One of the fun things about creating a win expectancy model is that you can retroactively go back and remember the crazy times. Often, this sort of statistical analysis is meant to unearth some nugget of insight…this post is about “hey, you remember when…”

If you are a fan of these teams, then if there is a game that stands out in your memory, it’s likely on this list. Because these are the games that objectively fall under the heading of: insane comeback. If you have your own memories of these games, let us know in the comments.

To get to this ranked list, we went through every game in our database, calculated the win probability of both teams after every recorded play, and then compared those chances against the final outcome. The games that we flagged as interesting fit one main criteria: at some point in the second half, the winning team had a very slim chance of winning the game

Without further ado, a selection of the greatest comebacks of the past two years:

#4 – April 19, 2015 – Binghamton @ Stony Brook – Stony Brook wins 14-13

Our granddaddy of all comebacks, this game actually started out well for Stony Brook. They scored just 3 minutes into the game to take a 1-0 lead. After that, it was not a particularly stellar 35 minutes. Binghamton went on 4 goal run to end the first quarter. 4-1 heading to the 2^nd, the game was actually pretty even; Binghamton scored 3 of the first 4 goals to push the lead to 7-2, but then Stony Brook closed it to 8-4 at halftime. After half, Binghamton scored the first two goals, making it 10-4 with 10:59 remaining in the 3^rd. Stony Brook scored to make it 10-5, but after some back and forth, they hit their lowest win odds at just 17.4% with 7:16 left in the 3^rd after an SBU shot was saved and Binghamton was able to clear.

Stony Brook was able to create a turnover, and they scored not long after to make it 10-6. From there, they ran off 4 straight goals to make it 10-10 by the end of the period. In the 4th, as they so often do, the comeback lost a little steam at that point, and Binghamton was able to retake the lead. Stony Brook tied it up again and eventually took a lead 5 minutes to the 4th quarter. Eventually Binghamton tied it up, the game went to overtime, and Stony Brook was able to get the winner 3:10 into the extra period.

#3 – March 8, 2016 – Robert Morris @ Mercer – Robert Morris wins 11 – 10

An away team triumphs! It may be coincidence that all the games on this list (except this one) resulted in eventual home team victories. Perhaps the home team is complacent coming out, which allows them to eventually assert their surperiority? Maybe the away team expends too much energy racking up their lead. Compound this with the energy required in travelling and they run out of steam? Either way, Robert Morris is the lone visitor to crack this list.

They actually were able to build a three goal lead with 5 minutes to go in the 2^nd quarter, so this was not a case of the eventual winner coming out flat. But over the next 20 minutes, Mercer asserted themselves with a 6-0 run to take their own 3 goal lead seconds into the 4^th quarter. This left Robert Morris with just a 16.4% chance of pulling out the victory.

Then RMU went on their own 4-0 run over a period of 10 minutes to take a one goal lead with about 6 minutes left. Mercer eventually tied it up, but RMU was able to pull it out in overtime. It’s always amazing when a game of runs like this ends up equalizing, resulting in an overtime period.

#2 – March 8, 2016 – St. John’s @ Yale – Yale wins 13 – 8

Can someone check whether March 8, 2016 was a full moon or something.  I mean, what are the odds that #2 & #3 happened on the exact same day.  Date shenanigans aside, the thing that stands out about this game is the final margin: 5 goals. You would not expect that a team that won a game by 5 goals would have ever really been in much danger, let alone enough danger to appear on this list. But that is exactly where Yale found themselves with 4 minutes gone in the 3^rd quarter, down 6 goals, 7 to 1. Over the first 34 minutes, they really didn’t have anything good happen (with the exception of one measly goal). This left them with only a 16.1% chance of winning with 8:45 left in the 3^rd after a series of face-off wins and offensive activity by St. John’s.

But of course, they did win. They won courtesy of a 12 – 1 run over the final 25 minutes of the game. This game definitely takes the cake in terms of symmetry. Both teams had huge runs (7-1 for St. John’s and 12 – 1 for Yale), but fortunately for the home team, their run was longer. I will have to do another post on this, but the 11 goal margin in just 25 minutes is probably some sort of record as well.

# 1 – February 6, 2016 – Vermont @ Jacksonville – Jacksonville wins 17 – 16

This was a high scoring affair, which is always going to lead to more volatility in the win expectancies. But it does take the cake as far as a deep hole, late in the game.

With 3:36 left in the 3^rd quarter, Jacksonville had just given up a goal and Vermont controlled the following face-off. At this point, Jacksonville’s win odds fell to a paltry 16.1%. They were only down by 3 goals, but because it was so late in the game, our model was pretty down on their chances at that point.

One thing that really helped Jacksonville in making their comeback was penalties; 4 60-second penalties for Vermont from that point on. They, in part, helped Jacksonville go on a pivotal run flipping the score from 14-12 Vermont at 3:36 left in the 3^rd to 16-14 Jacksonville with 10:27 left in the 4^th.