CSS Button No Image Css3Menu.com

Baseball Prospectus home
  
  
Click here to log in Click here for forgotten password Click here to subscribe

Futures Guide 2014 is Now Available in Paperback and Three E-book Formats.

Premium and Super Premium Subscribers Get a 20% Discount at MLB.tv!

<< Previous Article
Premium Article You Could Look It Up: ... (09/14)
<< Previous Column
Premium Article Lies, Damned Lies: Wha... (09/07)
Next Column >>
Lies, Damned Lies: Sec... (09/20)
Next Article >>
Under The Knife: Doom ... (09/14)

September 14, 2006

Lies, Damned Lies

ELO Considers the Yankees

by Nate Silver

the archives are now free.

All Baseball Prospectus Premium and Fantasy articles more than a year old are now free as a thank you to the entire Internet for making our work possible.

Not a subscriber? Get exclusive content like this delivered hot to your inbox every weekday. Click here for more information on Baseball Prospectus subscriptions or use the buttons to the right to subscribe and get instant access to the best baseball content on the web.

Subscribe for $4.95 per month
Recurring subscription - cancel anytime.


a 33% savings over the monthly price!

Purchase a $39.95 gift subscription
a 33% savings over the monthly price!

Already a subscriber? Click here and use the blue login bar to log in.

On Sunday afternoon, as most of the nation watched football, the Yankees defeated the Orioles to take over the best record in the American League from the Tigers. This result was more or less inevitable; the Yankees have been playing better baseball than the Tigers for most of the last two months, and had overtaken them in the ELO standings on August 19th. It's also a result which, coupled with the AL's victory in the All-Star Game, sets up the Yanks to have home field advantage throughout the playoffs.

Relatively little attention is paid to home field advantage in baseball, and rightly so. The home team wins between 53 and 54% of major league baseball games, as compared with about 60% in the NFL and NBA. Once a baseball team has secured a playoff spot, it generally does not pull out all the stops to ensure it has home field advantage; if the Yankees and Tigers enter the final day of the regular season tied, you're more likely to see a call-up start the game to set up the playoff rotation than you are to see Mike Mussina or Justin Verlander start on short rest.

But does home field advantage take on a special character in the postseason? And does it take on a special character when the home field is Yankee Stadium, with all of its attendant mystique, aura, and history?

The current Jeter/Rivera version of the Yankee dynasty, present in various incarnations since 1995, has gone 39-20 in postseason games at the Stadium, including a combined 12-1 mark in 1998 and 1999. If we roll back a bit further to 1976, the first year that the Yankees won the pennant following Yankee Stadium's remodeling (mystique and aura didn't begin with Derek Jeter, but some sort of cut-off is reasonable), the Yankees have gone 52-31, for a .627 winning percentage. That's an impressive result, but is it statistically significant?

We can address this question by using ELO ratings, which allow us to estimate the probability that one team beats another on any given day. The ELO ratings build in a 3.5% advantage for the home team, corresponding to the very stable home field advantage of between 3-4% that I described earlier. For example, the Yankees had a 63.4% chance of beating the visiting Padres in Game 2 of the 1998 World Series, and a 55.0% chance of beating the Dodgers to close out the 1977 World Series in Game 6 at the Stadium.

Below, we document all of the Yankees' postseason results since 1976 in this fashion. W(Ex) and L(Ex) designate the expected number of wins and losses based on accumulating the ELO ratings from the individual post-season games. For example, while the 1998 team went 6-1 at home, that team should have won nearly two-thirds of its playoff home games based on the superior caliber of that year's club. Conversely, the 6-3 home performance of the comparatively weak 2001 club, which rated as an underdog even in its home games, is relatively more impressive.

Year     W(Ex)    L(Ex)   W(Act)  L(Act)   +/-      Opponents
--------------------------------------------------------------------
1976     2.70     2.30      2        3     -0.70    Royals, Reds
1977     2.70     2.30      3        2     +0.30    Royals, Dodgers
1978     2.83     2.17      4        1     +1.17    Royals, Dodgers
1980     0.56     0.44      0        1     -0.56    Royals
1981     4.49     3.51      5        3     +0.51    Brewers, A's, Dodgers
1995     1.12    -0.12      2        0     +0.88    Mariners
1996     3.68     3.32      3        4     -0.68    Rangers, Orioles, Braves
1997     1.22     0.78      1        1     -0.22    Indians
1998     4.58     2.42      6        1     +1.42    Rangers, Indians, Padres
1999     3.42     2.58      6        0     +2.58    Rangers, Red Sox, Braves
2000     3.45     3.55      5        2     +1.55    A's, Mariners, Mets
2001     4.25     4.75      6        3     +1.75    A's, Mariners, D-Backs
2002     1.10     0.90      1        1     -0.10    Angels
2003     5.22     3.78      4        5     -1.22    Twins, Red Sox, Marlins
2004     3.19     2.81      3        3     -0.19    Twins, Red Sox
2005     1.10     0.90      1        1     -0.10    Angels
Summary 45.61     36.39    52.00    31.00
            0.556               0.627

All told, the Yankees have won 52 of their 83 home games during this period, when they "should" have won between 45 and 46. Note that the Yankees' expected winning percentage of .556 exceeds the usual home winning percentage of .535 because they were generally a stronger team than their opponents.

