« Cano: Four More Years! | Main | The Clemens Report »

Sunday, January 27, 2008


Feed You can follow this conversation by subscribing to the comment feed for this post.

Run, Chuck, run!

Good analysis YF.

In addition to the power numbers, one of the major benefits of steroids and HGH, at least to my understanding, is the quickened recovery rates, of users, from injuries. That, and the general prolongation of careers. These are two factors which cannot really be measured for reasons already mentioned in YF's post

A few brief notes:

"PF doesn't seem to me a reasonable indication of anything." It is an indication of the force with which a given batter, or set of batters, meets a well-struck ball on average. It tallies almost exactly with HR/H, for those who prefer a simpler datum. It is an excellent measure of batter power, and nothing else is. It is clearly responsive to all the various elements mentioned--none of which, however, can ever have an overnight, "step" effect on power or its measure, PF.

Nor does it "fail to explain" anything, and I can't see how anyone can say it does. What "explains" Hank Aaron or Babe Ruth or any power hitter? That they were competent batters with excellent lower-body strength. PF *measures*, not explains. But, as a measurement tool, it does show us whether there is even anything needing explanation.

A perhaps clearer tool is "relative PF", which takes a given man's PF as a ratio to the prevailing MLB-average PF year by year; I have a set of graphs of relative PF for the top-10 home-run hitters, and will be publishing them soon. But I can tell you now that there is nothing distinctive about modern-era players vs. "classic-era" players.

As to recovery rates, there is an entire page on the site that addresses those issues--check the site directory (linked at the top of every page).

More generally, I am willing and able (as time allows) to engage in dialogue on all these matters, but it helps a lot if folk making posts (and I do not by this mean any in particular, here or elsewhere) would first read the whole site before commenting on a paragraph or sentence here and there.

It seems concluding "steroids had little effect" is a bit of a leap. Another conclusion one could draw is that fewer people than assumed have used steroids. If, for example, Bonds, McGwire and other high-profile home run hitters were using because it had better results for their already prodigious talent but the vast majority of other hitters were not, then it wouldn't show up in the overall numbers, would it? The affects of better training, smaller ballparks, a livelier ball all would show up (like Walker says), but if the number of steroid users was smaller than assumed and limited more to the high-profile sluggers, then the evidence would be seen more anecdotally.

It's a plausible explanation, I think, for why those looking at the overall power numbers are arguing steroids have little effect, while those looking at individual seasons say the effect is clear.

Thanks for taking the time to respond here Eric, and I apologize for the phrase "explains nothing," which is inaccurate. I will review your site, but in the interim I have some questions based on what you have written here, and what Alan wrote in the Times (assuming it's an accurate reflection of your beliefs).

I wonder, first of all, did you suspect that there was little steroid effect before you began your investigation?

You write that PF "is an indication of the force with which a given batter, or set of batters, meets a well-struck ball on average." I ask, what is your evidence of this assertion? The force with which a ball is struck is a measure in ppsi. Why should I believe PF measures anything except what it says it measures, which is how many bases a player advances per hit, a number that might be controlled by any number many factors.

In his article, Alan essentially assigns to you the assumption that "any added strength hitters would get from steroids would not help them make solid contact with the ball, but only hit it farther when they do." Is your argument predicated on this assumption? What demonstrable evidence do you have that this is true? What if it's wrong? I recall reading that Bonds's alleged drug cocktail improved his eyesight.

Okay. I've looked over Eric's site a bit and it's pretty convincing. It goes a good way toward responding to the two central questions I posed above (though not all the way). If the question is, did PEDs have a major overall effect on baseball offense, there's a pretty convincing argument that the answer is no. But based on my cursory reading I remain extremely skeptical at the suggestion that PEDs have not been extremely beneficial to a certain segment of the baseball population.

Let's push this argument further: should the game, our national pastime, be encouraging usage? Should we ignore that many players flouted the law (and the rules of the game secondarily) to gain a supposed advantage (whether that advantage existed or not)?

Quotes like this from Walker are troublesome:

"A third issue is that investigations and discussions typically refer to "athletes", but in an overwhelming fraction of cases they really mean high-intensity weightlifters and bodybuilders, whose abuses of these substances typically far outdo what even professional athletes such as ballplayers may do (as evidenced by the classically freakish-looking bodies that such bodybuilders cultivate). Evidential results from "abusers" need care in evaluation: there is overuse and there is overuse, and to assume that what may be detected in gym fanatics necessarily represents what effects may occur in less rabid users is to tread thin ice indeed."

