This post explains the data in the PWR tournament possibilities page. You can follow along there if you’d like to browse the raw data yourself:

NCAA Tournament Participation Probabilities

Who’s in?

The following 10 teams are guaranteed an invitation to the NCAA tournament on the basis of their PWR.

• #1 St Cloud State
• #2 Massachusetts
• #3 Minnesota St
• #4 Minnesota-Duluth
• #5 Quinnipiac
• #6 Denver
• #7 Ohio State
• #8 Northeastern
• #9 Clarkson
• #10 Arizona St

Who else can get in?

The winner of each conference tournament can get in:

Atlantic Hockey

• American International (*)
• Robert Morris (*)
• Niagara (*)
• RIT (*)

Big Ten

• Notre Dame (*)
• Penn St (*)

Hockey East

• Massachusetts (in)
• Boston College (*)
• Boston University (*)
• Northeastern (in)

ECAC

• Cornell
• Brown (*)
• Harvard
• Clarkson (in)

NCHC

• St Cloud St (in)
• Colorado College (*)
• Denver (in)
• Minnesota Duluth (in)

WCHA

• Minnesota St (in)
• Bowling Green

Those marked with a (*) can only get in if they win their conference tournament. Those marked with (in) will make the tournament regardless of conference outcome. But, what about the rest?

There are 16 spots in the NCAA tournament. 10 participants are already known (see above). The winners of the Atlantic Hockey and Big Ten tournaments are not only guaranteed spots, but are the only teams from those conference tournaments that will get spots, so you can think of those two spots as reserved for one of the 4 AH teams and one of the 2 AH teams, respectively.

That leaves 4 spots total for a combination of the champions of the remaining 4 conferences and the 5 teams that can make it at-large (there is some overlap between those two groups). How many teams are selected at-large depends how many of the 4 remaining slots are taken by teams that aren’t already in by virtue of their PWR. For each of Hockey East, ECAC, NCHC, and WCHA that are won by a team that is already counted as being in, an additional spot is available for a different team to make it at-large.

The following teams can make it at-large:

• Cornell (64% of scenarios in which they win 0, 97% in which they win 1)
• Harvard (81% of scenarios in which they win 0, 94% in which they win 1)
• Bowling Green (29% of scenarios in which they lose)
• Providence (not playing, 65% of scenarios)
• Union (not playing, <.1% of scenarios)

With some conference tournaments beginning this weekend, the NCAA tournament outlook for most teams is pretty well known. But, a few teams in the middle are jockeying for bubble position, hoping to get an at-large bid should their conference tournament runs fall short. This article looks at what is likely to happen in the PWR rankings this weekend.

I’ll pull out the highlights, but you can follow along in the table that shows the likely outcomes for all teams:

Wins needed to likely have PWR rank on March 11, 2019

#1 St Cloud is almost certain to still be #1 next Monday, no matter what happens this weekend.

#2 Massachusetts through #7 Ohio State should still be the same set of teams, but perhaps reshuffled based on outcomes (even the teams not playing, #5 Quinnipiac and #7 Ohio State, could move a bit).

Those teams include:
#2 Massachusetts
#3 Minnesota-Duluth
#4 Minnesota St
#5 Quinnipiac
#6 Denver
#7 Ohio State

#8-#20 are fighting for bubble position and (where applicable) hoping to move on for additional opportunities to improve their PWR and at-large bid likelihood.

Those teams include:
#8 Arizona St
#9 Providence
#10 Clarkson
#11 Northeastern
#12 Cornell
#13 Harvard
#14 Western Michigan
#15 Notre Dame
#16 Penn St
#17 Bowling Green
#18 Mass.-Lowell
#19 Union
#20 North Dakota

#21 Minnesota and below are almost certain to remain outside the top 16, regardless of this weekend’s outcome, so facing long shots to make the NCAA tournament at large.

With just one weekend of play remaining until the first conference tournament begins, we know a lot more about who is likely to get an at-large bid to the NCAA tournament than we did even just a couple weeks ago.

Follow along, if you’d like:
Wins needed to likely end regular season at PWR rank

Remember that any team that wins its conference tournament gets an autobid. Because we don’t know how many tournament slots will go to autobids that wouldn’t have made the tournament at-large, we can’t be sure what PWR ranking will be required to make the tournament at-large, but it’s likely to be in the 13-14 range. So, we’re primarily looking for the likelihood of teams that don’t win their conference tournaments finishing around #13-14 (or higher), but will call anything in the 12-15 range “the bubble” because of the uncertainty.

Near locks

#1 St Cloud State through #8 Arizona State are probably in, no matter how they perform. Note that there are some unlikely scenarios in which teams like #7 Ohio State, #5 Denver, and even #3 Minnesota-Duluth can fall to the teens by the end of the regular season, which would open the door for missing the tournament with a subsequently poor performance in the conference tournament and some extremely bad luck. But, that’s unlikely.

