# Study of Home Ice Advantage - Stanley Cup Playoffs

1. ### Kyle McMahonRegistered User

Joined:
May 10, 2006
Messages:
11,599
2,171
Trophy Points:
196
SB Cash:
\$ 200,000
Gender:
Male
Occupation:
Sports Investor
Location:
Evil Empire
Hey folks,

I recently ran a study on home ice advantage in the playoffs dating back to 2006. I posted it to a new website (link in my bio) as a blog-style article with additional info related to sports betting. I figure stat junkies might be interested in this whether or not they are gamblers, so here are the data tables and the non-gambling related commentary. All feedback/suggestions, positive or negative, is more than welcome.

There is an inherent advantage to playing at home versus playing on the road in all sports, and the NHL is no different. We know the team with home ice advantage is more likely to win a playoff series. But is this simply due to the fact that they're a better team, as indicated by the standings? Is the advantage limited to the mere fact that a team is more likely to win a home game and just so happens to get an extra one should the series end in five or seven games? What sort of edge does home ice provide a team in a single game? These are all questions I was curious to know the answer to, so I conducted an analysis of each individual game within a playoff series. The sample chosen to start with is every first round playoff series dating back to the beginning of the modern salary cap era in 2006.

This table is presented from the vantage point of the home team for each individual game. The higher-seeded team is the home team for Games 1, 2, 5, and 7. The lower seed is the home team in Games 3, 4, and 6. I have also included goal differential, both raw and per game. The winning percentage of the series favorite (who is not the home team in three of the games) has also been included for ease of viewing.

Legend: Gm=game of series. W=home team wins. L=home team losses. Home%=home team winning%. Fav%=series favorite winning%. HomeGD=home team goal differential. HGD/Gm=home team goal differential per game. FavGD=series favorite goal differential. FGD/Gm=series favorite goal differential per game.

First Round Playoff Series, 2006-2018

 Gm W L Home% Fav% HomeGD GD/Gm FavGD GD/Gm 1 63 41 0.606 0.606 60 0.58 60 0.58 2 57 47 0.548 0.548 37 0.36 37 0.36 3 48 56 0.462 0.538 -1 -0.01 1 0.01 4 47 57 0.452 0.548 -29 -0.28 29 0.28 5 49 41 0.544 0.544 27 0.30 27 0.30 6 34 32 0.515 0.485 4 0.06 -4 -0.06 7 14 13 0.518 0.518 0 0.00 0 0.00
Contrary to what I anticipated, it seems that almost the entirety of the advantage provided by home ice is realized in the opening game of the series. Home teams (and by definition, favorites) have by far their best showing in the very first game, both in terms of raw wins and goal differential. Their advantage in games 2 and 5 drops by abut 10%, and by the time you reach Game 7, it is basically a coin flip. Overall, the team favored to win a series posts a nearly identical winning percentage in Games 2 through 5, even though two of those games take place on the road. Game 6 is the only instance where the series underdog is actually more likely than not to win an individual game. When broken down further into goal differential, the favorite does seem to perform better in their second road game, much as the underdog does in theirs.

The next question you may be wondering is whether or not these trends remain intact as the playoffs progress. Here are the same tables presented for the 2nd round of the playoffs, as well as the 3rd round/final. I combined the last two rounds as the sample size would too small otherwise.

Second Round Playoff Series, 2006-2018

 Gm W L Home% Fav% HomeGD GD/Gm FavGD GD/Gm 1 32 20 0.615 0.615 37 0.71 37 0.71 2 30 22 0.577 0.577 11 0.21 11 0.21 3 29 23 0.558 0.442 16 0.31 -16 -0.31 4 30 22 0.577 0.423 17 0.33 -17 -0.33 5 32 14 0.696 0.696 30 0.65 30 0.65 6 17 15 0.531 0.469 4 0.13 -4 -0.13 7 8 10 0.444 0.444 -15 -0.83 -15 -0.83
Third Round/Final Series, 2006-2018

 Gm W L Home% Fav% HomeGD GD/Gm FavGD GD/Gm 1 25 14 0.641 0.641 19 0.49 19 0.49 2 22 17 0.564 0.564 1 0.03 1 0.03 3 27 12 0.692 0.308 23 0.59 -23 -0.59 4 19 20 0.487 0.513 -5 -0.13 5 0.13 5 20 16 0.556 0.556 28 0.78 28 0.78 6 15 9 0.625 0.375 15 0.63 -15 -0.63 7 5 6 0.455 0.455 -7 -0.64 -7 -0.64
Overall Results for All Playoff Series, 2006-2018

 Gm W L Home% Fav% HomeGD GD/Gm FavGD GD/Gm 1 120 75 0.615 0.615 116 0.59 116 0.59 2 109 86 0.559 0.559 49 0.25 49 0.25 3 104 91 0.533 0.467 38 0.19 -38 -0.19 4 96 99 0.492 0.508 -17 -0.09 17 0.09 5 101 71 0.587 0.587 85 0.49 85 0.49 6 66 56 0.541 0.459 23 0.19 -23 -0.19 7 27 29 0.482 0.482 -22 -0.39 -22 -0.39
When we look at the later round results, we can see that overall Game 1 remains the most advantageous in favor of the home team. The effect increases as the playoffs roll on in terms of winning percentage. When goal differential is examined the results are less dramatic, though by the third round of the playoffs we are generally dealing with the strongest teams and closer scores can be expected.

