<$BlogRSDUrl$>

Michael/Male/26-30. Lives in United States/Pennsylvania/Wexford/Christopher Wren, speaks English. Spends 20% of daytime online. Uses a Fast (128k-512k) connection. And likes baseball /politics.
This is my blogchalk:
United States, Pennsylvania, Wexford, Christopher Wren, English, Michael, Male, 26-30, baseball , politics.

Wednesday, May 24, 2006

BaseRuns, BattingRuns, Runs Created ... 

Here at A Citizens Blog I’m always on the lookout for new ideas and concepts. Recently I began to start using new run estimators other than Bill James Runs Created. Everyone knows the story of Runs Created, but I’ll back up:

When he began his work on baseball stats Bill James sought an over-all stat to measure a player’s output. Since RBIs and Runs Scored depended on the contributions of his teammates, James looked for a stat that would independently estimate a player’s contribution to his team. When James plugged numbers into the following formula he found that he could usually estimate a team’s runs accurately:

((H + BB) * Total Bases) / (AB + BB)

And thus, Runs Created were born. Since then James has tinkered with the formula to make it more complex and more accurate. Here is the formula today:

A: H + BB + HBP – CS – GIDP
B: (S * 1.125) + (D * 1.69) + (T * 3.02) + (HR * 3.73) + (.29* (BB + HBP – IBB)) + (.492 * (SB + SF + SH)) – (.04 * K)
C: AB + BB + HBP + SF + SH

Basically. A represents base runners, B represents advancement of those runners and C represents opportunities. The final formula is:

Runs Created: (A * B) / C

Proponents of Runs Created argue that it is the most complex and accurate system devised for run created. I’d agree with that. I know of no other system that takes the totality of the game and gives you values to see how a player contributes.

Here are how the Phillies did with Runs Created in 2005:

Bobby Abreu: 116
Pat Burrell: 109
Chase Utley: 102
Jimmy Rollins: 100
Ryan Howard: 63
Kenny Lofton: 62
David Bell: 55
Mike Lieberthal: 49
Jason Michaels: 47

Obviously some players did better than others: sure David Bell had more Runs Created than Jason Michaels, but Michaels did his in 289 At-Bats, compared to Bell’s 557. Ryan Howard only played in 88 games but got 63 Runs Created. Hypothetically, if he had as many At-Bats as Chase Utley he’d have 109 Runs Created. Let’s move on…

There are other systems out there. I’m going to be comparing two of them. First is Pete Palmer’s Batting Runs, which is part of Palmer’s Linear Weights system:

BattingRuns: (.47 * H) + (.38 * D) + (.55 * T) + (.93 * HR) + (.33 * (BB + HBP)) + (.22 * SB) + (-.38 * CS) + (-.10 * (AB – H))

It is interesting stuff, though I doubt its overall accuracy because it doesn’t take into account something like Grounding Into Double Plays (GIDP), which Runs Created Does. On the plus side, it is a simple formula.

One small caveat: the value of an out in BattingRuns varies with league factors and whether or not the out ends an inning or not. We’re just using -.10 here.

BattingRuns:
Bobby Abreu: 118
Pat Burrell: 109
Chase Utley: 108
Jimmy Rollins: 99
David Bell: 63
Ryan Howard: 59
Kenny Lofton: 59
Mike Lieberthal: 55
Jason Michaels: 48

Those results are all remarkably similar to Runs Created, though David Bell makes out well with BattingRuns because it doesn’t penalize him for grounding into double plays twenty-four times in 2005, double that of any other Phillie (next closest: Pat Burrell with 12).

Moving along …

The second I am using is BaseRuns, which was developed by David Smyth. BaseRuns is a more complex formula:

BaseRuns:
A: H + BB + HBP - HR
B: (.8 * S) + (2.1 * D) + (3.4 * T) + (1.8 * HR) + (.1* (BB+HBP))
C: AB – H
D: HR

(A* (B/ (B+C) ) ) +D … confusing? Okay, divide B by B + C, then multiply the result by A, then add D.

The B & C numbers combine for a “scoring rate”, then multiplied into the baserunners. Supposedly proponents of BaseRuns argue that it is the most accurate because it delivers the right numbers even in “extreme” environments like beer league softball games.

BaseRuns:
Bobby Abreu: 116
Chase Utley: 105
Pat Burrell: 99
Jimmy Rollins: 93
David Bell: 64
Ryan Howard: 59
Kenny Lofton: 58
Mike Lieberthal: 55
Jason Michaels: 49

Again, David Bell’s 24 GIDP’s aren’t a factor so he’s making out fairly well. What was very interesting to me was that Pat Burrell dropped so dramatically. I’m really not entirely sure why.
So how accurate are the systems? I took the Phillies 2005 team numbers and plugged them into the equasion

Phillies
Actual Runs Scored: 807
BaseRuns: 810
BattingRuns: 824
Runs Created: 838

This tells me a few things:

1. The Phillies offensively weren’t that efficient in 2005: their actual runs were far below the numbers they “should” have scored. They score 4% less under Runs Created, 0.4% fewer with BaseRuns, and 2% less than with Batting Runs. If you look at the Phillies pitching numbers, Bill James thinks the Phillies should have won 95 games last year and badly beaten the Braves (by seven games) for the division. Instead the Phillies finished second and missed the playoffs again.

2. Proponents of BaseRuns have a point. They were nearly on the money. Maximum respect.

Last night ... the Phillies dropped a tough one to the Mets in 16 innings. This will probably be the longest game of the year and it will be interesting to see if the Mets pitching might be a little worn out after all of this. I thought Ryan did a good job and I can't help but think that Gavin Floyd and not he deserved to be demoted to the bullpen.

Cole Hamels won't pitch tonight, so Jon Lieber will go in his stead. I think Lieber will pitch well: I suspect batters will be fatigued and won't enjoy trying to hit 70 mph curveballs and sinkers.

Comments:
If total plate appearances is in the denominator for Runs Created, then why would it matter that Bell got more at bats than Howard or Michaels? The metric looks like it should result in runs created per plate appearance, but instead it looks like a total season number. What am I missing?
 
It is a total season number, so the more opportunities you have, the more runs you create. The significance is that Ryan Howard had 60% of Bell's TPA's and had nearly the same # of runs.

Pawnking: a team full of Bobby Abreu's would probably score 1,200 runs!
 
Post a Comment

This page is powered by Blogger. Isn't yours?