The Baseball Graphs Blog

Friday, March 24, 2006

Baseball Prospectus 2006

I finished reading the 2006 edition of Baseball Prospectus, and I’m mighty proud about it.  The thing is 554 pages long.  I’ve actually never read the entire Prospectus from front to back before (I’ve bought it for the last five years and just read the parts I found interesting).

I sometimes give the Baseball Prospectus writers a hard time, but I’m really a big fan.  I’m particularly a big fan of the PECOTA projection system, not because I think it’s more accurate than other systems (I have no idea if it is or isn’t) but because I think it’s the right way to make projections.  Instead of a single line, or a line plus a “reliability score,” PECOTA projections include a likely range of potential player projections including “breakout” and “attrition” percentages (among other things).  And they’ve put a robust set of PECOTA projections in the Annual, including a list of comparable players.  If you don’t subscribe to their website, this alone makes the book worth the price.

The book has three basic parts: a management editorial of each team, player stats, projections and comments for many, many players (including most team’s prospects) and general baseball research articles (called “Fungoes”) in the back.  It’s the same format they’ve used for several years.

The editorial sections are generally quite critical, calling out team managements for their mistakes.  My favorite was probably Oakland’s, which included a number of great insights about player values and drafts.  The Philadelphia article also had a nice analysis of farm system productivity.  They feel that Ned Colletti is off to a good start in Los Angeles and they are curiously upbeat about San Diego’s offseason.

They feel Jim Leyland won’t do a good job of managing Detroit’s young pitchers, and found that Colorado players tend to acclimate to their ballpark over time, meaning they become less effective on the road the longer they play at Coors.  As someone who follows baseball quite closely, I probably learned something new from about 20% of the articles, which ain’t bad.  Your actual results may vary.

The player projections and comments are the core of the book, and I thought they were generally excellent.  I really liked one change in the statistical format: they changed the batted ball info from Groundball/Flyball ratio to percent of batted balls that were groundballs.  This is a much more effective way to show batted ball tendencies and I hope it’s adopted by a lot of other sites.

My main criticism is that they sometimes get too cute in their comments, which undermines good writing.  I can’t blame them, really.  It’s hard to write so many good player comments.  They have some great lines (Ryan Howard: “sometimes good fortune is the residue of incompetence”) but they also have some head scratchers (Paul Wilson: “Further deponent sayeth not.”  Huh?).

Analogies seem to be the coin of baseball writers, and BPro writers are no exception.  But it’s worth remembering that analogies should primarily be used to help readers understand what the writer is trying to say.  Lots of sportswriters like to use them for entertainment purposes and BPro sometimes uses them to show how smart they are.  Unfortunately, I had no idea what they were talking about in those instances (Matt Capps = Gunther’s Dik-Dik?).  There’s such a thing as being too clever, and BPro 2006 sometimes falls into that trap.

But that’s a stylistic thing.  In general, I enjoyed the player comments very much.

I have to say, however, that I was disappointed in the “Fungoes.”  I thought the article about limitations of statistical analysis was kind of silly and self-serving (as was the St. Louis team article, in which the Cardinals were praised for hiring lots and lots of statisticians).  As a WPA fan, I was very much looking forward to Keith Woolner’s Win Expectancy article, but it generally didn’t have many new insights.  I did appreciate the chart of event win values changing by run environment (that’s a mouthful!) but that was about it.

The one Fungo I did enjoy was the business review, which focused on several issues that are typically not covered by the media.  In the future, BPro might want to think about dropping “Fungoes” altogether and keeping the book focused on team and player comments.  The Fungo articles can be published on their website.

My bottom line is that the Baseball Prospectus 2006 was up to their usual levels except in the “Fungoes” section.  I suggest you purchase it primarily for the team editorials and player projections/comments.  I doubt you’ll regret it.

You can read Patriot’s review of Baseball Prospectus 2006 at his blog.

Posted by Studes on 03/24 at 02:03 PM
Wednesday, March 15, 2006

Expected Win Shares Review

At the beginning of each of the past few seasons, the Baseball Crank has analyzed the “established Win Shares” level of every player on every team, to make a gross projection for each team.  I’m not a big fan of using Win Shares to project future performance, but the Crank has done something better by analyzing how each team did last year, compared to their established Win Shares levels.  This is insightful.

What’s most interesting to me are the “added” Win Shares by players not originally accounted for.  This is a way of gauging which teams have come up with the most new talent (the Braves and A’s), and which ones haven’t (the Reds and Phillies).

Posted by Studes on 03/15 at 08:27 PM
Win SharesPermalink
Monday, March 13, 2006

WPAing the blown call against Japan

Ben Kabak takes a look at the WPA repercussions of the blown call against Japan yesterday.  Great example of what WPA can do.

Posted by Studes on 03/13 at 06:53 AM
Win Probability AddedPermalink
Friday, March 03, 2006

Salary Treemap

Erik Wannebo has created a great treemap of this year’s major league salaries.  You’ll need a java-enabled browser (don’t worry, your browser is almost certinaly enabled for java).  This is a great (and useful) way to look at baseball salaries.  Just click on the page once it’s loaded.

Posted by Studes on 03/03 at 01:17 PM
Graphs and GraphicsPermalink