Like the first round data, the home ice advantage diminishes quite significantly for Game 2. This is true of both winning percentage and goal differential. From the perspective of the underdog, home ice appears to provide much more of an edge in the later rounds. However, the weakest playoff entrants tend to be eliminated in the first round, and these one-and-done teams naturally provide half of the sample size. The over-arching trend that emerges is that road teams fare better in their second road game than their first. While certain pockets of the data will naturally show greater discrepancies than others, this can likely be attributed to variance. The overall numbers indicate that the road team performs about one-third of a goal per game better in their second matchup compared to their first.

Continuing on to the latter games of a series, we see that Game 5 tends to fall in between Game 1 and 2 in terms of home ice impact. From a goal differential standpoint, the effect definitely increases later on into the playoffs, though the sample size diminishes considerably by the third round/finals (just 39 games). We observe a similar trend for Game 6; the home team tends to be more dominant in those situations towards the end of the playoffs.

This leaves us with Game 7. It is often said by commentators that "this is where the hard work in the regular season pays offâ€¦Team X gets to be the host for the deciding game in the series" or something to that effect. But in apparent opposition to popular belief, it seems that by the time a series reaches the maximum number of games, home ice advantage no longer exists. A series has gone the distance 56 times in the salary-cap era, and in every pocket of data as well as the overall sample, Game 7 is a virtual coin flip. From the perspective of the series favorite, it is by far the least advantageous situation in which to have home ice. When we expand our examination to goal differential, the outlook is even more bleak for the home team.

Edit: First line should have read "dating back to 2006", as I'm sure most people could guess after reading through.

Last edited: Apr 10, 2019
2. ### Puck DoggPuck life

Joined:
Mar 13, 2006
Messages:
1,271
104
Trophy Points:
112
SB Cash:
\$ 200,000
Interesting read, especially that the favorite is likely to be strongest in game 1. My belief was that underdog can pull off a surprise in first game, but these findings show exact opposite. Again, statistical proof shows that common sense and hunch is not something to rely on.

The info I'd like to see in addition is how games are won on regulation vs overtime (although some conclusions can be drawn from the goal difference) and which games are high and low- scoring ones, and for example does the amount of goals decrease as the series advance.

Kyle McMahon likes this.
3. ### Doctor NoRegistered User

Joined:
Oct 26, 2005
Messages:
8,415
2,254
Trophy Points:
166
SB Cash:
\$ 200,000
Interesting analysis.

One thing to consider are mix issues - my guess is that a partial explanation for the home teams' Game One performance relative to (say) Game Seven is that all underdogs play Game One, but if you even make it to Game Seven you've shown that you're a decent opponent.

Stated differently, the Game Seven record may be different if the teams that were blown out in Game One got to play in Game Seven as well.

4. ### Kyle McMahonRegistered User

Joined:
May 10, 2006
Messages:
11,599
2,171
Trophy Points:
196
SB Cash:
\$ 200,000
Gender:
Male
Occupation:
Sports Investor
Location:
Evil Empire
Yes, I've also wondered if certain games are more predisposed to higher or lower offense than others.

There would be a couple factors to consider though. One is empty net goals. It would be quite time consuming to try and remove these from the data unless somebody has already done the legwork. Fortunately they are rare enough that it likely wouldn't corrupt the data overall in a large sample size.

The other consideration would be overtime goals. A 4-goal game is just that if the score is 4-0 or 3-1. But 2-2 assures a 5th. Fortunately it's easy to make note of these goals.

I might have a look at this when time permits.

5. ### Kyle McMahonRegistered User

Joined:
May 10, 2006
Messages:
11,599
2,171
Trophy Points:
196
SB Cash:
\$ 200,000
Gender:
Male
Occupation:
Sports Investor
Location:
Evil Empire
You raise good points. Something I might take a look at is the average regular season points difference between teams that reach a Game 7 versus the global average. If it's typically closely matched teams that go to Game 7, it makes sense that the results would be as they are. I have a hunch that the chances of a series reaching Game 7 isn't correlated to the gap in team strength in a linear fashion, but I'd like to see the results.

Canadiens1958 and Doctor No like this.

Joined:
Aug 2, 2005
Messages:
4,325