Does Walker know how much players used? If so, how? How is able to speak with such authority that the players didn't use disproportonately to their recommended dosages for medical purposes? Does he know how much their usage deviated from recommended patterns of usage for medical purposes when prescribed by a physician or in a hospital, something Walker goes out of his way to bold as a case where there are also adverse effects, ridiculously ignoring the idea that in a hospital steroid treatment may be done for explicit, case-specific purposes where their benefits are weighed by a physician against the risks they pose.

Walker seems to live in a world of moral equivalencies (witness the section that asks "do we ban cleats?", or the comparison of the "crime" of steroid use to the use of "clown guns"), equivalencies which don't help his case.

Some answers to questions asked:

"Did you suspect that there was little steroid effect before you began your investigation?"

Yes, from work I did many years ago for a page on my main baseball site, a page entitled "The Silly Ball" (it was written at a time people would still remember the A's slogan "BillyBall"). I was annoyed at how obvious the ball juicing of 1993 was and how very stupid were the things sportscasters and writers were saying about home runs and offense. I set out then to see if the data supported the position, and they did, solidly. Even then, there were early rumblings about steroids, and I mentioned them dismissively. But never underestimate the power of bigots looking for witches to burn.

"You write that PF 'is an indication of the force with which a given batter, or set of batters, meets a well-struck ball on average.' I ask, what is your evidence of this assertion? The force with which a ball is struck is a measure in ppsi. Why should I believe PF measures anything except what it says it measures, which is how many bases a player advances per hit, a number that might be controlled by any number many factors."

You can replace PF, if you like, with a simple ratio of home runs to hits, and the curves, given correct scaling, overlay just about perfectly, so the points some have raised here and there about speed contributing to doubles and triples are irrelevant. It is true that the force delivered to a ball is a physical measure, but it translates directly to the distance the ball will travel for a given launch angle. Given the simple assumption that launch-angle distributions don't change much over time, for one man or a league, then the number of home runs, given some fixed basis--which is to say hits, that datum which home runs are a fraction of--are going to go up in proportion to the average distances balls are travelling, which comes back to the average energy being imparted to them. Again: home runs are determined--on an average basis--solely by the average distances balls travel, which is solely a function of the average force with which they are struck. We don't simply count home runs because when balls get easier or harder to hit at all--as, for example, from changes in the effective size of the strike zone--the home-run total will vary with the hit total. it is the proportion of balls going for hits that are given enough energy to leave the park that interests us.

"'[A]ny added strength hitters would get from steroids would not help them make solid contact with the ball, but only hit it farther when they do.' Is your argument predicated on this assumption? What demonstrable evidence do you have that this is true? What if it's wrong? I recall reading that Bonds's alleged drug cocktail improved his eyesight"

An extensive examination--and please believe me, i don't use adjectives lightly--of the scientific literature discloses no indications that any PED can improve vision or reflex response time, and several assertions that they do not. If that is wrong, a lot of physicians are in for a big shock. These things are the reason the "Medical Effects" page of the site is so very heavily laden with quotations from the scientific literature.

"Does Walker know how much players used? If so, how?"

Eyes. I included on the site some images of what freako bodybuilders end up looking like expressly so that people could see that ballplayers are really quite far from achieving the sorts of effects possible to really heavy abusers.

Let's get clear here: I do not advocate or recommend PEDs for baseball players. To begin with, my firm belief, which I hope I have documented, is that they don't have any meaningful effect on baseball performance. I think the medical risks of using them are not grave, but what is the point of any risk at all for essentially no gain? My belief is that all parties here--MLB, the Players Association, even the writers and the public--ought to be seizing the "no effects" argument with both hands, because it is the first and only persuasive argument to offer players at both the major-league and minor-league levels as to why they should not even think of using the stuff. As Will Carroll says somewhere, 1000:1 odds against something bad are no comfort if you're the 1.

For sports in which PEDs can and almost surely do have some effects on results (such as events where the winner of a race event can be determined by one or two hundredths of a second, or in weight-lifting or other activities emphasizing upper-body strength), the issues are more complex. It is, though, hard to argue with the many professional medical ethicists, some of whose
work I present on the site (recall, there is an entire separate site page on ethical issues, which I wish those who talk about "moral equivalencies" would read through). Those men are highly skilled and experienced physicians who have specifically trained for and become world-wide recognized authorities on issues of medical ethics. As I say elsewhere, that doesn't make them divine oracles per se, but it does suggest that we might give their opining more weight than that of the boys down at the Dew Drop Inn.

