Statistically Regressing the Phillies’ Offense

EDIT: Petti noticed an error in his calculator. He e-mailed me the updated numbers. As such, the post has been edited to reflect the changes.

At Beyond the Box Score, Bill Petti (@BillPetti) did some number-crunching, looking at the actual production (in terms of wOBA) from baseball’s best and worst hitters and comparing it to a regressed version, using each hitter’s numbers over the past three years. Petti described his process:

To gain some perspective on the early returns this season I decided to regress all batters with >=40 plate appearances by their 3-year average batting average on balls in play (BABIP) and their 3-year average HR/FB rate. The chart above presents the top-15 and bottom-15 batters in terms of the difference between their actual wOBA this year and what we would expect given their 3-year averages in BABIP and HR/FB.

Petti made his calculator public, so I downloaded it and ran the numbers for the Phillies. Below are the results.

Name Actual Adjusted Regressed
Raul Ibanez 0.231 0.236 0.293
Ryan Howard 0.362 0.344 0.408
Carlos Ruiz 0.304 0.309 0.328
Jimmy Rollins 0.306 0.307 0.321
Wilson Valdez 0.279 0.274 0.282
Ben Francisco 0.337 0.322 0.328
Shane Victorino 0.363 0.35 0.336
Placido Polanco 0.415 0.419 0.347

In graph form (click to enlarge):

Five of the eight hitters come out looking better after the regression. Ibanez goes from a .231 wOBA to .293, but that is still well below average (around .315). Howard had a .046 point regression up to .408, and Rollins went up .015 to .321. Ruiz jumped 24 points to .024 while Valdez had barely any change at all, moving only three points.

Polanco fell the hardest, from .415 to .347, an 68-point plunge. Victorino and Francisco were hit for 27 and 9 points to .336 and .328 respectively.

It’s not the most encouraging analysis, for sure, but things could be worse. The offense hasn’t been good lately and there aren’t any excuses for that, not even bad luck. That won’t change by any significant margin unless the Phillies get Chase Utley back and at his previous level of production, Domonic Brown makes a speedy rebound from a broken hamate bone, or if they get creative as Eric Seidman hypothesizes at Brotherly Glove.

Leave a Reply



  1. bje79

    April 27, 2011 10:14 AM

    Wow, great (although depressing) stuff!

  2. joe

    April 27, 2011 12:40 PM

    Any stats that makes Ibanez look good or at least better than anyone else on this roster must be a crock of s#*t. He sucks and he is getting worse every time he takes the field. I would bench him and let Mayberry Jr play more, he could do any worse.

  3. Richard

    April 27, 2011 01:18 PM

    It doesn’t make him look good, it tells us that he is likely to improve.

  4. MplsPhilsFan

    April 27, 2011 01:54 PM


    Thanks for providing this data, I find it fascinating.

    I work with regression in my job regularly, so I understand its strengths and limitations. The three year timeframe does provide some smoothing out of the data, but does not account for a player who may have been liminted because of a reduced number of plate appearances, such as Rollins (injury) or Francisco (lack of opportunity).

    Now in looking at this chart, there are not a ton of surprises. Polanco is playing out of his mind right now, so it is no surprise to see his regression line will track downward.

    I will say I was a bit surprised by Shane’s regression, and wonder if that is because of the large fluctuation in his numbers over the past few years. Offensively, the guy has been all over the place, so the R squared for him (showing how predictive a regression line will be) is probably low, especially when compared with someone like Ibanez whose numbers are easily predictable.

    Perhaps I am being too optimistic, but I do think this this team will be much better by mid June when Utley, Brown, Lidge and Contreras will be back.

    Utley at even 75% will be a huge upgrade, Brwon will struggle, but provides some speed at the very least. The bullpen will have Lidge, Contreras, Madson, Bastardo, Stutes and Romero (4 pitchers with high strikeout rates and one LOOGY)

    I just don’t see anyone matching up with us

  5. Bill Pettti

    April 27, 2011 06:27 PM

    A Giants fan that was non-too pleased with Sandoval’s numbers led me to look into it and the error was depressing some of the estimates.

    Of course, as a Mets fan I liked the numbers better when they were lower, but intellectual honesty and all…

  6. Rob

    April 27, 2011 09:51 PM

    You had me at “Phillies offense” and “regressing”

  7. Css228

    April 28, 2011 09:35 AM

    Honestly it doesn’t look so bad now that the numbers have been fixed. I also am optimistic about our roster once healthy.

  8. Css228

    April 28, 2011 09:35 AM

    All that said, Ibanez needs to be replaced. .293 is just all around unacceptable for a corner outfield who can’t play defense

  9. Smitty

    April 28, 2011 12:32 PM

    I hope it gets better but Ibanez ain’t geting any younger and now that the beard is gone he ain’t getting any more attractive. Wouldn’t Jermaine Dye or even an Aaron Rowand be a better fit for left/right field help?

  10. BM

    April 28, 2011 02:20 PM

    In order regress Ibanez HR/FB rate, you have to assume that his age has not caused permanent loss of power. I am not sure you can count on that.

    Zips still thinks Ibanez will end up with 17 or 18 HRs this year. I think Zips is nuts on this one.

  11. Css228

    April 28, 2011 06:25 PM

    @Smitty. Hunter Pence. Perfect fit.

  12. Tyler

    April 28, 2011 07:46 PM

    It is interesting to see that not only is Ibanez not hitting, but when he does hit he is not getting the extra base hits. Most all of the other starters have a bit of a higher wOBA exept for Valdez. It is good to see that Ruiz and Howard are getting on base and getting the exta bases when necessary. Please correct me if I am incorrectly interpreting the data.

Next ArticleMets Series Preview with Joe Janish