THE BOOK--Playing The Percentages In Baseball

Tuesday, October 07, 2008

Vote: Most Outstanding Players of 2008

Complete WAR, 2008

Justin gives us a best-of-breed.  He also applies my revised positional adjustments, which, as of today, are these:
+1.25 C
+0.75 SS
+0.25 2B/3B/CF
-0.75 LF/RF
-1.25 1B
-1.75 DH ... Should be -2.25, but then I add in 0.50 for the DH penalty of how hard it is to hit off the bench.

Monday, October 06, 2008

Last time each team won a playoff series

Nice list.

Power or Finesse Pitchers in the Post-season?

Bill James, in what will surely be an article to appear in the next Gold Mine, looks at the issue of whether the Power or Finesse pitchers perform better in the post-season.  He does his typically enjoyable study of matched pairs, where he proceeds to select 100 power pitchers and 100 finesse pitchers (they match in a variety of ways, except in K and BB).  They match up quite well in the categories he selected. He also notes:

But the power pitchers had averaged 183 strikeouts, 76 walks; the finesse pitchers had averaged 107 strikeouts, 57 walks.  The two groups were nearly even in terms of home runs allowed (a few more for the power pitchers), but the finesse pitchers had given up, on average, 18 more hits.  18 more hits, 19 less walks, one less homer. . .the same results overall.

As you guys know, I’m big on simply doing K minus BB, per PA.  And just looking at the bolded part, you can see that I think the two groups are biased. I responded:

Read More

Evaluating the 2008 Forecasting Systems

David compares THT, Chone, PECOTA, and ZiPS. 

Saturday, October 04, 2008

While it is next to impossible for us (as outsiders) to evaluate managers…

Sometimes you can get an idea as to how well a manager actually “understands” the game.  To wit:

Jerry Manuel says:

“You don’t see a lot of guys that have statistical numbers play well in these championship series,” Manuel said. “What you see is usually the little second baseman or somebody like that carries off the M.V.P. trophy that nobody expected him to do. That’s because he’s comfortable in playing that form of baseball, so therefore when the stage comes, it’s not a struggle for him.”

I pity the poor Met fans.

Friday, October 03, 2008

Complete Linear Weights, 2008

Colin provides his data for easy access, along with his intro article.

Thursday, October 02, 2008

How can we “predict” which teams will do well in the post-season?

Here is an SI,com article by John Donovan (I don’t know who he is).  In it, he says:

Still, there are smart ways to pick the teams that will fare best in the playoffs. Nate Silver and the hard-thinkers over at Baseball Prospectus have looked at tons of data and come up with a formula that identifies the three main characteristics of a successful playoff team. They are:

You might notice there’s no mention of home runs or the ability to squeeze a guy to second with one out against a left-hander. There’s not anything in there about crafty managers or experience or a versatile bench, either. Momentum? History? Don’t even bother. Speed? Pssh. Clutchness? Please, save it.

Read More

Baseball Thesis

Long-time reader Mike has his thesis posted on my site:

The home advantage has been consistently demonstrated across a number of sports, but conclusive evidence of the origin of the home advantage has yet to be found. One factor thought to contribute to the home advantage is familiarity; the home team is more familiar with their stadium and playing field and thus should have an advantage in competition. To isolate this variable, we compared the records of teams in their last year at a stadium, where familiarity should be high, with their records their first year in a new stadium, where familiarity should be low. Professional baseball, hockey, football, and basketball data from the four major U.S. leagues were examined. Results showed no differences in home winning percentage between a team’s high-familiarity season and its following low-familiarity season, suggesting that familiarity does not play a major role in the home advantage.

Situational Wins

Here is my first stab at trying to describe Situational Wins.  Please provide comments, especially as it pertains to readability.  I will then make the necessary modifications, and I’ll submit it to THT for publication for the general public to consume.

Wednesday, October 01, 2008

Nate Silver meets Dan Rather

Sweet.

Tuesday, September 30, 2008

Sabermetric Playoffs

Put your thoughts here on the games you are watching…

HR rates by height

Reason?  Sampling bias.  Who are the players 6’5” and greater?  And do they appear in both sets (aging)?  You only have 10% of the sample, and so, much more likely for wild swings.  Create 4 groups of 25% of the ballplayers, and I’d bet you get smooth results.

