Michigan Tech's NCAA Chances

In this week's preview I gave a very small glimpse into some of the numbers for Michigan Tech's chances on making the NCAA tournament based on data published by CollegeHockeyRanked.com (CHR). Now I'm going to dig a little deeper for those of you who are interested and pull information from CHR, CollegeHockeyNews.com (CHN) and our friend Geof Morris from Alabama at wchaplayoffs.com.

Geof Morris used his own BELOW (Bring ELO to the WCHA) rankings, which are based on the ELO ratings used to rate chess players and 538 uses them quite regularly when talking about professional sports teams. Using Geof's analysis, he predicts that Michigan Tech will lose to FSU 28.6% of the time and win the Broadmoor Trophy 43.0% of the time. The other 28.4% of the time, MTU will lose in the championship game.

Based on CHN's probability matrix and knowing they use KRACH, I was able to determine the probabilities their matrix predicts for each of the three possible outcomes for MTU as well. CHN has MTU losing 33.8% of the time, winning it all ~39% and losing in the title ~27.2%. The last 2 are estimates because they only report their data to the nearest percent.

As you can see both Geof and CHN get very similar numbers for the outcomes for the WCHA tournament. If we average them, we're looking at 31.2% chance of losing to FSU, 27.8% chance of losing in the championship game and a 41.0% chance of winning the WCHA Tournament and earning the automatic qualifier. Jim Dahl at CHR calculated the probability of each team getting in depending on how many wins they get this weekend. For Michigan Tech, their odds are around 1% if they lose to FSU, 14% if they lose the championship game and obviously 100% if they win Broadmoor.

Also of note, Jim has broken down the seeding based on all of the above. Michigan Tech will be ranked somewhere between 13 and 15 if they make the tournament and fall into the fourth band, meaning they would play a #1 seed somewhere, with some combination of the following extremely likely: 

  • North Dakota, Denver, or St. Cloud in St. Paul or Cincinnati 
  • Providence, Quinnipiac, or Boston College in Worcester or Albany

So based on Jim's probabilities and Geof's data, we can predict MTU's overall probability of making the tournament at about 47.2%. That's slightly lower than CHN's Monte Carlo predictions, showing MTU at 51% but much better than CHR's 29.1%. Now remember, CHN is using KRACH to remove some of the least likely combinations, while CHR includes every possible combination, including something like Wisconsin winning the Big Ten Tournament (which is impossible now after their loss yesterday to Penn State).

What the Huskies Need

Now to actually show you who to root for this weekend. First of all, Michigan Tech's best chance for an at-large bid rest in the majority of games going chalk (the better seed winning), with one big exception, Michigan winning the Big Ten tournament as the #2 seed. Now to break down Michigan Tech's preferences for each conference (most of this applies to staying ahead of Cornell):

WCHA Tournament - Grand Rapids, MI

  1. Ferris State
  2. Bowling Green
  3. Minnesota State

In the WCHA, things are pretty obvious. They want to win the games themselves and forget about everyone else, but if that doesn't happen, Minnesota State is the team Michigan Tech wants to win the least, simply because BGSU and FSU help MTU's RPI more based on playing them more times. So in Friday's early game, Michigan Tech should be rooting for Bowling Green and in the event that MTU loses Friday night, the Huskies will have a better shot of getting in with a Ferris State championship.

NCHC Tournament - Minneapolis, MN

  1. Denver
  2. North Dakota
  3. St. Cloud State

      Bad: Minnesota-Duluth

Minnesota-Duluth is one of the main teams that Michigan Tech should be concerned with this weekend when it comes to getting an at-large bid. If the Bulldogs win a game, either the the semifinal or third place game, they'll jump ahead of the Huskies and make it more difficult for MTU to earn an at-large bid. Beyond that, in most situations, UND, SCSU and DU make little difference, but it appears Denver winning is slightly favorable.

ECAC Tournament - Lake Placid, NY

  1. Harvard
  2. Quinnipiac
  3. Dartmouth (but not Championship)

       Bad: St. Lawrence

The ECAC is an interesting conference from Michigan Tech's perspective. If they lose to Ferris State, additional focus is placed on staying ahead of Cornell, who was eliminated last weekend. Because of that, Harvard gets the top ranking here. Also in that scenario, Dartmouth needs to beat Quinnipiac, but in general, a Bobcats win is preferable so they remain second on the list, with St. Lawrence making the bad list. Dartmouth winning the championship is also a bad thing.

Hockey East Tournament - Boston, MA

  1. Massachusetts-Lowell
  2. Boston College
  3. Providence

      Bad: Northeastern

ortheastern is the other main team that Michigan Tech is fighting with for one of the last at-large spots. If they manage to win a game, the odds of the Huskies finding their way in without the Broadmoor drop considerably. Beyond that it appears that a Massachusetts-Lowell championship is the most beneficial result followed by Boston College and Providence.

Big Ten Tournament - St. Paul, MN

  1. Michigan

       Bad: Minnesota, Ohio State, Penn State

n the Big Ten, things are pretty simple. Michigan needs to win the tournament and every game Michigan State would have won would have helped the Huskies due to their three regular season matchups. Beyond that, what happens between Ohio State, Penn State and Minnesota isn't a big deal. 

Atlantic Hockey Tournament - Rochester, NY

  1. Robert Morris
  2. Air Force
  3. RIT
  4. Army

tlantic Hockey appears to be the least important of all to Michigan Tech with slight changes in RPI which may make the difference if other results don't go quite as ranked above in other conference tournaments.

These rankings also change depending on which team MTU is worried about at the time, whether its Cornell, Minnesota-Duluth or Northeastern. Now lets take a closer look at things based on when Michigan Tech loses.

Loss to FSU

If the Huskies were to lose to FSU, the easiest way for MTU to get in was for all of the favorites to win except for Michigan State over OSU and Minnesota before losing to Michigan. If that had happened, results in the WCHA and AHA tournaments would have become irrelevant, as would who wins HEA, as long as Northeastern didn't win a game. For the ECAC, St. Lawrence can't win a game and Dartmouth can't win the tournament.

With Michigan State losing to OSU in overtime, things take a dramatic turn. Michigan Tech will require the WCHA to go all underdogs* in order to get in so we'll know before the Huskies even play if this is possible. Under this scenario, the only result in the NCHC that matters is UMD losing both games. BC also needs to make the HEA final, and Providence can't win the autobid, while the ECAC must go Harvard over Dartmouth in the final. The Big Ten has to go to Michigan, but their opponent is likely irrelevant.

*I managed to find a few exceptions: NCHC - 1 DU, 2 UND, 3 SCSU; Harvard over Dartmouth, Michigan over Minnesota, BC winning HEA title, Air Force in the AHA title game. There are multiple results where MTU finishes tied with Cornell in RPI to 4 decimals, but was listed after Cornell and at least one example where it was a similar tie with UMD.

Loss in Championship Game

If Michigan Tech manages to make the championship game, everything will come down to how many of the following events happen:

  • Minnesota-Duluth wins a game (and under certain scenarios ties the 3rd-place game)
  • Northeastern wins its HEA semifinal match
  • Michigan doesn't win the Big Ten tournament
  • Dartmouth or St. Lawrence wins the ECAC tournament

If more than one of these things happen, Michigan Tech is likely eliminated. Of note, MTU can survive one of those if UMD ties the third place game under certain circumstances. But in general, the Huskies' best bet is simply to win the WCHA tournament for the first time since 1965 (the format has changed many times, and from 1966 to 1981, no true champion was determined), and first time ever in the current format (first round total goal/best of three series followed by a single elimination tournament involving at least 4 teams).