MLB All-Star Game: Has homefield advantage really mattered?

facebooktwitterreddit

Going deep inside the data shows homefield advantage in the World Series is unimportant to winning the World Series. What does that mean in relation to the All-Star Game? 


On the surface it seems like this is a fairly easy question to answer. Reason would have it to determine if home field advantage has mattered, just look at which teams won the World Series in each season since 2003—the first year the All-Star Game winner determined home field advantage.

But that is actually a faulty use of logic. One, winning the World Series is a highly pragmatic view, one which is so results-oriented it screams “OVERCOMPENSATION”. Without presenting any numbers at all, I can simply point you to the fact—and yes it is a fact, even without numbers—few of the best teams since the inception of the Wild Card in baseball in 1994 have actually won the World Series.

More from MLB All-Star Game

And yet at the end of the year, we can crown the World Series winner as “the best team in baseball”. Most of the time, it is not true. They were the team who got hottest in October. That’s really all the World Series title proves.

They were good enough to get in, and then played better than anyone else for anywhere from 10 to 20 games after the regular season. Now, in some cases, they truly are the best team. But not always.

But the point here isn’t really to answer that question, as fascinating as it is, and notable as it may be to really go back all the way through 1994 and try to figure out for each season who really was “the best team in baseball” (great future article idea, am I right?).

The point is to determine if the All-Star Game has had any merit to determining the World Series champion? And if it has, does it play a role in this whole “the best team doesn’t always win the World Series” (true) narrative?

Before we move on to the numbers, though, one more caveat must given. To make this determination there are several different, logical and legitimate ways of determining this. We’ll take a look at each, and then decide if in fact, it is advantageous.


World Series Winners

Okay, so I’ve seemingly contradicted myself. First, I basically said this is a horrible way to determine if “the All-Star Game winner leading to World Series homefield advantage is advantageous” question is a false dichotomy. And now I’m considering it legitimate.

It’s legitimate, only in so far as it acts as one of the parts. It’s like Fox News. If Fox News was the only news channel, the media would be quite unbalanced. But with it, there is a greater sense of speculative opinion and chaotic chatter—which is really all the media is good for anymore, anyway.

So this category is our Fox News. It’s our balancing piece. It will keep us from going too far off the deep end of rationalism and allow us to maintain some sense of common appeal (maybe where the analogy breaks down).

YearAll-Star Game Winner (Held Homefield Advantage)World Series ChampionLeague
2003American LeagueFlorida MarlinsNational League
2004American LeagueBoston Red SoxAmerican League
2005American LeagueChicago White SoxAmerican League
2006American LeagueSt. Louis CardinalsNational League
2007American LeagueBoston Red SoxAmerican League
2008American LeaguePhiladelphia PhilliesNational League
2009American LeagueNew York YankeesAmerican League
2010National LeagueSan Francisco GiantsNational League
2011National LeagueSt. Louis CardinalsNational League
2012National LeagueSan Francisco GiantsNational League
2013American LeagueBoston Red SoxAmerican League
2014American LeagueSan Francisco GiantsNational League

There you go: eight of 12 years, the World Series champion came from the team who had homefield advantage—which of course was earned by that team’s league winning the All-Star Game in July of the same season.

At 67 percent, it is most definitely a majority, and might be semi-relevant statistically. It may even give us some notion of correlation, if not causation.

If nothing else, it is interesting. But it ultimately proves nothing (not even a perfect 12-for-12 would have proved anything; that would be a far too pragmatic approach). But this does give us a nice baseline, and even works to further motivate us to continue down the path toward figuring this out.

The next logical ally to travel is to look at the records of each team in the World Series.


World Series Opponents

Okay, now we’re getting somewhere in looking at the records—though if we really believe a 162-game season win-loss record tells us definitively who the best team is, we’re again fooling ourselves. It is just another piece in the puzzle.

In truth, this isn’t too much of an advanced approach to what we just did. It’s only adding the important element of competition. But to really determine the true effect of the All-Star Game we’ll have to compare not World Series opponents but World Series winners to themselves based on more advanced methods.

We’ll get there. For now, let’s compare the opponents.

