Win Probability Update

Good news! Win Probability and Run Expectancy stats have all been park adjusted for 2008. It looks like the run environment during the season was set a little too high, so you’ll notice batter stats are slightly higher, while pitcher stats are a bit lower. Other than that, not much changed. Manny Ramirez is still #1 in WPA for batters and Cliff Lee is still #1 for pitchers.

Print This Post

David Appelman is the creator of FanGraphs.

10 Responses to “Win Probability Update”

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

    Has or another website looked into calculating playoff probability update? In other words, calculated win probability added for a regular season game multiplied by the game’s impact on increasing or decreasing a team’s probability of landing a playoff slot. I realize that a small number of plate appearances will have a disproportionately large impact on making the playoffs and that it would be very volatile with most of the statistic accumulated during the last 2 weeks of the season, but it might be fun to quantify it and see who ranked the highest.

    Vote -1 Vote +1

  2. Steve says:

    Geez, what happened to Roy Halladay? He must have gotten a ridiculous park-adjustment, because he’s now well behind Lee in WPA/LI.

    Vote -1 Vote +1

  3. The Edge says:

    David, your “park adjustments” are ridiculous. Do you simply use the average runs per game of the stadium or something? There’s no other explanation for the Ballpark in Arlington having roughly a 6.21 run environment. That would take a 130 park factor which not even Coors Field achieved in its prime. The effect is manifest to a lesser extent in Detroit—5.49? Really?—but is no less indicting.

    On the flip side, McAfee Coliseum and the Rogers Centre are not 3.96. That would take a 83 park factor on their parts. If you’re going to come up with such ridiculous figures, you’re far better off not using these so-called park adjustments. You’re not adjusting for park. You’re adjusting for the quality of the team that plays its home games in that park.

    Vote -1 Vote +1

  4. The Edge says:

    Steve, it was the ridiculous “park adjustment” he gave the Rogers Centre. It essentially penalized all Blue Jays pitchers for being so good and having such a poor offense. Conversely, it rewarded Blue Jays hitters for the opposite reasons.

    Vote -1 Vote +1

  5. The Edge says:

    Well, no, the same reasons is what I should have said. Basically, his park adjustments compare a team to itself.

    Vote -1 Vote +1

  6. Steve says:

    The Edge: As much as I like seeing Lee overtake Halladay, I have to agree that the Rogers Centre penalty seems excessive. I mean, A.J. Burnett actually has a negative WPA/LI. Now, I know his season wasn’t quite as good as it looked, but a guy who led the league in Ks and had at the very least an ERA+ of 106 was worth negative wins? That doesn’t seem right.

    Seems like what happened to the Indians in the 90s, when The Jake was tabbed as a hitters park simply because we had great offenses, even though if anything it rates as a very slight pitchers park, though it’s really pretty neutral.

    Vote -1 Vote +1

  7. For the park adjustments I’ve always used the “Quick Park Factors” described here:

    Vote -1 Vote +1

  8. tangotiger says:

    The Edge: excellent point.

    Yes, the park environment is based on the number of runs scored in that park. And, if you followed the link, you correctly surmised that if you happen to have a bunch of pitchers who happen to pitch great in the same park, that this depresses that run environment. Also as noted in the link, this is neither the best way to do this, nor the worst way. You believe however that while this may work out most of the time, the few times it doesn’t work really obfuscates things enough as to call into question the whole process to begin with.

    Let me think about it, and feel free to post additional relevant comments.

    Vote -1 Vote +1

  9. tangotiger says:

    TheEdge: I elaborate more on my site, at post 3 (click my name).

    Feel free to continue the discussion there. You can do it here as well, but I don’t monitor the threads here as intently as on my own site.

    Vote -1 Vote +1

  10. Sky says:

    Patriot has five-year regressed park factors already worked out for you:

    Vote -1 Vote +1