While recently conducting research for a chapter in my book I found myself investigating the post-season page of Baseball-Reference to see whether the home or road team won more often in the playoffs. The research was more in-depth than that sounds, though, as I recorded the results for each individual game of each individual series through the years. As in, I recorded how often a home team won Game 3 of the Divisional series, and so forth. This way we can see not only if teams really benefit from home field advantage but also when they benefit most and least.
In order to do this I had to look at separate sets of criterion for each. The divisional series did not come into play (other than 1981) until 1995, with the institution of the Wild Card. Therefore, results for the Divisional Series go from 1995-2007. The Championship Series went all the way back to 1969, however from 1969-1984 the series were best of five. From 1985-2007 the Championship Series has been a best of seven. With that in mind my results for the Championship Series stem from 1985-2007. We could go all the back to 1969 but I want to keep everything constant.
The World Series goes all the way back to 1884, however from 1884-1902 all playoff games were considered exhibition games. Additionally, from 1903-1922 there were World Series games that resulted in ties, which cannot be included in this study. So, the periods used for my results are:

• Divisional Series: 1995-2007
• Championship Series: 1985-2007
• World Series: 1923-2007

Overall Results
Without looking at the individual games within each series the results for home field advantage for each overall series are:

• Divisional Series: Home – 55.5%, Road – 44.5%
• Championship Series: Home – 52.5%, Road – 47.5%
• World Series: Home – 55.9%, Road – 44.1%

Just looking at these overall results we can see that the Divisional Series and World Series have a higher home field advantage than the Championship Series.
Game Results – Divisional Series
Below are the individual percentages for games one through five of the Divisional Series, from 1995-2007. In parenthesis are the total number of games recorded. There were exactly 200 games in the Divisional Series from 1995-2007.

• Game One (52): Home – 50.0%, Road – 50.0%
• Game Two (52): Home – 63.5%, Road – 36.5%
• Game Three (52): Home – 48.1%, Road – 51.9%
• Game Four (31): Home – 54.8%, Road – 45.2%
• Game Five (13): Home – 53.8%, Road – 46.2%

From 1995-2007, both the home and road teams have literally had a 50/50 shot at winning game one. Game Five looks like more of a significant discrepancy than it is considering there have only been 13 instances. Even though the percentage difference is nearly 8%, the results are only 7-6 in favor of the Home Team. What really interested me was the discrepancies between Game Two and Game Three. Home Teams have won almost 2/3 of Game Two’s but are under 1/2 of Game Three’s. I will have to look at how many of these Game Three’s were with the series tied as opposed to a team being up 2-0 since this seems to directly correlate with Divisional Series sweeps by teams that began the series at home.
Game Results – Championship Series
Below are the individual percentages for games one through seven of the Championship Series, from 1985-2007. In parenthesis are the total number of games recorded. There were 255 games in the Championship Series from 1985-2007.

• Game One (44): Home – 54.5%, Road – 45.5%
• Game Two (44): Home – 43.2%, Road – 56.8%
• Game Three (44): Home – 61.4%, Road – 38.6%
• Game Four (44): Home – 50.0%, Road – 50.0%
• Game Five (39): Home – 53.8%, Road – 46.2%
• Game Six (27): Home – 44.4%, Road – 55.6%
• Game Seven (13): Home – 69.2%, Road – 30.8%

Game Seven has an advantage for Home Teams but, again, there have only been 13 instances. This time the results were 9-4 in favor of the home team. Though more in favor of the Home Team than the 7-6 of Div. Series Game Five there have not been many instances.
Game Results – World Series
Below are the individual percentages for games one through seven of the World Series, from 1923-2007. In parenthesis are the total number of games recorded. There were 487 games in the World Series from 1923-2007.

• Game One (84): Home – 61.9%, Road – 38.1%
• Game Two (84): Home – 58.3%, Road – 41.7%
• Game Three (84): Home – 54.8%, Road – 45.2%
• Game Four (84): Home – 50.0%, Road – 50.0%
• Game Five (67): Home – 50.7%, Road, 49.3%
• Game Six (51): Home – 62.7%, Road – 37.3%
• Game Seven (33): Home – 51.5%, Road – 48.5%

The really interesting part of this World Series data is how the home field advantage basically decreases until Game Six. The Home Team has had less of an advantage as the games progress before jumping up in Game Six. Game Seven then reverts back to the 50/50 idea. Apparently, Games Four, Five, and Seven are up for grabs regardless of the home or road situation.
Analysis
There are nineteen types of games (5 Div, 7 Champ, 7 WS) and seven of them present basically an equal split between home and road. The biggest 50/50 split, in terms of individual games and not the overall series, comes in the Divisional Series, where Game One, Game Three, and Game Five are pretty much 50/50. The World Series also brings with it three games that are relatively 50/50 and the Championship Series has only one 50/50 game (Game Four).
The ironic part of that is the fact that the Championship Series, overall, has the least home field advantage of the three, despite having only one 50/50 game. The other games in the Championship Series are balanced out by some leaning toward home and other toward road, rather than many being even. There will be those who argue that home field advantage does not matter, no matter what the All-Star Game says, but statistically it has.
Of the nineteen types of games, the only ones wherein the Road Team has a clear advantage are Game Two and Game Six of the Championship Series. That leaves us with two games favoring the Road Team, seven games being pretty much 50/50, and ten games favoring the Home Team.

1. Pizza Cutter says:

Eric, you can test to see whether the difference between the numbers is actually statistically significantly different from 50/50. Are you familiar with binomial distributions?

2. You mean like probability mass functions?

3. Pizza Cutter says:

It’s related to pmf, but this is more of a hypothesis testing method. This site does the heavy lifting for you.
http://faculty.vassar.edu/lowry/binomialX.html

4. Okay, so if we look at all of these games in terms of Binomial Probability, we get the following results for the home team.
Div. G1 = 0
Div. G2 = +1.80
Div. G3 = -0.14
Div. G4 = +0.36
Div. G5 = 0
Ch. G1 = +0.45
Ch. G2 = -0.75
Ch. G3 = +1.36
Ch. G4 = 0
Ch. G5 = +0.32
Ch. G6 = -0.38
Ch. G7 = +1.11
WS. G1 = +2.07
WS. G2 = +1.42
WS. G3 = +0.76
WS. G4 = 0
WS. G5 = 0
WS. G6 = +1.68
WS. G7 = 0

5. Pizza Cutter says:

So that says that the only one that reaches statistical significance is WS G1. There’s also the issue that lately, HFA has been assigned to the team with the better record (although in the WS, it alternated for years, then became tied to the ASG), so we would expect something other than 50/50. Hmmm…

6. CT_Steve says:

Isn’t some of the analysis tempered by the fact that the team with the better record gets the home field in the DS and the CS. That is, if that team wins more it might be because it’s the better team, not because of where the game is.
The comparison makes more sense in the World Series, where home-field is determined by some irrelevant factor.

7. Steve, what you brought up is the whole idea behind the article. Is home field advantage really that great? The idea was to explore if HFA really does help. It might be becasue they are a better team or it could be because they were aided by HFA.
With better records, it could revolve around your division, too, and not the quality of your team. I would have to imagine it is tougher to play 18-19 games against the Dodgers, Padres, Rockies, and DBacks, than against the Pirates, Astros, Reds, and (last year) Cardinals.
As Pizza Cutter said, though, and as MGL mentioned over at the Inside the Book Blog, only World Series Game 1 is statistically significant as being in favor of the home team. Next up would be Game 2 of the Divisional Series and Game 6 of the World Series.