Forecasting the 2010 Series based on precedent

Two days ago, after Game Two of the World Series, I wrote a piece here at THT Live looking at what had happened in previous World Series when a team won each of the first two contests. Short version: Based on precedent, the Rangers had a one in five chance to win the Series. However, if they won the Game Three, their chances rose to two in five, but would become zero in five if they lost. (Again, this is only based on World Series precedent, not probability theory).

Well, the Rangers won Game Three. So now what does the future hold? Rather than look at times a team was up two games to one (that’s almost all the Series), or even those when one team won the first two and the other won the third decision, let’s note one key fact about this Fall Classic: The home team has won every contest. At the risk of paying too much attention to small sample sizes, that strikes me as noteworthy. After all, in the 50 previous times a team won the first two decisions in a World Series, only 14 happened to the road team and 34 to the home squad (two Series split the first games between towns).

What does the past say about a World Series in which the home team wins each of the first three games? Well, it’s happened 20 times previously. Here’s how those 20 played out (and I’ll present the results from the point of view of the up team, which is the Giants in 2010).

Four won the Series in five games (1916, 1933, 1988, 2000)
Four won the Series in six games (1917, 1953, 1980, 1995)
Four won the Series in seven games (1947, 1987, 1991, 2001)
One won the Series in eight games. (1919 was a best-of-nine. It’s a Series where one team was on the take, so its predictive value is highly questionable)
Five lost the Series in seven games (1955, 1956, 1958, 1965, 1971)
Two lost the Series in six games (1978, 1981)

So, based on this, the Giants have a 13/20 chance of winning, and the Rangers a 7/20. My own hunch is that it’s probably a bit better for Texas than that. (I like the previous two-fifths guesstimate). If nothing else, Saturday’s win guarantees Cliff Lee will start again. While Game One proved he’s human and that Tim Lincecum fella ain’t bad, Texas has to like getting two starts from Lee this October.

For baseball fans the good news is this: There’s an 80 percent chance for the Series going a least six games, and about a 50/50 chance for a seven gamer, based on precedent. About time, as we haven’t had a seven-game Series since 2002. Heck, the Ser9ies has featured only one six-gamer since 2003.

One related question: If 20 previous Series had the home team win each of the first three games, how long did it take until the road team finally notched a win? Let’s see .. .

Game Four had the first road victory seven times (1916, 1919, 1933, 1958, 1988, 1995 and 2000). The team up two-games-to-one won six of those (all but 1958).

Game Five had the first road victory three times (1947, 1953, 1980), all of which saw the team that won the first two games of the Series win the world title.

Game Six had the first road victory three times (1917, 1978, 1981), only one of which (1917) saw the team that won the first pair of games claim the entire Series.

Game Seven had the first road victory four times (1955, 1956, 1965 and 1971). In every case, the team that LOST the first two games of the Series won the Series.

Three times the home team won every game (1987, 1991 and 2001).

Putting it another way, 13 of these 20 games had the home team win Game Four. And of the remaining all-homer Series, the home team won 10 out of 13 times in Game Five, and then seven out of 10 times in Game Six. Oddly, the home team lost a majority of Game Sevens: four out of seven.

Based on the above, if the Rangers win the rest of their home games, history says they have a 60 percent chance of winning it all, because six of the 10 times the home team won all of the first five games saw the squad in the Rangers’ position win it all. Flipping it around, if they lose just one of the two remaining games, their chances become much bleaker: Nine of the 10 times the first road win came in Games Four and Five, the squad in SF’s position claimed the title. Basically, the numbers above say Games Four and Five matter the most. That make sense: The Rangers are the team most needing wins and they’re the home team, so a loss would be especially bad for them.

My own hazy guess is that Game Four will be key. If the Rangers take that, they have Lee going at home, and if he can avoid being beaten twice in one postseason series, the Rangers need only one of two games in San Fran.

Then again, none of this really matters. That’s why they play the game and the great thing about baseball is its continual ability to confound everything.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Ichiro Suzuki, meet linear weights
Next: And That Happened »


  1. DrBGiantsfan said...

    I tend to agree with the notion that tomorrow’s game of pretty important.  From the Giants perspective, if they lost tomorrow, it’s tough to count of Cliff Lee being subpar two consecutive games, then they have to win both games back in SF, something I don’t think they can count on at all.  If they win tomorrow, they can lose Game 5 and still win the series if they split the last two in SF.

  2. MWoo said...

    Although I generally like this article, this reasoning falls into the classic trap:
    “While Game One proved he’s human and that Tim Lincecum fella ain’t bad, Texas has to like getting two starts from Lee this October.”
    Texas has to like getting two starts from Lee, only if both starts are his established post-season level.  In the next 4 games, the Rangers can only hope for getting ONE start they would like from Lee this World Series (unless they weirdly decide to swap Lee for Lewis in a possible Game 7). 

    The comment from DrB has the related “conditional statement” problem.  Yes, it’s unlikely for Lee to be subpar for two straight games, but if you add the conditional antecedent, “given Lee was subpar for his last game,…”  To be subpar for two straight, Lee has to be subpar for just ONE game, his next one.

    In 28 regular season games this year, Lee had 10 games where he gave up >=4 earned runs and lost 7 of those games.
    In 33 regular season games, Lincecum had 9 games where he gave up >=4 earned runs and lost 7 of those games. Yes, I know that Tim plays at AT&T and without a DH.  But they’re not miles apart.

    I think that the Lee/Lincecum matchup, at home, does favor Texas, but it’s not the slam dunk that people are making it out to be. And the Wilson/Lewis/Hunter v. Cain/Sanchez/MadBum matchups are going to decide the majority of what happens.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>