The Unlikeliest Wins in History

What struck me while watching the Oilers defeat the Blue Jackets 3-1 this past Sunday was how unlikely the win was, considering the shot clock. The Jackets outshot the Oilers by a tally of 40-14, or 40-13 if you take out the empty net goal. Devan Dubnyk was completely barraged by one of the worst teams in the league, facing an increasing amount of rubber as the game went on, including a ton of work on the penalty kill. It was a minor miracle that the Oilers left Ohio with 2 points that night. But how unlikely was this win?

I'm sure there are many ways to define what an unlikely win is: a) a 30th place team beating a 1st place team, b) a team managing to win through many injuries, c) a team winning their 3rd game in 4 nights, d) any Edmonton Oiler win in calendar year 2010, etc. However, I wanted to research how many games there had been since the 2004-05 lockout where a team won a game in regulation while posting less than 15 non-empty net shots for. This will form the basis of my definition for the term 'unlikely'.

It turns out that only 35 such wins have been recorded in the 8800 games played since that time (as of this writing), or approximately 0.4%. In other words, one out of every 250 games played in the NHL in the last 8 seasons has featured a win with less than 15 shots for. There have only been 15 games with less than 14 non-empty net shots for in a regulation win, or one game out of every 587 played -- meaning your team could play 7 entire seasons and still not see a win like Dubnyk posted on Sunday (on average).

Here's a quick chart with some summaries of these 35 unlikely wins:

Which goalie has posted the most unlikely wins? Why, it's Oiler MVP Nikolai Khabibulin, who has posted 3 such wins since 2005 (2 with the Oilers, 1 with the Black Hawks). Only 4 other goalies have won multiple games, meaning the feat is extraordinarily rare to replicate.

The team which has won the most unlikely games is the Dallas Stars, followed by a group of teams at 3, including your home squad. Which team has been victimized the most by unlikely wins -- allowing a tiny amount of shots against while still losing the game? It's the New Jersey Devils, which seems unlikely considering the Hall of Fame quality goaltending in net (har har). Columbus has been victimized thrice (those poor things), as have the Godless heathens down Highway 2 in Calgary.

Interestingly, only 10 of the 35 unlikely wins (or 28.6%) have been posted at home. A thorough person would check to see how likely a team is to get outshot on the road in general, but I am certainly not that person.

So what are THE most unlikeliest wins out of this bunch? I'm going to find the ratio of shots against to shots for to get a sense of the weight of the victory: someone winning while getting outshot 50-14 is a more unlikely win than someone getting outshot 15-11 and still winning. Here are the top ten most unlikely wins:

So you can see that Dubnyk's win on Sunday was tied for the 4th most unlikely since 2004, with a ratio of 3.1 shots against for every shot for. The most impressive win was a little over 3 years ago, when Jonas Hiller stymied the Black Hawks 3 to 1 while getting outshot 43 to 11 (non-empty net shots).

Just when you think the Oilers aren't really accomplishing much, they turn around and do something historic! What a lucky fanbase we are!