YearAL Team/Regular season recordNL Team/Regular season recordASG Game winner/HomefieldWorld Series winner
2003New York Yankees (101-61)Florida Marlins (91-71)American LeagueFlorida Marlins
2004Boston Red Sox (98-64)St. Louis Cardinals (105-57)American LeagueBoston Red Sox
2005Chicago White Sox (99-63)Houston Astros (89-73)American LeagueChicago White Sox
2006Detroit Tigers (95-67)St. Louis Cardinals (83-78)American LeagueSt. Louis Cardinals
2007Boston Red Sox (96-66)Colorado Rockies (90-73)American LeagueBoston Red Sox
2008Tampa Bay Rays (97-65)Philadelphia Phillies (92-70)American LeaguePhiladelphia Phillies
2009New York Yankees (103-59)Philadelphia Phillies (93-69)American LeagueNew York Yankees
2010Texas Rangers (90-72)San Francisco Giants (92-70)National LeagueSan Francisco Giants
2011Texas Rangers (96-66)St. Louis Cardinals (90-72)National LeagueSt. Louis Cardinals
2012Detroit Tigers (88-74)San Francisco Giants (94-68)National LeagueSan Francisco Giants
2013Boston Red Sox (97-65)St. Louis Cardinals (97-65)American LeagueBoston Red Sox
2014Kansas City Royals (89-73)San Francisco Giants (88-74)American LeagueSan Francisco Giants

So this chart actually tells us a bit more than I’d originally intended it to, and I’m glad. It’s actually quite insightful if we allow it to tell us what it is intending.

The best piece of information it gives us is the final results of the “better” team, record-wise, in all the World Series when the All Star Game winner determined homefield advantage era (my goodness, we really need a better shorthand for this; it’s far too wordy).

The record for the higher win total team in 12 World Series in this era is 5-7. Seven out of 12 times the team with the worse record went on to win the World Series. Let’s be careful not to make too much of those details. Again, record is not necessarily the best indication of who is best.

And even if it were, it’s only among teams who made it to the World Series. But at least head-to-head it is helpful to note towards my ultimate point: the best team does not always win the World Series. But we’re still trying to figure out how much the All-Star Game plays into the phenomena.

The other terribly important thing this particular chart does for us is tells us which team would have hosted the World Series, if Major League Baseball was run like the NHL or NBA where the team with the better record gets home court/ice in the finals.

And from there it tells us at an elementary level if homefield was important.

Interestingly nine out of 12 times the team with the better record also hosted the World Series (only in MLB would the “also” be necessary in that sentence). So argument over, right? Clearly Major League Baseball has it right, more often than not, the best team hosts the World Series.

Sounds fair?

Well not to the two teams who missed out on homefield advantage because their league didn’t win the All-Star Game. And not to the Red Sox and Cardinals who earned the same 97-65 regular season record in 2013 only to see the homefield advantage determined by which league won the All-Star Game, instead of something sensical such as interleague win-loss, or if you want to go the other way, win-loss in all intra-league games.

Or most sensical of all, run differential (but we’ll get to this much more later).

Besides, saying the All-Star winner automatically gains the homefield advantage in the World Series has worked because it’s been right 75 percent of the time, is the same thing as saying the BCS was a good system because most of the time it pitted the best two teams in the country against one another in the National Championship Game.

It may be a true statement, but the question isn’t the truth of the statement. The question at stake is the process. Again, pragmatism (“it turned out to be right”) is the enemy here. We want to get it right because the process is correct.

Not to mention, yet, the greater fault even: What if there had been a different team on either side representing their league in the World Series? At least once or twice we’d have to assume it would change who should have been the team with homefield advantage (Sorry folks, I’m not going to go down that rabbit trail here and investigate that far. It would be superfluous to our purposes here.).

The next step is to take the same data we had, but add a couple categories.


Winning Percentage versus Pythagorean expected win-loss record

For the sake of not writing a million words on the topic we’re skipping ahead a bit (sounds like an ESPN midnight replay, right?).

In this category, we’re going to take both World Series team’s records—again, we’re going to assume since the rest of the postseason uses the “fair” homefield advantage method, the right team reached the World Series in each league for each year this method was used to determine homefield advantage—and compare it to their expected win-loss record.

