a close.

  3. The BU season is OVER!

  4. Actually, the selection committee has never explained how it breaks ties. Some think it's RPI, some think it's the head-to-head PairWise comparison. If it's head-to-head, BU wins the tiebreaker over CC. Also, if it's head-to-head, Merrimack wins the tiebreaker over Denver for the final one seed. Wisconsin wins the tiebreaker over Western Michigan either way.

  5. We actually have atleast six games left plus the HE tournament so it is not over. Good effort though

  6. No trophies this year - the season is over!

  7. We actually already have a trophy from the Icebreaker in October so ha!

  8. LOL - you lost to Harvard to capture LAST place in the Beanpot!

  9. I don't see how that is relevant to us winning the icebreaker. Go back to BC interruption where you belong

  10. Who will Parker call out and throw under the bus next?

    Based on performance, it should be useless Parker Suck Up Warso. That -3 did as much to help BU come in last place in the Beanpot as Parker's incompetent coaching!

    Kudos Warso!

  11. Re: tiebreakers. I think there has only been one time ever that the team with the lower RPI has been placed above the higher RPI team. I know that part of it is not definitive, but it's a pretty determinant factor. I hope more of these articles continue, with a couple of tweaks, this is a topic that can produce some really good reads. Usually you don't see anyone making PWR projections until conference championship weeks, which by then it's very late in the process, so again, props for taking this on a month earlier than everyone else. PWR is a somewhat complicated system that is unfamiliar to most, even to many dedicated fans of the sport. Might be worth it to preface a future article with a step by step breakdown of how the process works, to provide some more clarity on how definitive it is, and that predictions are very difficult to make in advance, and that you're not really trying to predict what the committee will do, in terms of teams getting in or not, but predicting how results will shake out, and how those subsequently affect the PWR. Ignore the other BS in these comments. And as always, Go Whalers.

  12. "We actually already have a trophy from the Icebreaker in October so ha! "

    Do they actually give BU a trophy for that? If they do, I've never seen a picture of it. Anyone have definitive proof here?



    btw, nice one on the handy picture of the icebreaker trophy, scott