September 16, 2014

How the Home Team Wins

October 24, 2008 by · 10 Comments 

In baseball, the home field advantage is relatively small compared to professional football or basketball. In his study, Cyril Morong found that the home field advantage in baseball is around 7 – 8 % (see http://www.geocities.com/cyrilmorong@sbcglobal.net/HomeRoad.htm for the exact details).   That’s still significant enough to get me to wonder how it is that the home team had an advantage.   Notice I said HOW not WHY the home team has an advantage.  Why generally involves theories about getting to sleep in their own beds, or home cooking, or not having jet lag, or not being out late a night partying, etc.   What I want to know is HOW the home team has an advantage.  Is it because they walk more?  Is it because they hit more home runs, or it is more singles?    Knowing how should lead us to at least more intelligent guesses about the why.

To answer the how question, I used the Seamheads.com Parks Database (http://seamheads.com/db/databases.htm, then select Ballpark Stats Splits).  Looking at the data 1956-2007 (excluding 1999) revealed the following ratios between the home team and the away team (100 = Home and Away rates identical):

(Note:  Runs & AB based on per game, 1B/2B/3B/HR per AB, all others per est. PA’s)

Category Factor
Games Played 100
Runs 103
Est. Plate Appearances 96
At-Bats 95
Batting Average 103
Singles 102
Doubles 104
Triples 119
Home Runs 104
Sacrifices 109
Sacrifice Flies 107
Hit by Pitch 105
Walks 109
Intentional Walks 118
Non-Intentional Walks 108
Strikeouts 95
Stolen Bases 104
Caught Stealing 94
Grounded Into DP 98

Taking them by groups:

PAs/ABs  – These rates ‘favor’ the visitors.  Due to the home team not batting or not batting a full inning in the last half-innings if ahead or if they won the game, we can probably dismiss this data as not significant to our inquiry.

BA/1B/2B/3B/HR – Most of these are close to the 3% difference in runs, except for one that really stands out – Triples!   The home team triples 19% more than the visiting team.

Sac/SF – Sacrifices favor the home team by 9%, and sac flies by 7%.

HPB/BB/Strikeouts – Walks favor the home team by a rather large 9%, with INTENTIONAL walks favoring the home team by 18%.  Even taking out the intentional walks due to their elective nature, walks still favor the home team by 8%.  In conjunction with more walks, the home team strikes out 5% less than the visiting team.

SB/CS/GDP – The ‘running’ categories favor the home team, with 4% more stolen bases coupled with 6% fewer caught stealings.

Now we have some basis to venture a few guesses about the WHY.    Triples certainly stand out.   On the offensive side, it makes sense that the runners may have a better idea in their home park of when it makes sense to try for a triple, given where the ball was hit.  On the defensive side, outfielders apparently know better how to play any unusual caroms that may occur in their home park to prevent doubles from becoming triples.

Sacrifices and Sacrifice Flies seem to somewhat stand out, but I’m a bit skeptical to conclude much from these results.   We used PAs for the denominator, but since these two events are very context sensitive and, in the case of sacrifices, an elective strategy, it would probably be more enlightening to use some type of opportunity measurement, such as runners on first, no out situations for Sacs, and runners on third and less than two outs for Sac Flies.   Just looking at the other components we have, it’s safe to conclude that the home team has more opportunities per PA for each of these, and that using a more appropriate denominator is going to reduce the 9 and 7 percent numbers.

Walks and strikeouts, taken together, certainly seem to indicate the home team has a strike zone advantage.  If we look at the home/visitor non-intentional walks and strikeouts together as a ratio, the home team’s Non-BB/K ratio is 14% better than the visitor’s.  Now, this COULD be an umpiring issue, with the umpire ‘favoring’ the home team, but that would take some study of balls, swinging strikes, and called strikes to try to find umpire bias.   A more straightforward explanation would be that home batters are more accustomed to the hitting background, meaning that their visibility and therefore pitch recognition is simply better than the visiting team.  If true, this would also partially explain why other components, such as doubles and home runs, also favor the home team, as pitch recognition should result in not just better BB/K ratios, but in better contact results.

It also appears at first glance that there’s a stealing home field advantage.   To add to what is in the chart above, home teams stole 3% less than visiting teams, and home teams were successful on 68% of steal attempts vs. 66% for visiting teams.  There have always been stories of teams ‘watering down’ the base paths when facing a good running team, while on the other hand there are many instances of teams with ‘fast turf’ trying to build their rosters with fast players when possible.  Those instances don’t really explain what we see in the data.   With stolen base attempts being an elective event, and knowing that the visitors are more often behind in the score than the home team, it’s very possible that those extra 3% of attempts are simply being taken by lesser skilled base stealers.  We would need to control for the quality of the runners to make any definite conclusion about a stealing home field advantage.

When it comes to the home field, familiarity DOESN’T breed contempt!   The triples data shows strong indications that outfielders can play the corners and walls much better in their home parks than they play them on the road.  The walk/strikeout data combined with the base hit components show almost as strongly that batters recognize pitches coming out of the center field background much better in their home parks than they do on the road.   Now, whether that’s because they haven’t been out late the night before when they’re at home is for someone else to study…..

Comments

10 Responses to “How the Home Team Wins”
  1. Justin Murphy says:

    That’s great! A good way to prod it further would be if you looked at teams the year after they moved into a new stadium to see if the triples effect was lessened, since they didn’t know the walls as well yet.

  2. Rally says:

    Cool study. As to the strikeouts/walks, it could be the batters being more familiar with the hitting background, but it could also be the home team pitchers being more familiar with the mound.

    Or probably both of those, in some combination. I don’t want to dismiss umpiring, but I doubt it. To the umpire, every game is a road game.

  3. Xeifrank says:

    Nice work! Is it possible to get the component of BABIP listed? Thanks.
    vr, Xei

  4. This is exactly the sort of thing I’m looking at for basketball. I too want to know HOW (FG%, turnover rates, rebounding, etc).

    Thanks for giving me a few more things to chew on. :)

  5. KJOK says:

    Xeifrank – Yes, I can easily pull HR’s out of the AVG calculation. I will do that in the follow-up article. THANKS.

  6. Xeifrank says:

    Thanks KJOK, I will bookmark your site and wait for your follow up. I am sure many others will too.
    vr, Xei

  7. Evan says:

    Very nice work. I also look forward to the follow-up.

    As far as sacrifice bunts, I think the difference in sac bunts may be a result of managers following the mantra of “play for the lead on the road, play for the tie at home”. This might be more apparent if the percentage of sac bunts in so many opportunities changes inning by inning.

  8. kds says:

    Good work.

    A question on interpreting your numbers. When you say that Home (or Road), is 110 for one of the factors you measured, isn’t 100 the average rate, so that the other side (Road or Home), would be 90 or so? In this case the side with the 110 would be 10% better than average, but about 22% better than the other side.

    “Play for the tie at home and the win on the road,” would seem to indicate that home teams would be more likely to use one run strategies, such as sac hits and stolen bases.

  9. KJOK says:

    KDS:

    Thanks. This may not have been clear, but when I said “ratios between the home team and the away team (100 = Home and Away rates identical)” what I meant was a 110 factor means the home data was 10% greater than the road data, NOT 10% greater than the average of the two.

    Then you mentioned:

    “Play for the tie at home and the win on the road,” would seem to indicate that home teams would be more likely to use one run strategies, such as sac hits and stolen bases.

    True, and the data seems to somewhat support that.

Trackbacks

Check out what others are saying about this post...
  1. [...] Tom Meager looks at home field advantage by component, as does KJOK. [...]



Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!