The Pythagorean expected win-loss model is based on a team’s run differential, but does not consider strength of schedule or ballpark factors like third order win percentage does. It is the simplest way to determine if a team is underperforming or exceeding expectations.

YearAL Win %AL Expected W-LDifferenceNL Win %NL Expected W-LDifferenceExpected HF
2003.623.593+.030.562.537-.025AL
2004.605.593+.012.648.617+.031NL
2005.611.561+.050.549.561-.012PUSH
2006.586.5860.000.512.506+.006AL
2007.593.623-.030.555.562-.007AL
2008.599.568+.031.568.574-.006NL
2009.636.587+.049.574.568+.006AL
2010.555.561-.006.568.580-.012NL
2011.593.605-.012.555.543+.012AL
2012.543.537+.006.580.568+.012NL
2013.599.617-.018.599.623-.024NL
2014.549.519+.030.543.537+.006NL

There are so many different things we could do with all these numbers; so many different kinds of assertions we could make, especially if and when we add in who actually hosted the series and who  won the series and the like.

Of course one problem with the assertion here is the idea a team is more or less deserving of making the World Series (which I’ve already said is based less on how they played in the regular season, and more about just playing well in October) than someone else.

But again, we’re trying to determine if the better regular season only among the two teams who did make the World Series won the series and determine if it had anything to do with homefield advantage.

Of course in doing this, we are also leaving out home and away records during the regular season, which would be a good place to go if we wanted to know how much having home field advantage in the World Series helped or hurt each team.

You might even say it makes more sense to go in this direction than the one we’ve gone. And it is undoubtedly important.

But let’s assume, for the sake of argument, the constant with any team in the World Series is they are at least a good (multiple games over .500) home team, and likely are aided by having home field advantage once they got to the World Series (even the Marlins and Rays—teams with historically poor regular season attendance—saw an uptick in homefield advantage from an atmosphere perspective when they reached the World Series).

But one of the things we’re trying to determine is simply if the better team won. In only a few cases they did, if we take into account the Pythagorean expected win-loss record.

One of the first things we notice, the “better team” using the Pythagorean expected win-loss model had homefield advantage six out of 12 times, or literally half the time, with one series (“PUSH”) not netting a homefield winner.

What is really striking is only four times in the 12 years did the “better team”, or team with a better expected win-loss record, actually win the World Series. And only two of the four times did they win with homefield advantage in the series.

And one of those was the 2009 Yankees, who by any measure were clearly the superior team to the Phillies. Homefield advantage certainly had no bearing on the eventual winner.

So what?

The data present, if it is indeed representative of a true approach to who the better team is in the World Series, and which one should have had homefield advantage, seems to indicate at the very least homefield advantage is not really a factor in a team winning the World Series.

It is not a decided advantage, statistically-speaking anyway, to have homefield advantage. In fact, more often than not, in this era where the All-Star Game determines which team has homefield, the team without it has often “overcome the odds”.

And it seems to be about half the time a team who isn’t deemed to be the better team actually wins the World Series, regardless of who has homefield advantage.

So in a sense I just wrote over 2,000 words to prove nothing, or to prove what I, at least, already believed. The World Series is ultimately a crapshoot. Homefield or not, the “better team” sometimes wins, and sometimes does not.

Everything I said previously was itself a process—the writing process. I was trying to figure it out as I wrote. My final analysis follows.

As for what it means for the All-Star Game: It seems to me to mean it’s actually OK if Major League Baseball wants to push the idea of the game being more important because it earns that league homefield advantage in the World Series.

The importance of it is false, which may actually mean it is in a sense false advertizing. But the point is really this: If Major League Baseball wants to continue to sell the All-Star Game as important because it determines homefield advantage, all power to ‘em.

If they determine to end it and go back to a more standard win-loss record homefield determiner, great too. In reality it doesn’t matter. The All-Star Game itself is ultimately no more than an exhibition, as it always has been.

It is “The Mid-summer Classic”. And that’s all it will ever be. It has no bearing on “The Fall Classic”.

All Statistical data taken from baseball-reference.com.

Next: Each MLB team's worst All-Star Selections of All-Time

More from MLB All-Star Game