FanPost

Nolan Ryan and BABIP

First I want to make clear that I think that Nolan Ryan was a good baseball player. I don’t like to pass out terms like “great” or “best” lightly and so won’t use them to describe him, but at the very least he was very good at what he did, far better than I’ve ever been at anything I’ve done, for instance, and there is certainly a longevity-based argument for him to be in the Hall of Fame. I’ll even go so far as to say that his inclusion in the Hall does not cheapen it unless you are a fan of a very, very small Hall, which I am not.

However, there is a massive gulf of difference between “very good for a long time” and “the best ever”. This seems to be a fairly controversial opinion. For instance, in the 1999 All-Century team sponsored by Mastercard and Major League Baseball, Ryan was named by the fans as the top right-handed pitcher of the period (Sandy Koufax was voted best lefty). His career numbers at a glance also hint at greatness: most strikeouts ever, most no-hitters in the history of the game, 8 All-Star appearances, 2nd all-time in career games started, lowest hits/9 innings ratio of all time… it’s quite a pedigree. And by one definition of “great” – the one that means something similar to “large” – there isn’t a lot of question here either: Ryan was the most extreme power pitcher of his or of any era.

Any time a very good or “great” player like Ryan is discussed, there are always arguments as to the proper contextualization of his career accomplishments. One argument against Ryan is that as gaudy as his strikeout total and no-hitters look, his peripheral numbers don’t all appear to add up to a level that a Tom Seaver, for example, achieved.  His career ERA+ is a solidly above average but nothing special 112, and while his FIP, which better takes into account the things that a pitcher can control than ERA does, bumps his career runs-allowed rate from 3.07 per 9 to 2.97, it’s only enough to put him on a similar plane with Tom Seaver, Bob Gibson, and Roger Clemens once era (not ERA!) and the ballparks each player played in are taken into account.

Still, there is the knotty question of Ryan’s BABIP, which was only .265 in his career, a number that by itself certainly seems to indicate that he ought to be given even a little more credit than what is due from a glance at his FIP. Is this a big deal? I’ll look at this question in three parts.

 

Part I: Explainability

I am no scientist – I graduated with the highly unscientific degree of English with an even more unscientific emphasis on creative writing – but I do read a lot of science-related material in general and biology in particular, and one thing which comes back in over and over again is the concept that statistical anomaly by itself is not of a lot of use without a means to explain it. For example, with BABIP there is a hypothesis out there which states that knuckleball pitchers have some ability to post lower BABIPs than what would be expected from them due to several factors. Charlie Hough, for instance, finished his career with a very low .250; Tim Wakefield’s .274 is also well below the average over the past few years of around .290. Looking at the data, I might extend this out to include any pitcher who throws a single pitch 70% or more of the time, which also allows us to explain Mariano Rivera and his cutter (.261 career BABIP).

The important point about this isn’t that sabermetricians have been able to create a post hoc explanation for certain phenomena. Humans are very, very good at assigning meaning to patterns, even if those patterns turn out to be meaningless. No, the point is that these sabermetricians have created a hypothesis which can be tested against future performance. For instance, when R.A. Dickey’s career ends in a few years, we’ll be able to look back and see if his BABIP was abnormally low or not. If it isn’t in his case (so far it’s sitting at a very un-Hough .300), it might be due to luck. Enough R.A. Dickeys, though, will disprove that hypothesis.

In this respect, it’s not enough to simply observe that Ryan posted a lower than average BABIP over the course of his career. It is entirely possible that other factors beyond his control may have driven this metric down (I will look at this more in Part 3) or even, as improbable as this may sound, that his performance in this regard over the course of his career was due to luck.

It’s not as amazing as it sounds. Yes, the chances that Ryan himself had a low career BABIP from luck alone is extremely improbable, but given enough pitcher careers, there are bound to be some which end with higher or lower totals than what they “should” have based on little more than random chance. Unfortunately, I simply do not have the time or the background in statistical work to come up with a P value for Ryan’s performance, but I am sure that it is above zero because that is the nature of probability.

