2008 ALCS: Same Picture So Far

With the Red Sox forcing a game 7 last night with their win against the Tampa Bay Rays, I thought it’d be fun see just how similar this series looks to last year’s ALCS when the Red Sox played the Indians.

A couple years ago I combined all the Win Probability charts of the 2004 ALCS. Here’s what the 2007 and 2008 ALCS look like when superimposed. 2008 is in red and 2007 is in black.

ALCS 07 vs 08 Small

The Rays have thus far had a much better chance at winning the series than the Indians ever had. In the top of the 7th with Rays runners on 1st and 2nd and no outs, they had a 99.3% chance of winning game 5. The Indians in 2007 had, at their best, a 57.9% chance of winning game 5, and never had better than a 50% chance to win game 6 and game 7.

When compared to the 2004 ALCS, the Yankees at one point had an 83.8% chance of clinching in game 4, and then an 87.8% chance of clinching in game 5.




Print This Post



David Appelman is the creator of FanGraphs.


5 Responses to “2008 ALCS: Same Picture So Far”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. billso says:

    I knew the Rays really should have finished off the BoSox in Game 5… this just makes it painfully obvious.

    Vote -1 Vote +1

  2. lookatthosetwins says:

    It would be interesting the see a comparison of the Yanks/Red sox series from 04. The Yankees had to have at least a 99.3% chance of winning that series at some point.

    Vote -1 Vote +1

  3. Scappy says:

    It would be nice to see a WPA graph that instead of going from -0.5 to 0.5 it goes from -2.0 to 2.0. So any event in game 7 is effectively multiplied by 4 to reflect the added importance of the game.

    Vote -1 Vote +1

  4. sibpin says:

    Scappy, what’s your basis for using 4 as a WPA multiplier? I’ve set up a table to describe what WPA multipliers should be, based on the difference in probabilty a team wins the World Series before/after a given game. That data, plus a description of the Red Sox’ performance, is available at http://sonsofsamhorn.net/index.php?showtopic=37598&st=340

    Vote -1 Vote +1

  5. Lee Gregory says:

    I find it more interesting to re-plot the whole WE data set converted into and ALCSeries win expectancy. To do this I have to make a simplistic assumption though that the outcome of each game, going into it, is a 50/50 proposition. Then I can calculate the expected series win probabilities at the beginning and end of each game, and convert the game WE for each event to a series WE.

    An example can be seen at http://sonsofsamhorn.net/index.php?s=&showtopic=37598&view=findpost&p=1956974

    Vote -1 Vote +1