#8 Arizona State is actually a bit safer than those teams by virtue of not having a conference tournament in which they could perform poorly and push themselves lower.

Those teams include:
#1 St Cloud St
#2 Massachusetts
#3 Minnesota-Duluth
#4 Minnesota St
#5 Denver
#6 Quinnipiac
#7 Ohio St
#8 Arizona St

Likely in control of their own destiny

#9 Northeastern through #14 Harvard can come close to locking up an at-large bid with a strong finish, but would still be on the bubble even with a disastrous finish. Then, it would come down to conference tournament performance and quite a bit of luck around other teams’ performances to determine who gets the at-large bids.

Those teams include:
#9 Northeastern
#10 Providence
#11 Western Michigan
#12 Clarkson
#13 Cornell
#14 Harvard

I debated a bit whether to include Harvard in this list or the next, they’re right on the margin. There are scenarios, though not the most likely, in which even with a pair of wins Harvard could end the regular season around #13.

At-large bids are possible

#15 Bowling Green through #19 Penn State could make the tournament at-large with a near perfect close to the season, a decent conference tournament performance, and some luck.

Those teams include:
#15 Bowling Green
#16 Mass.-Lowell
#17 Notre Dame
#18 North Dakota
#19 Penn State

Probably need to win their conference tournament

None of #20 Lake Superior, or below, can end the regular season in the top 16. So, it would take an improbably strong conference tournament performance and some luck to get an at-large bid (remember that winning the conference tournament guarantees an autobid, so the at-large bid would only come into play with a loss in the conference tournament, which would likely also have pushed down the PWR ranking a bit).

Who doesn’t find themself looking at the PairWise Ranking (PWR) each week? (Ok, very few of us do, but I have a feeling you might be among the select few). While a lot of the tools and analysis here focus on the end of season and who’s going to make the NCAA tournament, it’s also sometimes fun to look at the week-to-week ebb and flow.

You can always use the PWR calculator to find out for yourself what would happen with just a few games’ results determined, e.g. what would St. Cloud St’s PWR be if its weekend results were already in. But, using the same simulations I use for end-of-season projections, focusing on just one week’s results, can provide a more nuanced answer to the same question. If the calculator tells you “what would St. Cloud St’s PWR be with nothing changed but this weekend’s results being set”, the simulations tell you “what is St. Cloud St’s PWR likely to be after all teams’ games this weekend are completed?” Got it?

One more programming note–it’s not uncommon to see teams in the 40s have one week upside potential in the range of 10 spots. Those generally require an improbable sweep of a much higher ranked team, and even if completed, don’t result in the team being anywhere near the bubble. I include a game as interesting only if it involves a team that has a notable chance of making the NCAAs at large.

Upside potential: #20 North Dakota

A stretch of tough competition provides PWR opportunities for #20 North Dakota. After the Fighting Hawks earned a much needed split against #1 St Cloud St last weekend, they are now at the bottom of a PWR hill going into a pair of games at #6 Denver this weekend. By the current rankings, UND “should” lose those road games at a ranked opponent, so any success will push the Hawks upward, while failure won’t result in much of a fall (though UND can’t afford to squander such an opportunity to climb at this point).

A pair of wins would most likely shoot UND into the #12-#16 range, with a break into the single digits unlikely, but mathematically possible. A split would most likely result in a modest climb to the #16-#19 range. Going into games you’re expected to lose results in very little downside risk, as even getting swept would likely leave the Hawks in the #19-#22 range.

Using the PWR Calculator, you can see that if no other games were played this weekend, a pair of wins would land UND at #12, a split at #17, and a pair of losses at #20.

Downside potential: #10 Providence, #9 Cornell

#10 Providence faces the opposite situation from UND, hosting #39 Vermont for a pair of games this weekend. The Friars are at the most treacherous PWR ranking, where RPIs start to become pretty compressed, numerous PWR comparisons are won by 1, and your resume doesn’t really look that different from the team 10 spots below you. Providence is supposed to win; a sweep won’t help their ranking, but underperforming could result in a sharp fall.

A pair of wins would most likely keep Providence in the #9-#12 range. A split would likely result in a modest fall to the #12-#16 range. Getting swept in a pair of games the Friars are supposed to win could be catastrophic, most likely resulting in a PWR anywhere in the #14-#20 range.

Using the PWR Calculator, you can see that if no other games were played this weekend, a pair of wins would land PC at #9, a split at #14, and a pair of losses at #17.

The narrative for #9 Cornell is the same as Providence’s, but with a pair of road games at #19 Union and #49 Rensselaer.

A pair of wins would most likely keep Cornell in the #7-#10 range. A split would likely result in a modest fall to the #9-#15 range. Getting swept in a pair of games they’re supposed to win would result in a fall for the Big Red, likely to anywhere in the #13-#19 range.

