2038 Strikeout Proficiency

[jleddy]

Note: As a member of the UMEBA, I would have loved to have run this analysis for the league, but alas, due to the limited number of teams and thinned out total number of starting pitchers, there wasn't enough of a sample size to provide an interesting discussion. Therefore, the following is analysis of BBA pitchers that I think some of you will enjoy.

To K or Not to K, Is That Really a Question?:
A Study on Strikeout Proficiency
"If you can get an out on one pitch, take it. Let the strikeouts come on the outstanding pitches. Winning is the big thing. If you throw a lot of pitches, before you know it, your arm is gone." ~Dwight Gooden


If you ask any pitcher, they are likely to tell you that the strikeout is arguably their out of choice. Sure, some think strikeouts are boring and facist, and of course a tapper back to the mound or a lazy pop-up will do just fine...especially if the at-bat is only one or two pitches.[/extreme foreshadowing] However, in terms of dominance, bragging rights, Steve Nebraska Award votes, and salaries, the strikeout rules all. But what is the best way to measure the value of a strikeout as well as a pitcher's overall talent and potential?

Read the typical BBA article about a pitcher or scan the back of his baseball card and the sheer number of strikeouts he's compiled is likely to be cited. Here are the strikeout kings from the 2038 season of qualified* pitchers (*1 1P per team game, so some workhouse relievers were included):

NAME	TEAM	K
Jubal Troop	EDM	247
Arthur Dempster	RCK	241
Feliciano Rafael	SFB	235
Aki Kondo	SA	231
Carlos Rodríguez	CLG	231
Héctor Márquez	SEA	221
Peter Grady	JAX	220
Chris Kelly	TC	217
Miguel Ramos	CAL	214
Jefferson Pierce	CLG	213

An impressive list of a counting stat but it is missing context, like the fact that Dempster trailed Troop by only six whiffs despite pitching in nearly twenty less innings. It's interesting to note that Grady made the top ten as a reliever.

Now perhaps you want to go even deeper and obtain some context and examine a pitcher's strikeouts per nine innings (K/9) as a metric, which provides a useful rate statistic hinting at a pitcher's additional dominance:

NAME	TEAM	K/9
Carlos Rodríguez	CLG	11.5
Aki Kondo	SA	10.5
Jefferson Pierce	CLG	10.1
Manuel Andrés	CAL	10.0
Diogo Lindt	LBC	9.8
Arthur Dempster	RCK	9.7
Peter Grady	JAX	9.4
Miguel Ramos	CAL	9.4
two tied	n/a	9.2

Rodríguez jumps from 5th in total strikeouts to 1st when taking into account innings, whereas "strikeout king" Troop falls into a tie for tenth, meaning his title is purely titular based on the most rudimentary of analysis. Andrés makes the top half of the K/9 list despite finishing the season outside the top ten (13th) in strikeouts.

So total strikeouts is good, but whiffs per nine innings is even better. That said, K/9 tends to favor hurlers who face more batters and have a greater opportunity at a strikeout. Conversely, K/9 penalizes pitchers who don't give up a lot of hits, walks and hit by pitches. Striking out the side in order is a much greater outcome than a pitcher who strikes out the side along with a few baserunners, so strikeouts per batters faced (K/BF) is even better than K/9.

NAME	TEAM	K/BF
Aki Kondo	SA	0.289
Carlos Rodríguez	CLG	0.282
Diogo Lindt	LBC	0.270
Jefferson Pierce	CLG	0.270
Arthur Dempster	RCK	0.258
Peter Grady	JAX	0.256
Yrrigs Carpenter	SA	0.254
Manuel Andrés	CAL	0.253
Feliciano Rafael	SFB	0.252
Miguel Ramos	CAL	0.250

A third list and a different name atop the leaderboard for the third time! Kondo, who finished 4th in total Ks and 2nd in K/9, gets his due and shows his combination of both dominance and efficiency. A newcomer on this list is Carpenter, who was awarded with the Steve Nebraska Golden Arm Award for the Johnson League in 2038. "Only" pitching 208 innings kept him off the counting stat leaderboard and he narrowly missed landing among the K/9 leaders with 9.0, good for 12th best.