Sabermetric Moves of the 2009 Pre-Season

Here we go again…

Splitting the batting lines into binomial metrics

Pizza lays out the idea.  As studes noted, we talked about this alot in the past. 

What Brian suggests in the comments is the way I normally approach the problem, as the way Voros did it.  Here are my aging patterns by these metrics.

I also echo Pizza’s position on where to put the HR.  Sometimes I do it the way Pizza says it, and sometimes the way Voros says it.  The fact of the matter is that you can construct two equally plausible scenarios.

There is an undeniable relationship between K, BB, and HR.  There is also an undeniable relationship between HR and FB (and to a lesser extent LD).  The only rigtht way to do it is to model this relationship.  If for example you do it as Pizza proposes it, then you need to have an additional function on the HR/FB rate that includes the K and BB rate.  If you do it as Voros proposes it, you need to include the FB rate to apply to the HR rate.

Double Plays and Groundball Chances

Colin gives us lots of data to consider.

The worst hitting season of all-time?

It seems that this year’s Tony Pena, Jr is in the discussion.

His Situational Wins (aka WPA/LI) is -2.45 in only 235 PA, making it a rate of -7.3 wins per 162G.  However, he was super clutch (for him), Ortiz-like.  On Aug 10, in extra innings, he got an RBI single.  That was his most high-leveraged at bat.  In his next most high-leverage situation he faced, he got another RBI single on opening day.  On June 12, in a tie game in late innings, he scored on a wild pitch.  (Ok, that may have been lucky.) Anyway, the point is that with the game on the line, he went from being one of the worst hitters in the entire history of baseball, to simply being one of the worst hitters in baseball this year.

That’d be like someone who couldn’t score at a whorehouse, to becoming someone who managed to be able to speak to the girl next door.  If that isn’t clutch, I don’t know what is.

Relative League Strength

Rally looks at the relative strength of each league by looking at players that move league-to-league.  But, instead of looking at the players in-season, or season-to-season, he looks at decade-to-decade.

I have some concerns, especially if age-wise, the movement is biased and there are no age adjustments.  That’s why in-season works so well.  That also means the sample size kills you, as Rally is noting.  In any case, it’s an interesting way to look at it, especially if you can handle the potential age bias.

Monday, September 29, 2008

End of year Sabermetric stats

Courtesy of Patriot.

I don’t really have much to add.  Patriot noted that he uses a 73% offensive replacement level, likening it to a .350 OW%.  Using PythagenPat and 4.5 RPG per team, I get .364.  No biggie.  Just wanted to point out that he should probably be saying .360 not .350.  However, what if you look at it as one replacement guy with 8 average guys?  In this case, this team will win .486 games, making our replacement level -.014 wins per game (or more accurately per one-ninth of a game slice).  Adjusted to a per game basis, that’s -.014 times 9 equals -.126, or a .374 win%. 

That is, rather than presuming 9 replacement-level hitters with a team of average defense, we presume 1 replacement-level hitter, 8 average hitters, and average defense.  That gives you a .486 win%.  The marginal impact is .014, which you “annualize” by multiplying by 9.  Kinda like ERA for relievers.  Anyway, to get it to my replacement level, I’d use 74% or 75%.  We’re pretty much in agreement here.

With starters, if we repeat this process, but presume 5.4 IP per replacement start, and the bullpen gives him average support, then Patriot’s 125% gives you a starter win% of .390.  To make it .380, you’d want 1.27 or 1.28.  So, 125% is perfectly fine.

For relievers, it’s the same process as hitters, if you presume 1 IP per replacement relief.  You’d want 106% or 107% of league average.

Anyway, basic core agreement, with just a smidge of disagreement on the peripherals.

Who should switch positions?

I’m running the Fans Scouting Report for each player, giving him a valuation per position.  Since speed matters little for a catcher, that component is underweighted if I put that player as catcher.  Speed matters alot in CF, so I overweight that component.  I simply put each player at each position, compare that to the position he played the most often this year (through games of Aug 20, 2008), and give you the list:

Read More