So what, then, could be causing Ryan to be so hard to make good contact off of, that lowers his batting average even more than all of his strikeouts dictate? One hypothesis put out there is that perhaps really good strikeout pitchers also only allow weak contact on a higher percentage of the balls they do allow to be put into play. Here are the top 10 K/9 pitchers of all time and their BABIPs:

 

At first glance, there may be something there, but if so, it’s rather subtle: several guys are just a tick below average, with all of the real outliers either guys who played in different eras or, in the case of Trevor Hoffman, the kind of player whose BABIP could be explained in other ways (he’s a relief specialist, the sample size is a bit smaller than with many of the other players, and throughout much of his career his modus operandi was to throw his 95 mph heater as much as possible and only occasionally mix in a circle change or slider to keep hitters off balance. FWIW, 2-pitch specialists might have much the same advantage as one-pitch guys: witness Kaz Sasaki and his .246 career BABIP). Nonetheless, there also isn’t anyone on this list *over* the average, which could mean a correlation. Further studies are needed.

 

Part II: What About The Other Numbers?

 

Earned run average is nowadays only a little more well-respected by sabermetricians as a predictive tool than wins are. There are good reasons for this, which I’ll list here in descending order of importance:

-          Earned run average does not take park factor or the era a pitcher played in into account. Back when ERA was still king, pitchers like Wes Ferrell were not even considered for the Hall of Fame because of their relatively high ERAs – high ERAs in a historical sense, that is, but not necessarily in the specific context of the 1930s.

-          For relief pitchers, earned run average doesn’t take into account the runs a pitcher allowed but which were scored against the man who put those runners who came in on base. Additionally, it doesn’t take into account the increased effect on winning percentage a pitcher who enters the game with the score tied in the 9th inning gives to his team if he pitches well compared to the same level of performance from a guy who enters the game in the 6th inning down by 12 runs.

-          Earned run average measures things that pitchers don’t really give up, namely runs. Pitchers give up walks and home runs and hit batters every now and then and prevent hits by striking out batters, but don’t really give up runs per se, just the elements which contribute to runs.

-          In addition to being lucky enough to not give up a lot of hits, relatively speaking, pitchers can also get lucky in the sense that the men they face just happen to not come around to score. Some people think some pitchers have an innate ability to control this, but statistical proof is lacking.

-          Pitchers who allow a lot of runs on errors aren’t necessarily any better that pitchers who allow the same amount of runs on hits because the difference between an error and a hit is largely subjective. This is a bigger issue when looking at pitchers from early baseball history than modern guys.

ERA+, which takes the top point into account by adjusting for park and era, has interesting things to say about Ryan and his contemporaries:

  • Tom Seaver 128
  • Jim Palmer 126
  • Dave Stieb 123
  • Ron Guidry 119
  • Bert Blyleven 118
  • Gaylord Perry 117
  • Steve Carlton 115
  • Fergie Jenkins 115
  • Phil Niekro 115
  • John Candelaria 114
  • RYAN 112
  • Tommy John 111
  • Don Sutton 108
  • Frank Tanana 106
  • Jack Morris 105

 

Now, I don’t want to make it seem as though I actually think that all those guys in front of Ryan were actually better pitchers; longevity, again, does provide a compelling argument in and of itself, and Ryan pitched longer than any man in the history of the game. That being said, the ERA+ list is interesting as much for what it does *not* adjust for as what it does. ERA doesn’t care whether the players it looks at got to where they were on a career level by getting lots of strikeouts, avoiding walks, inducing more double plays than one would expect, allowing fewer hits on balls put into play, or leaving more runners stranded over the course of a career. If Nolan Ryan’s BABIP is based on some as-yet-unquantified skill which ought to be taken into consideration above and beyond what his DIPS numbers indicate, shouldn’t we also take into consideration at least the possibility that his inability to prevent runners from being stranded also represents an as-yet-unquantified lack of skill? ERA+ says that before you adjust out the former and adjust in the latter, Ryan helped his team out on a per-inning basis as much as Tommy John, Phil Niekro, and John Candelaria, and significantly less than Tom Seaver or Jim Palmer.