Okay, we've established that a K is useful, K/9 is better and K/BF is even more desirable. Alas, not all batters faced are equal, right? Two pitchers may have the same K/BF but what if one achieves the strikeout in fewer pitches? If a pitcher is missing bats at a higher percentage of the time, he's more likely to pitch deeper into games and record a greater number of strikeouts (and outs in general) than his counterparts. Now we're really lasering in on dominance AND efficiency.

NAME	TEAM	K/100
Aki Kondo	SA	7.423
Carlos Rodríguez	CLG	6.985
Peter Grady	JAX	6.890
Diogo Lindt	LBC	6.889
Jefferson Pierce	CLG	6.880
Arthur Dempster	RCK	6.870
Yrrigs Carpenter	SA	6.776
Feliciano Rafael	SFB	6.573
Manuel Andrés	CAL	6.532
Tristan Alfama	YS9	6.504

There's your strikeouts per 100 pitches (K/100) leaderboard for 2038. (The stat is simply strikeouts divided by total pitches, multiplied by 100 to give it a relatable number more akin to K/9, plus most pitchers target approximately 100 pitches per start.) The list is made up with virtually the same names as before, but in just a different order and with Alfama filling in for Ramos. Kondo stays on top and by a massive margin, leading some to possibly question the validity of this statistic. More on Kondo later, but here's proof that K/100 is the best defensive independent pitching stat (DIPS) you can find.

Among those qualified BBA pitchers from 2038, below are the correlations of the three strikeout statistics used above, (K/9, K/BF and K/100) against four run-prevention metrics: Earned Run Average (ERA), Adjusted ERA (ERA+), Fielding Independent Pitching (FIP), and Component ERA (CERA).

	ERA	ERA+	FIP	CERA
K/9	-0.425	0.442	-0.508	-0.505
K/BF	-0.507	0.526	-0.569	-0.595
K/100	-0.513	0.532	-0.581	-0.608

It's clear that K/100 has the strongest correlation to all four run-prevention metrics. [The correlation coefficient measures the strength of a linear relationship between two variables. The correlation coefficient is always between -1 and +1. The closer the correlation is to +/-1, the closer to a perfect linear relationship.]

Now back to Kondo, who went 13-8 this past season en route to an All-Star Game nod. On the surface, the San Antonio Outlaw hurler has rather pedestrian traditional stats, and even some average advanced stats as well [keep in mind this is based on the 103 qualified pitchers from 2038]:

STAT	TOTAL	RANK
ERA	3.99	26th
IP	198.2	51st
BABIP	.303	72nd
ERA+	115	78th
HR/9	1.5	85th

While a 3.99 ERA isn't horrible (top quarter of all qualified pitchers), you can see his ERA+ took a beating. Kondo finished middle of the pack in innings pitched, which can be attributed to his Stamina rating of 6. The Japanese fireballer also was prone to the long ball (Movement rating of 5), which led to an inflated ERA and, along with some bad luck, a mediocre BABIP. Kondo did perform substantially better on the road than at home but the San Antonio ballpark isn't an extreme hitters park by any means. Additionally, the Outlaws ran out an absolutely average, middle-of-the-pack defense behind Kondo. Who's to say a few upgrades in defense couldn't lead to a major breakout for Kondo?

There are some outstanding signs of dominance that tie back into Kondo's league-best K/100:

STAT	TOTAL	RANK
K/100	7.42	1st
K/9	10.5	2nd
K/BB	6.6	3rd
K	231	4th
WHIP	1.09	9th
BB/9	1.59	10th
FIP	3.53	11th
CERA	3.37	16th

As you can see, Kondo just about strikes out everyone and walks no one...what more could you ask from your starter? In fact, Kondo lead the league in K/9 for a fourth-straight season. Kondo's command, control and dominance points to a very, very special talent. Cutting down on gopher balls, going deeper in games and maybe some help in the field will only make Kondo that much greater.

All in all, it's clear that the more strikeouts, the better and the fewer pitches, the better. Combining the two (K/100) is a recipe for success and could be the go-to stat the next time you do some scouting and evaluating. Who really cares if the strikeout is boring and fascist as long as it's efficient.

[RonCo]

I like K/100. Especially because we finally get a Nine pitcher in there.

Overall, I like thinking about the value of the K.

[Ted]

This is great work. Thanks.