Using the PWR Calculator, you can see that if no other games were played this weekend, a pair of wins would land Cornell at #8, winning either but losing the other at #11, and a pair of losses at #17. Surprisingly, there doesn’t seem to be a big difference between losing to #19 Union or #49 Rensselaer.

How it works

Each forecast is based on at least one million monte carlo simulations of the games in the described period. For each simulation, the PairWise Ranking (PWR) is calculated and the results tallied. The probabilities presented in the forecasts are the share of simulations in which a particular outcome occurred.

The outcome of each game in each simulation is determined by random draw, with the probability of victory for each team set by their relative KRACH ratings. So, if the simulation set included a contest between team A with KRACH 300 and team B with KRACH 100, team A will win the game in very close to 75% of the simulations. I don’t simulate ties or home ice advantage.

The calculator uses the same PWR formula that mimics the NCAA’s tournament selection process that is used to produce the PWR tables on this and many other sites. But, in addition to running the formula against the games that have already been played, it lets you test scenarios by seeing what the PWR would be with different results for existing games, new results for not-yet-played games, or entirely fictitious games.

With most teams having just 10-12 regular season games remaining, we can draw some conclusions about who’s likely to make the NCAA tournament. In this article, I’ll look at the regular season forecasts (see First 2019 PWR forecast available for an explanation, if you’re not familiar) to discuss how teams need to perform in their remaining games to earn an NCAA tournament berth.

You can also always browse the weekly forecast yourself at PWR by wins forecast.

Who’s a lock?

No one is mathematically a lock at this point, but in 1,000,000 simulations of the rest of the regular season, there wasn’t a single one in which #1 St. Cloud St won fewer than 2 games or dropped below #13 PWR. Of course, those simulations include assumptions about the Huskies’ strength based on their performance so far, so it’s possible that an unprecedented collapse could result in a tournament miss. But, based on their performance to date, I’ll say St. Cloud St is the one team positioned to be assured a tournament berth.

Who’s in barring a collapse?

Even #2 Massachusetts doesn’t have quite as guaranteed a path. Winning just 4 games could result in a regular season PWR in the 10-15 range, which is right on the bubble of making an at-large NCAA tournament berth.

Down through about #7, Northeastern, teams will finish on or above the bubble by winning half their remaining games. Those teams include:

#2 Massachusetts
#3 Denver
#4 Minnesota-Duluth
#5 Ohio State
#6 Quinnipiac
#7 Northeastern

Who controls their own destiny?

From #8 Minnesota State down to #30 Wisconsin control their own destiny. That is, they can each end up on the bubble, or better, with a sufficiently good performance over the rest of the regular season. The lower the current ranking, the better a performance is required. For example, Wisconsin would really need to win out 11 or 12 of its remaining 12 games for the Badgers to be well positioned for an at-large berth.

Those teams include:

#8 Minnesota State
#9 Cornell
#10 Clarkson
#11 Arizona St
#12 Bowling Green
#13 Western Michigan
#14 Notre Dame
#15 Union
#16 Providence
#17 Harvard
#18 Penn St
#19 Mass.-Lowell
#20 Lake Superior St
#21 North Dakota
#22 Northern Michigan
#23 Yale
#24 Boston U
#25 Minnesota
#26 Michigan St
#27 Michigan
#28 New Hampshire
#29 Miami
#30 Wisconsin

Who won’t make the tournament at large?

From #31 Bemidji St down, there aren’t many possibilities of climbing to the bubble. A few seem to have the possibility of climbing to the bottom of the bubble, the 13-15 range, if they win out; but, even such a performance would still be pretty unlikely to result in an at-large tournament berth with a loss in the conference tournament. If you wanted to make a super long shot bet on one of these, #31 Bemidji St, #32 Dartmouth, #33 RIT, #34 Maine, #39 Boston College, and #41 Nebraska-Omaha stand the best chance.

But, all of these teams should really plan to win their conference tournament if they want to make the NCAAs:

#31 Bemidji St
#32 Dartmouth
#33 RIT
#34 Maine
#35 Michigan Tech
#36 Colorado College
#37 Air Force
#38 American International
#39 Boston College
#40 Princeton
#41 Nebraska-Omaha
#42 Niagara
#43 Vermont
#44 Mercyhurst
#45 Colgate
#46 Rensselaer
#47 Brown
#48 Connectictu
#49 Alaska
#50 Merrimack
#51 Army
#52 Canisius
#53 Bentley
#54 Robert Morris
#55 Sacred Heart
#56 Ferris St
#57 Alabama-Huntsville
#58 Holy Cross
#59 St. Lawrence
#60 Alaska Anchorage

What next?

I’ll keep updating the forecasts weekly, so you can always browse them yourself. I’ll also try to post interesting interpretations here, with increasing frequency as we near the end of the regular season. Meanwhile, you can explore more of the data yourself:

• PWR Rankings
• PWR Forecast