SB Nation - Login for mobile commenting

Lookout Landing

Somebody Explain This To Me

A table of league-total baserunning numbers, via Baseball Prospectus. Note that:

EqGAR = Equivalent Ground Advancement Runs

EqSBR = Equivalent Stolen Base Runs

EqAAR = Equivalent Air Advancement Runs

EqHAR = Equivalent Hit Advancement Runs

EqOAR = Equivalent Other Advancement Runs

EqBRR = Equivalent Base Running Runs (Sum of all components)

Year EQGAR EQSBR EQAAR EQHAR EQOAR EQBRR
1985 0 -155 4 0 0 -150
1986 0 -223 0 0 0 -223
1987 0 -143 0 0 0 -143
1988 0 -208 1 0 0 -207
1989 0 -217 1 0 0 -215
1990 0 -232 0 0 0 -232
1991 0 -293 2 0 0 -291
1992 0 -266 2 0 0 -265
1993 0 -292 0 0 0 -291
1994 0 -182 0 0 0 -181
1995 0 -197 0 0 0 -197
1996 0 -212 0 0 0 -210
1997 0 -340 1 0 0 -338
1998 0 -317 1 0 0 -315
1999 0 -299 0 0 0 -298
2000 0 -296 1 0 0 -296
2001 0 -321 2 0 0 -318
2002 0 -310 0 0 0 -308
2003 0 -172 0 0 0 -170
2004 0 -171 2 0 0 -169
2005 0 -174 1 0 0 -171
2006 0 -141 0 0 0 -139
2007 0 -47 0 0 0 -45
2008 0 -66 1 0 0 -65
2009 0 -88 1 0 0 -85

I imagine this pattern continues as you go beyond 1985. I just got tired of copying and pasting.

Why is it that, over the past 25 seasons, BP has all Major League baserunners as contributing -5362 runs via stolen base attempts? Why does everything else add up to ~zero?

1 recs  |  25 comments

Comments

And what happened between 2003 and 2009?

That’s quite a dropoff.

Are all of these stats against an average for individuals?
Don't steal, mkay, if you steal you're bad, mkay, because stealing is bad, mkay...
Every time I see John Clayton, I expect to hear "mkay" at the end of his sentences.
I'm guessing that there are too many CS compared to SB hence the negative numbers.

The fact that everything else is zero bothers me more. How can the running game for SB be so (apparently) unoptimal but every other aspect is almost exactly perfectly optimized. Looks like somebody doesn’t know how to treat some values so they are making sure that they add up to zero. That is more bothersome. At least I can understand how players tend to have bad SB% so that can lead to negative values.

I bet if you account for leverage in the SB situations, EQSBR +/- 50 each year or at least a lot closer than it is now.

Do you have SB% for the league as a function of year?

Using the method I posted just prior

1997 was -175 runs for MLB. So based on a sample size of 2 years (shut up), I have no problem believing that baserunners as a whole have cost a shitload of runs with SB attempts over the last 20 years, but without knowing how BP calculated it their numbers seem extreme. Maybe they take leverage into account?

I'm betting they just use a higher baseline for CS% and run values.

Up it by a little bit (say, .16 for SB and .50 for CS) and the numbers would changed dramatically.

I would assume leverage would hopefully help alot

Pinch runners have good SB% and usually steal in critical times. Fat players often have bad SB% and try to steal in low leverage spots. (Hopefully)

Well for 2006 using

Using 0.16 runs for a SB and -0.45 for a CS that year, the AL was at -25 runs total and the NL -32 for a grant total of -55. Not sure how EQSBR runs is calculated…

Yeah

Calculating the league numbers manually every year always yields a negative number, but not this negative. So it’s not so much the sign as it is the magnitude.

Maybe they do use leverage and guys actually hurt their team a lot by getting out in critical portions of the game

Or maybe we are ignoring guys who get caught trying to steal 3B which hurts the team more than trying to steal 2B.

By going in and checking on guys with 1 attempt each year

It doesn’t appear that leverage is taken into account. And I imagine the 3B case is included when calculating the familiar +0.16 / -0.45 figures, no?

Hmm. Maybe it is included. I'm not sure. I'll have to check after I get home.
I couldn't find the values for stealing (or caught stealing) 2B or 3B by themselves

Anybody know what they are?

Of course its easy to imagine BP messing up

SB runs seems like the easiest to calculate out of all of those, though…

Inclues 3B and home

You are right

Yeah, what Fett said.

This article has a few numbers that are interesting.

Did anyone email BP?

The few times I have emailed anyone at BP, they have responded within a day.

Well I'm going to rec this even if you guys aren't!

Very interesting find, Jeff. I’d like to think it’s caused by some sinister doomsday-plot, but it’s probably just a quirk in the calculations. Doomsday calculations.

But seriously, couldn’t this signify a serious flaw in the metric?

So they're using baserunning above/below league average?

But stolen base attempts are calculated as run values as if in a linear run equation?

They conjecture that runner advancement tactics are roughly optimal, but....

…base-stealing tactics are not. We know that teams have been stealing too much for at least the last hundred years or so, but I don’t know of comparable research demonstrating whether runner advancement policies are/are not foolish. They may be assuming that teams are optimizing advancement fairly well.

Regarding the 5000+ runs lost to base-stealing, it’s unsurprising. Only the highest-percentage basestealers help their teams, while generally everyone else is harmful. For an amusing illustration of this effect, check out these two outliers: 1. The previous all-time record holder for steals, Lou Brock, made a negative net contribution to his teams in that category; 2. Rickey Henderson’s worst stealing season by far, in terms of run value, was his record-setting 1982 season. He set a new steals record in September, but broke the record for being caught in July.

The slide from 2003 onward, including the sharper drop since 2007, can be explained by two kinds of actual decrease. The “default” amount of steal attempts most teams make is lower than it was ten years ago, and the number of teams each year that adopt crazily aggressive basestealing is falling. Teams have been wising up about this in recent years.

Or they changed their run values or in some other way modified their calculations
Its a little strange that runner advancement tactics would be perfectly optimal
What about pickoffs?

If CS numbers include pickoffs as well as actual times caught stealing (which is my understanding) then this doesn’t necessarily equate to stealing too much.

Yuniesky Betancourt alone probably accounts for a couple of the EQSBR from times he dozed off on the bases.

I already checked and pickoffs do not make up the difference

You must Login with your SB Nation account and be a member of Lookout Landing to post a comment.