There is another factor at play with career statistics. One of the wondrous things about Ryan’s career is that he never had an ugly decline phase the way Steve Carlton, for example, did. While on the one hand this is certainly in Ryan’s favor, as it speaks to his ability to continue to pitch well for a long, long time, it also means that a guy like Steve Carlton’s peak is less represented in his career rate stats than it is with Ryan’s. I will leave the reader to decide whether or not this is an important distinction to make.

Part III: What About Context?

The first thing which may shock you about the Ryan vs. Seaver debate: Tom Seaver’s career BABIP was .250; as low as Nolan Ryan’s total was, Seaver’s was significantly lower. Of course, Seaver pitched a good chunk of his career in pitcher-friendly environments such as Shea Stadium and Old Comiskey Park… but wait. The same can be said about Nolan Ryan, and in spades. Anaheim Stadium in the 1970s and the Astrodome in the 1980s both had R/A park factors of about .95, trailing just San Diego and essentially tied with the Oakland Coliseum during the late 70s and early 80s and Safeco Field as the most extreme pitchers’ parks since the deadball era.

In any case, here are Ryan’s BABIP numbers compared with his team:

Year

Team

Innings

BABIP

Team

Difference

1968

NYM

134

242

261

-19

1969

NYM

89.3

252

252

0

1970

NYM

131.7

228

260

-32

1971

NYM

152

273

265

8

1972

CAL

284

236

254

-18

1973

CAL

326

280

278

2

1974

CAL

332.7

255

281

-26

1975

CAL

198

264

281

-17

1976

CAL

284.3

269

272

-3

1977

CAL

299

263

282

-19

1978

CAL

234.7

292

275

17

1979

CAL

222.7

266

286

-20

1980

HOU

233.7

291

278

13

1981

HOU

149

246

261

-15

1982

HOU

250.3

263

276

-13

1983

HOU

196.3

247

261

-14

1984

HOU

183.7

274

280

-6

1985

HOU

232

293

278

15

1986

HOU

178

241

260

-19

1987

HOU

211.7

281

285

-4

1988

HOU

220

285

270

15

1989

TEX

239.3

258

274

-16

1990

TEX

204

244

279

-35

1991

TEX

173

230

290

-60

1992

TEX

157.3

301

300

1

1993

TEX

66.3

246

294

-48

 

Weighted by innings pitched, I come up with a difference of a little over 10 points in Ryan’s favor, a nice number, almost certainly statistically significant, but against the backdrop of the last 20 years, that’s slightly less (but overall comparable) to the effect Kerry Wood has had on hitters’ ability to make good contact on his pitches. In any case, it’s not that huge of an effect, on the order of an extra hit a month Ryan prevented over the course of his career.

Conclusion

I’ve looked at 3 different items regarding Ryan and come up with 3 things:

1.       Because it’s hard to pin down exactly why it’s so low in a way we can predict other pitchers will be as low, Ryan’s BABIP may or may not be due in part to luck or other circumstances beyond his control.

2.       Looking at a statistic most sabermetricians stay away from in large part because it does not differentiate well between pitchers who do a good job at controlling their game and pitchers who have a good defense or park behind them, Nolan Ryan does not look nearly as impressive as some of his counting statistics (strikeouts for instance) or that BABIP would lead you to believe.

3.       Looking at Ryan’s career against the backdrop of his era, the parks he played in, and the defenses behind him, Ryan’s .265 BABIP is actually more like a .281, all things considered.

 

I hope that this was enlightening for the reader as it was for me. I for one found the first bit especially interesting, and as to the third, while I came in with a general idea that Ryan’s BABIP wasn’t *quite* as impressive as it looks at first glance, I had no idea that park/era/defense adjustments could end up being so severe.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

Join Lookout Landing

You must be a member of Lookout Landing to participate.

We have our own Community Guidelines at Lookout Landing. You should read them.

Join Lookout Landing

You must be a member of Lookout Landing to participate.

We have our own Community Guidelines at Lookout Landing. You should read them.

Spinner

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9351_tracker