In a recent post, I outlined some of the information I found about best-of-seven Stanley Cup Playoff series from 1942-2014, and which of the seven games have statistically been the most important. I suggested some alternative formats, which would actually allow teams with home ice advantage to host more games in a series than their opponents.
In order to find that information, I looked at the winner of each series listed here and entered into a spreadsheet whether the winner of the series won Games 1-7. It looked like this:
That effort also allowed me to uncover information about the sequences of best-of-seven series. For example, I learned that there are 35 different combinations for a team to win four games out of seven. These are the combinations, where a 1= Win and 0 = Loss (for calculation purposes, series that ended in fewer than seven games had to have zeroes tacked on for the remaining games. So ignore any zeroes after the fourth “1.”
As you can see, I assigned an ID number to each possible sequence. Then, I charted the frequency of outcomes. Again, ignore any “0” after the fourth “1.”
OVERALL
What we see here is that a sweep (ID number 1, W-W-W-W, again, ignoring any zeroes after the fourth “1”) is by far the most frequent occurrence for a best-of-seven series. A sweep has happened 111 times out of 615 series, or 18% of the time. The next four likeliest outcomes are series that go five games.
FIVE GAMES
There are only four ways in which a series can end in five games, ID numbers 2-5 (broken record alert, sorry: we’re ignoring any zeroes that come after the fourth “1.”) ID number 5, which represents the series-winning team taking the first three games, losing the fourth, and then closing the series out in Game 5 (W-W-W-L-W) is the most common result among series that end in five, and the second most common result overall, behind a sweep.
SIX GAMES
The most common six-game result (and the fourth most common result overall, behind a sweep and ID numbers 4 and 5) is ID number 13, which is a sequence of W-W-L-L-W-W. It’s by far the most common result for six-game series, having happened 36 times. None of the other six-game sequences come even close to the frequency of the five-game sequences, although, as I pointed out here, with the average series getting longer over time, it’s possible that fact will change. By far the least common six-game sequence is ID number 15, which represents a team taking a 3-0 series lead, losing the next two, and then winning Game 6 (W-W-W-L-L-W). That’s only happened 10 times since 1942, whereas every other six-game sequence has occurred at least 16 times.
SEVEN GAMES
Series that go seven games are by far the rarest—only 151 (24.5%) of all best-of-seven series have actually gone the distance. We see that ID numbers 21 (L-W-L-W-L-W-W), 22 (L-W-L-W-W-L-W) and 33 (W-W-L-L-W-L-W) are the most common sequences for series that go seven. We know how rare it is for a team to go down 0-3 and then win the next four, as the Kings did to the Sharks last year. It’s only happened four times in history (the 1942 Maple Leafs over the Red Wings, the 1975 Islanders over the Penguins, and the 2010 Flyers over the Bruins are the others). However, that sequence of events is not the rarest in hockey history.
Another observation: obviously, trailing a series 3-1 is difficult to overcome, obviously, but not impossible. On the other hand, if the team down 3-1 had also been down 2-0, it’s in even deeper trouble. Teams have overcome a 3-1 series deficit 25 times since 1942. However, teams that were down 2-0 in the series only account for nine of those occurrences, which is fewer than we’d expect.
ID numbers 34 (W-W-L-W-L-L-W) and 35 (W-W-W-L-L-L-W) also have happened only four times since 1942. And, checking in as the rarest sequence in Stanley Cup Playoff history since 1942, ID number 27, the elusive W-L-L-W-L-W-W. The most recent example was 2012’s New Jersey over Florida seven game series:
Now, is this information predictive? It’s difficult to tell because, although I looked at 615 series, and that seems like a lot, the numbers could be skewed by random variance. We could see three straight of ID number 27, and although it would be surprising, we shouldn’t be shocked because hockey is a fluky game and past results don’t always predict future results. But if nothing else, these numbers are fun to play with and just might give us a better idea of what we’re really seeing when we watch the playoffs evolve.
(Another stick tap to my buddy Elliott for his excellent work with spreadsheets. Any time I implied I performed some function other than coming up with the idea for the project or simply entering data into Excel, it was actually he who did it).
