FanPost

A methodology to measure closer consistency.

Richard Mackson-USA TODAY Sports

I commented on Sean's article about Joe Nathan about how a couple of bad outings can throw off a pitcher's stats, especially so with relief pitchers because of their small number of innings pitched. I also said I should fire up a spreadsheet and play with medians, instead of means, to get a more accurate picture of what's going on with our Man Of The Ninth.

So I did.

Using Baseball-Reference's game log for Nathan this year, I dumped the data into a spreadsheet, cleaned it up a bit, and started sorting. I took out quite a few columns because I wasn't interested in the averaging-stats, merely the counting-stats (and only a few of those, to boot). He's had 43 appearances this year, which means the median in the ordered list would be #22 -- there are 21 entries above it and 22 entries below it.

Anyway, here's what I got for a statistically-typical outing.

DR IP  H  R ER BB SO HR BF Pt St GB FB LD PU
 2  1  1  0  0  0  1  0  4 16 10  1  1  0  0

(DR = days of rest, BF = batters faced, Pt = pitches, St = strikes, LD = line drives, PU = pop-ups)

Then again, maybe medians aren't quite the answer. I figured I'd test out the method on what we could probably universally agree-on as being a train-wreck of a season, Phil Coke's 2013. I did the same steps and came up with this:

DR IP  H  R ER BB SO HR BF Pt St GB FB LD PU
 2  1  1  0  0  0  0  0  3 11  7  1  1  0  0

Hmm. Oddly similar, except for one fewer batter faced. Then again, Coke in '13 was essentially a LOOGY. How about a "true" closer season which was a disaster, Jose Valverde in the first part of 2013?

DR IP  H  R ER BB SO HR BF Pt St GB FB LD PU
 2  1  0  0  0  0  1  0  4 16 11  1  2  0  0

That doesn't look too bad, actually... but we all know how that ended up. Who's the best closer of all time? Probably Mariano Rivera. His 2008 season was his biggest by WAR (2.8), so let's sort that one.

DR IP  H  R ER BB SO HR BF Pt St GB FB LD PU
 2  1  0  0  0  0  1  0  4 15 10  1  1  0  0

Not too different from above. So maybe the mark of a good closer isn't only found in their average stats, or in their median stats. Rivera was great because of his consistency, and a way to measure that is standard deviation. So maybe that'll get us closer to figuring all this out.

I dumped all the stats, their medians and their (SD) into a table, and here's where the differences start to pop out; the lower the SD, the less variability there is.

Pitcher IP H R ER BB SO HR BF Pt St GB FB LD PU
Rivera '08* 1 (0.19) 0 (0.85) 0 (0.41) 0 (0.41) 0 (0.24) 1 (0.87) 0 (0.24) 3 (1.16) 13 (4.83) 10 (3.06) 1 (1.08) 1 (1.06) 0 (0.70) 0 (0.57)
Nathan '14 1 (0.19) 1 (1.07) 0 (1.01) 0 (1.01) 0 (0.69) 1 (0.87) 0 (0.32) 4 (1.40) 16 (6.52) 10 (3.56) 1 (1.00) 1 (1.14) 0 (0.82) 0 (0.37)
Valverde '13 1 (0.10) 0 (1.45) 0 (1.27) 0 (1.27) 0 (0.57) 1 (0.76) 0 (0.66) 4 (1.40) 16 (6.39) 11 (3.94) 1 (0.81) 2 (1.56) 0 (0.83) 0 (0.69)
Rivera '08 1 (0.35) 0 (0.82) 0 (0.38) 0 (0.38) 0 (0.29) 1 (0.99) 0 (0.24) 4 (1.47) 15 (6.15) 10 (3.85) 1 (1.21) 1 (1.03) 0 (0.65) 0 (0.58)

I took Coke out of the table because he wasn't used too much like a closer, and it muddled-up the table. But, in Rivera's 2008 season he was used for more than one inning in 15 of his 64 outings, which naturally inflates some of these stats in comparison to a "ninth inning only" usage pattern (this shows up in the SD for BF); the row with the asterisk takes out the appearances of more than one inning, which settles things down a bit. I also put that row on top because, after juggling things around for a while, I realized that's how things sorted themselves out; Rivera's one-inning outings in '08 were the most consistent, then Nathan, then Valverde. (Rivera's full 2008 stats are included for reference. Holy hell he was good.)

The whole point of this exercise was to try and answer the question, How is Joe Nathan doing this year? Well, if the table above is to be believed... not bad. He's no Rivera, and he's no Valverde. But he's alright.

This is a FanPost and does not necessarily reflect the views of the <em>Bless You Boys</em> writing staff.