Streaking the Lawn -- SB Nation's Virginia site -- sets the scene:
Before fans could say Klöckner, the once eight goal #BEATEMDOWN had turned into a one goal deficit with 17 seconds to play.
It looked as if Virginia was going to cruise to a season-opening victory over Loyola in Charlottesville. For 45 minutes at Klockner, the Cavaliers surgically butchered the Greyhounds, keeping Loyola off the board for two periods and assembling consecutive four-goal quarters that built Virginia's 12-4 lead before the final period. There was no reason to believe that Loyola wouldn't leave Virginia with an ugly effort on its record; down eight goals and unable to stop the Cavaliers at the defensive end, the Greyhounds looked like toast.
And then Loyola ran off a nine-goal run and melted faces. Because reasons.
The Greyhounds' fourth quarter surge against the Cavaliers seemingly came out of nowhere. Looking at performance metrics1, Loyola was playing like a team set to have its teeth kicked and pushed through their toes:
|Pythagorean Win Expectation||6.62%||93.38%|
Based on the team's performances at that point in the game, if Loyola and Virginia played a bazillion games against each other at the same level of performance generated in the first three periods of play on Thursday night, the Cavaliers would be expected to win virtually 100 percent of the time. That's not just crushing souls; that's crushing souls and then crushing the memory of the souls that were crushed. Then 'Hounds exploded.
Tracking each team's Pythagorean win expectation and log5 values2 throughout the fourth quarter, Loyola's rush toward the lead -- and a potential win that nobody envisioned after three quarters of play -- looks ridiculous:
|14:37, Q4 (12-5, Virginia)|
|Pythagorean Win Expectation||10.70%||89.30%|
|14:14, Q4 (12-6, Virginia)|
|Pythagorean Win Expectation||15.25%||84.75%|
|13:11, Q4 (12-7, Virginia)|
|Pythagorean Win Expectation||19.96%||80.04%|
|12:39, Q4 (12-8, Virginia)|
|Pythagorean Win Expectation||24.62%||75.38%|
|11:26, Q4 (12-9, Virginia)|
|Pythagorean Win Expectation||29.07%||70.93%|
|8:42, Q4 (12-10, Virginia)|
|Pythagorean Win Expectation||36.23%||63.77%|
|6:50, Q4 (12-11, Virginia)|
|Pythagorean Win Expectation||43.30%||56.70%|
|2:30, Q4 (12-12)|
|Pythagorean Win Expectation||47.08%||52.92%|
|0:15, Q4 (13-12, Loyola)|
|Pythagorean Win Expectation||50.46%||49.54%|
Now, a win probability analysis potentially shows the Greyhounds at or close to a 100 percent favorite to win the game at the 0:15 mark. The building block metrics to an analysis of that nature, though, doesn't consider time or scenario or other variables. Those metrics only see Virginia as Loyola's equal (after, of course, the Greyhounds bootstrapped themselves up to the Cavaliers' level). The considered metrics reflect that at the end of regulation (and just a blip after Tucker's goal):
|Pythagorean Win Expectation||50.00%||50.00%|
That's an insane comeback from Loyola. It's just nuts. As late as the 6:50 mark of the fourth quarter, the team's performances on the field -- including Virginia's blitz of the 'Hounds through three periods and Loyola's blunderbuss effort through the fourth -- still implied that Virginia would earn a victory despite getting housed in their own building for almost 10 minutes. (At that point, though, the tide was pushing Loyola in the right direction.) Loyola was dead -- dead! -- and were *thisclose* to getting a win that it almost never should have had. It was a crazy night in Charlottesville, one where the teams weren't equal for extended periods of time but would have ended up that way save for the existence of overtime.
1 This isn't win probability. Win probability is a different kind of animal, but these metrics are building blocks to a win probability analysis.
2 Again, this isn't win probability. This focuses on how well a team is playing and their overall chances of victory against the same opponent if they played into infinity.