As to the argument that even if MLB as a whole hasn't benefitted, a few select men have: I have been spending so much time responding to emails and comment posts all over the internet that I still haven't finished my upcoming new page on individual case studies, so what I say here is just a short few words. Comparing career relative PF graphs--meaning a man's actual PF divided by the MLB-average PF for each year (so we see how he exceeded norms, independent of what those norms were)--the graphs for the modern players on the all-time top-10 home-run list (Bonds, McGwire, Palmeiro) do not show any systematic differences from those of list members from earlier eras. The two men whose power performances lasted to the greatest age are classic-era figures, Aaron and Frank Robinson. I'll try to remember to post back here when that page finally goes up.

Finally, as to the Bonds "spike", spikes are a well-established phenomenon in baseball and--as one can see by perusing Professor DeVany's paper "Steroids, Home Runs and the Law of Genius" (as linked on the site)--are not a reflection of any extrinsic factor but simply something that happens in a big enough world. Just to throw out a few names in connection with power spikes, try Davy Johnson, Luis Gonzalez, Terry Steinbach, Lonnie Smith, and of course Roger Maris. The thing about a spike is that it's just that: up, then DOWN again just as fast. To attribute that to some magic chemical reflects incomplete consideration of the matter (and please, please, no one bring up the "well, that's when they stopped using it" garbage, because you can't have it coming and going).

Thank you for letting me vent a little.

Thanks Eric. As noted, I think the arguments on your site are pretty convincing, but I still find your "spike" argument unpersuasive. Spikes don't happen simply as random occurances on the statistical fringe. You mention Maris (and his 61). A blip, yes. But a blip in an expansion year. So we can point to at least one external controlling factor that contributed to the spike. I just don't see how you can look at the Bonds case, after the Chronicle reporting, and then say "random spike." It's not a credible argument. The performance patterns of certain players implicated in various investigations strongly suggest amplification by PEDs (and I mean the whole ball of wax, not just the steroid group), and this correlation undermines the "spike" argument.

It's interesting that we see PEDs having an enormous impact on other sports. You firmly believe they have little impact on baseball. My impression is that the margin between a AAA and an average ML player (nevermind the star) are very fine indeed, or sometimes a nonexistant matter of circumstance. I fear that's where a lot of PED use is happening. How positive can we be that there is absolutely no benefit from PED usage?

As it is, I would not argue with your argument that, in a cost/benefit analysis, PEDS are absolutely a horribly bad gamble.

To me, in any case, PED use in baseball begs broader, more uncomfortable questions about our narcissistic society as a whole. We scapegoat ballplayers who wish to chemically improve performance when billions are spent on a drug that begins with a V and its ilk and plastic surgery. We somehow expect baseball to represent values above and beyond our own, but it is a business as venal as any other. That hipocracy has always been there, but we've seen it grow. Say one thing. Practice another. We've seen it from ballplayers. We've seen it from the Commish. From our president. From ourselves.

Sorry for the confusing "V and its ilk" line there. Our spam filter was blocking the name of that substance:


V is so 2004. All the hipsters are using C-I-A-L-A nowadays.

I'm think I share YF's view of the possibility that PED's might not have actually helped performance. I've always been partial to the view but it's hard for me to get beyond certain individual spikes in performance. Of course, we know of certain players who seemingly gained no advantage from steroid use (Lawton, Alex Sanchez--althogh who knows what their baseline was?), but then you have the compelling statistical cases of Bonds and McGwire. Could it be that PED's work for players with a particular skill set and at a certain level? One possibility is that players like McGwire and Bonds knew actually how to use PED's better than others who were also using. Perhaps, they were more intelligent (for lack of a better word) in using these drugs to maximize their performance. And perhaps they were part of a minority of players who actually knew what they were doing with these drugs, and therefore a statistical study seeking to measure the general effect of steroids would only see little, if no effect.

By the way, here's one thing, YF, I could cite as an "external factor" that could go a long way in explaining Bonds's spike. Didn't he start using that high-tech arm guard (that propped his arm in perfect swinging position and made him fearless on the inside pitch) around the same time he started putting up Ruthian numbers?

"Spikes don't happen simply as random occurances on the statistical fringe." Looking this over, I see this is an erroneous statement. My intended meaning was, "Spikes are a natural phenomenon in an abstract statistical measure, but in a non-closed system (like baseball) spikes may also be the product of non-random causation."

I'm no expert in statistics, and these systems are complex. What Walker has presented on his site is pretty challenging for those of us with poor quantitative skills.

The comments to this entry are closed.

Search YFSF

Sports Gambling


schedule & standings