We can perform a crude test for statistical significance by means of a binomial distribution. How likely is it that a team that should have won 55.6% of its home games actually wins 62.7% over an 83-game sample by means of chance alone? The answer is that there is an 8% chance of this result emerging randomly, which hints at statistical significance, but is below the 95% confidence interval that statisticians usually prefer. Nevertheless, when coupled with the fact that a disproportionate number of these victories came in dramatic fashion--12(!) of the 52 home victories came in the Yankees' last at bat--this is a fairly impressive result.

Whatever weak evidence there might be for the Aaron Boone effect, what about other post-season teams? We can evaluate their results using the ELO method as well. This time, we'll group things by decade, starting with 1960, the earliest year in my ELO database.

Decade     W(Ex)     L(Ex)    W(Act)   L(Act)    +/-
-----------------------------------------------------
1960's     35.35    30.65      34        32     -1.35
1970's     74.88    65.12      75        65     +0.12
1980's     93.92    82.08     111        65    +17.08
1990's    122.42   105.58     119       109     -3.42
2000's    109.20    92.80     107        95     -2.20
Total     435.77   376.23     446.00    366.00  +10.23
               0.537               0.549

Since 1960, the home team has won 54.9% of playoff games, when ELO would have predicted 53.7%. Save for a remarkable period during the 1980s when home teams were hotter than a July day in Fallujah, this is not a statistically significant result, particularly once the Yankees are removed from the equation.

However, one can be skeptical of the impact of psychological effects like momentum and clutch performance in professional spots in general, while maintaining an open mind about potential outliers. Fifty-five thousand screaming fans on a damp October night at the Stadium with a young closer on the mound and Derek Jeter at the plate is one potential outlier. Indeed, if the Aaron Boone Effect is real, it may be that the Yanks' very history of winning playoff dramatic games at the Stadium is what has reified it.

Nate Silver is an author of Baseball Prospectus. 
Click here to see Nate's other articles. You can contact Nate by clicking here

Related Content:  Significance

0 comments have been left for this article.

<< Previous Article
Premium Article You Could Look It Up: ... (09/14)
<< Previous Column
Premium Article Lies, Damned Lies: Wha... (09/07)
Next Column >>
Lies, Damned Lies: Sec... (09/20)
Next Article >>
Under The Knife: Doom ... (09/14)

RECENTLY AT BASEBALL PROSPECTUS
Fantasy Article Free Agent Watch: Week Four
Premium Article Minor League Update: Games of Wednesday, Apr...
Premium Article Skewed Left: The Squandered Promise of the D...
Premium Article Explaining Spending
Premium Article Pebble Hunting: Platoonies Never Say Die
Overthinking It: The Obligatory Article Abou...
Premium Article What You Need to Know: Pineda and Pine Tar, ...

MORE FROM SEPTEMBER 14, 2006
Under The Knife: Doom and Gloom
Premium Article You Could Look It Up: No Hitters Ain't That ...
Premium Article Schrodinger's Bat: The Whole, the Sum, and t...
Premium Article Future Shock: Short-Term Impact, AL

MORE BY NATE SILVER
2006-10-11 - Premium Article Lies, Damned Lies: Projection Reflection
2006-09-27 - Premium Article Lies, Damned Lies: Playoff Hurlers
2006-09-20 - Lies, Damned Lies: Secret Sauce
2006-09-14 - Premium Article Lies, Damned Lies: ELO Considers the Yankees
2006-09-07 - Premium Article Lies, Damned Lies: What to Do With Papelbon
2006-08-30 - Lies, Damned Lies: To Close Or Not To Close
2006-08-23 - Premium Article Lies, Damned Lies: Tufte Takes on the Red So...
More...

MORE LIES, DAMNED LIES
2006-10-11 - Premium Article Lies, Damned Lies: Projection Reflection
2006-09-27 - Premium Article Lies, Damned Lies: Playoff Hurlers
2006-09-20 - Lies, Damned Lies: Secret Sauce
2006-09-14 - Premium Article Lies, Damned Lies: ELO Considers the Yankees
2006-09-07 - Premium Article Lies, Damned Lies: What to Do With Papelbon
2006-08-30 - Lies, Damned Lies: To Close Or Not To Close
2006-08-23 - Premium Article Lies, Damned Lies: Tufte Takes on the Red So...
More...

INCOMING ARTICLE LINKS
2007-09-06 - Prospectus Toolbox: The Playoff Odds Report