Baserunning Runs, Retrosheet Era

Colin Wyers presents baserunning runs compared to average for all the years between 1953 and 2007. 2008 numbers will be available once retrosheet data is available, but here are the 2007 leaders and trailers:

6 Juan Pierre
6 Brian Giles
5 Dan Uggla
5 Luis Gonzalez
5 Johnny Damon
5 Orlando Cabrera
5 Marcus Giles
5 Hideki Matsui
5 Aaron Miles

-5 Ryan Garko
-5 Jacque Jones
-6 Barry Bonds
-6 Mike Lowell
-7 John Mabry
-9 Paul Lo Duca

Who wants to run some year-to-year-correlations? And who's going to beat me to a full slice-and-dice post?

7 months ago Sky Kalkman Comment 21 comments 1 recs |

Story-email Email Printer Print

Read More: colin wyers, baserunning, retrosheet

Comments

I done gone and drooled all over my lap.

And That Happened: Sports with a Hint of Sarcasm

by philkid3 on Nov 19, 2008 9:06 PM EST reply actions 0 recs

I ran a year-to-year correlation and only got 0.257…

by erosen on Nov 20, 2008 1:21 PM EST reply actions 0 recs

Nevermind… I think I “forgot” to combine the seasons split between different teams.

by erosen on Nov 20, 2008 1:24 PM EST up reply actions 0 recs

That didn’t make much of a difference 0.261…

by erosen on Nov 20, 2008 1:36 PM EST up reply actions 0 recs

I think we really want some sort of playing time adjustment, too

So we can use (runs/200 times-on-base) or (runs/600 PAs) instead of absolute runs.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 3:30 PM EST up reply actions 0 recs

I figured that short seasons might be impacting it— so tried a 300 AB+BB minimum and the correlation is up to .299.

Using that sample, going to a per 600 PA ratio ~~- perhaps players with inconsistant play-time have relativelyconsistant +/~~…? Nope .289 correlation

Or per 600 Times On Base – HRs…? Maybe based on how often they are on base will level it off alittle? Not there either [.286]

Here are the top and bottom for those rate versions for 1998-2007:

per 600 Times On Base – HRs
2005 Larry Walker +40
2000 Tom Goodwin +39
2004 Darrin Erstad +38
2005 Ruben Gotay +31
2002 Rob Mackowiak +27
…
2005 Xavier Nady -28
2007 Michael Barrett -29
1999 Mike Stanley -34
2002 Quinton McCracken -35
2007 Paul LoDuca -38

per 600 PA
2000 Tom Goodwin +13.3
2005 Larry Walker +13.2
2004 Darrin Erstad +12.5
2005 Ruben Gotay +8.3
2006 Alfredo Amezaga +8.2
…
2005 Kevin Millar -7.8
1998 Fred McGriff -7.8
2007 Paul LoDuca -11.0
1999 Mike Stanley -11.7
2002 Quinton McCracken -12.8

by erosen on Nov 20, 2008 3:10 PM EST reply actions 0 recs

I see you already handled my caveat I just posted above. Nice work.

Actually, a better opportunity denominator would be this (and require Colin’s help):

Every play gives the baserunner a positive score or a negative score. In total, someone might have +7 run from good plays and -5 runs from bad plays, for a grand total of +2 runs. If we use the absolute value of the changes, we get +2 runs per 12 “run-opportunities”. This would account for players who are often running the bases compared to those who aren’t.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 3:33 PM EST up reply actions 0 recs

I'll look into it.

I has a blog. | Ask a sabermagician!

by cwyers on Nov 20, 2008 6:03 PM EST up reply actions 0 recs

Or, to be more fair, add up both the positive and negative absolute values possible at each choice.

That wouldn’t discriminate against whether a guy made a good choice or a bad choice.

Well, there isn’t actually always a dichotomy, is there? You could go for an extra base and either make it or not make it, or else choose not to go for an extra base. So…. hm….?

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 6:13 PM EST up reply actions 0 recs

Mike Lowell is remarkably slow

He’s one of the worst baserunners I ever saw with a Red Sox jersey.

Mother---- him and John Wayne!

by MerryGoByeBye on Nov 20, 2008 3:19 PM EST reply actions 0 recs

Career, per 600PA, min 3000 career PA:

TomGoodwin 5.7
Mookie Wilson4.4
Bobby Tolan +3.7
Freddie Patek +3.6
Al Smith +3.5
Darin Erstad +3.4
Willie Wilson +3.3
…
Earl Williams -3.3
Bruce Benedict -3.4
Pat Borders -3.6
BengieMolina -3.6
Chris Hoiles -3.7

And here are some players who stole alot of bases, yet don’t fare very well:
RANK PLAYER +/- [SB/600]

Otis Nixon 65
Rickey Henderson +1.2 64
Tim Raines +1.1 48
Lou Brock +0.9 51
Billy North 52

by erosen on Nov 20, 2008 3:33 PM EST reply actions 1 recs

Again, more great stuff. How about combined baserunning and SB/CS run values?

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 3:36 PM EST up reply actions 0 recs

Stolen bases and caught stealings are included.

I should probably break those out explicitly at some point.

I has a blog. | Ask a sabermagician!

by cwyers on Nov 20, 2008 6:02 PM EST up reply actions 0 recs

Oh, really? That's an important point.

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 6:14 PM EST up reply actions 0 recs

Yeah.

And they’re situational as well… a successful stolen base in a “high leverage” situation relative to run expectancy (not win expectancy, however) is worth more under this system.

A lot of people like to look at them seperately, however. I’ve had that request when I used a similar setup before.

I has a blog. | Ask a sabermagician!

by cwyers on Nov 20, 2008 6:16 PM EST up reply actions 0 recs

Shouldn't be too hard to un-leverage them, right?

There are uses for both lines of analysis. And yeah, broken out is nice, because we already know how good players are at stealing bases – it’s the other side of the coin we’re foaming at the mouth for.

Are you going to apply this same methodology to outfield throwing arm? And would it be possible to combine the two with some sort of WOWY approach to account for “strength of schedule”. That is, getting thrown out by Ichiro is more acceptable than being thrown out by Johnny Damon…

Beyond the Boxscore // Calling BJ Upton lazy is lazy.

by Sky Kalkman on Nov 20, 2008 6:28 PM EST up reply actions 0 recs

SB/CS is easy to unleverage.

The other events could be unleveraged as well.

But I think that including context in baserunning is more important because the runner has more of a choice as to when he tries to go for extra bases and when he doesn’t, based upon the context. The runner is able to leverage himself and I think that matters.

It’s probably trivial to turn these scripts into arm ratings. I suppose at that point you could use something like Odds Ratio to adjust for “strength-of-schedule”, though it’d probably be a bear to do.

I has a blog. | Ask a sabermagician!

by cwyers on Nov 20, 2008 6:50 PM EST up reply actions 0 recs