Honestly, that makes perfect sense to me. I haven't RTFA yet, so I'm hoping it has some answer to the question of what steroids do, or may do, for pitchers.

Probably nothing.

Baron, from the article:

"Tobin applied a similar, though less extensive, mechanical analysis to pitching and found that smaller impacts were possible. He calculated that a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent, or four to five miles per hours for a pitcher with a 90 mile per hour fastball. That translates to a reduction in earned run average of about 0.5 runs per game.

'That is enough to have a meaningful effect on the success of a pitcher, but it is not nearly as dramatic as the effects on home run production," says Tobin. "The unusual sensitivity of home run production to bat speed results in much more dramatic effects, and focuses attention disproportionately on the hitters.'"

They may? Cool.

Actually, for this pseudoscience article they ought to take the man's degree away and revoke the accreditation of the university, unless it's a parody.

"He calculated that a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent"

How did he calculate that? I call bull

Actually, for this pseudoscience article they ought to take the man's degree away and revoke the accreditation of the university, unless it's a parody.

Indeed. What's a Tufts physicist talking standard deviations, outliers, and tails good for when we have the anecdotes of Tom House and Jim Leyritz?

I just hope this doesn't get out. Players might start using.

I just hope this doesn't get out. Players might start using.

dicks chig the long ball

Since so few people seem to RTFA, people like EddieA might want to look at the following paragraphs before passing instant judgment:

Tobin is quick to acknowledge that athletes in many sports today achieve at a higher level than athletes of the past, and that this trend is not evidence of cheating. He also points out that many other changes, including adjustments in ballpark dimensions, league expansions, entry of African-American athletes, and lowering of the pitcher's mound, could affect major league batting—although he says that none of those changes coincide with the sudden burst of home run production in the mid-1990s.

"Physics cannot tell us whether a particular home run was steroid-assisted, or even whether an extraordinary individual performance indicates the use of illicit means," says Tobin.

But analysis of the physics, combined with physiology, yields telling results. "These results certainly do not prove that recent performances are tainted, but they suggest that some suspicion is reasonable," he concludes.

Of course the fact that Tobin is heavily credentialed isn't necessarily all that conclusive, since I'm sure that other equally credentialed physicists might disagree with his conclusions, but this is certainly not "psuedoscience."

#10

I did RTFA. While I wouldn't call it pseudoscience, there are IMO way too many concrete assertions and way too little data backing up those assertions in the article.

Tobin reviewed previous studies of the effect of steroid use and concluded that muscle mass, the force exerted by those muscles and the kinetic energy of the bat could each be increased by about 10 percent through the use of steroids. According to his calculations, the speed of the bat as it strikes the pitched ball will be about 5 percent higher than without the use of steroids and the speed of the ball as it leaves the bat will be about 4 percent higher.

I would really like to see his study, and how he came up with those concrete numbers. Has Tobin been conducting a study with admitted steroid-using MLB players as his subjects?

Andy, just because the professor is smart enough to nuance his claim doesn't mean that he knows what he's doing.

In particular, the quote I highlighted before ("a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent") ... it seems like it would take a life's work to arrive at a number like that. Experiments with control groups would be extraordinarily difficult to arrange, and require years of evaluation. Were there experiments, or is this all theoretical? Did he ever have a coach that told him long toss would help him more than lifting weights would? That sentence smells like bulls--t so bad.

In addition, if you RTFA, you get this whopper:

According to Tobin, the explosion in home runs coincides with the dawn of the "steroid era" in sports in the mid-1990s, and that surge quickly dropped to historic levels in 2003, when Major League Baseball instituted steroid testing.

Huh? In 2003 there were more HRs hit than in 2002. I don't even know what he's talking about.

Has anyone actually read the study? This is a one-page press release from the Tufts P.R. office.

Unanswered in the release are questions like: if he thinks there's a 5% increase in pitch speed due to steroids, how much of that is transferred into ball speed off the bat? Is it part of the 4% increase he's attributing to steroids, or in addition to it?

Actually, for this pseudoscience article they ought to take the man's degree away and revoke the accreditation of the university, unless it's a parody.

The irony of this comment is sublime enough to be ridiculous.

If the methods aren't published, it's pseudoscience. Even if it's right.

EDIT: Sorry - I didn't RTFA until after my post (pseudoposting). One would be well advised not to believe a word of any press release on science. The MSM and most bloggers screw it up more than Joe Morgan explaining VORP. If Prof. Tobin's paper is published and contains a methods section that could be replicated or refined, it's science. Even if it's wrong. I was mostly reacting to far too many "reports" in all fields of science that don't bring a full description of methods and supporting data.

Steroid use by a Major League Baseball slugger may... boost home run production by 50 percent or more

I knew it! I knew Jerry Hairston Jr. was only a 1-2 home run guy!

But pondering Ozzie Canseco's adjusted power numbers just fried my mind.

Has anyone actually read the study?

If the methods aren't published, it's pseudoscience.

Paragraph 2 of the linked article: "Tobin, a specialist in condensed matter physics with a long-time interest in the physics of baseball, will publish his paper "On the potential of a chemical Bonds: Possible effects of steroids on home run production in baseball" in an upcoming issue of the American Journal of Physics."

It's not even as complicated as the professor makes it, though he's spot on in his statistical observations.

A small percentage increase in distance hit will lead to a much larger impact on home run numbers. Imagine 10,000 balls hit toward the 385 foot sign at Yankee Stadium. No ball hit 380 feet will be a home run. Every ball hit 395 feet -- an increase of less than 4% in distance -- will be a home run.

Steroid use by a Major League Baseball slugger may produce only modest increases in muscle mass and bat and ball speed but still boost home run production by 50 percent or more, according to a new study by Tufts University physicist Roger Tobin.

Yeah well, weight training without steroids or other PEDs may also lead to modest increases in muscle mass

I don't know if there's anything new here.

He's quantifying the affect of steroids on batted ball speed, which is, by his own admission, only a educated guess.

He's then demonstrating that a faster batted ball leads to more home runs, which doesn't take a condensed matter physicist to figure out.

I do like his explanation about the distribution of batted balls and how nudging the midpoint to the right can dramatically impact the tail.

But pondering Ozzie Canseco's adjusted power numbers just fried my mind.

Hey, there's your control case. Identical twins, one juiced, one didn't(?). One hit 462 home runs, one hit 0.

Also, the A's drafted Jose in the 15th round, while the Yankees drafted Ozzie in the 2nd round. Heh.

Imagine 10,000 balls hit toward the 385 foot sign at Yankee Stadium. No ball hit 380 feet will be a home run. Every ball hit 395 feet -- an increase of less than 4% in distance -- will be a home run.

Does data exist on fly-ball distances? How fine is such data collected by Stats or BIS? Has anybody looked at how the distribution of fly balls has changed over time?

He calculated that a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent, or four to five miles per hours for a pitcher with a 90 mile per hour fastball.

Likewise, does data exist on pitch speeds? It sure seems like there are a lot more guys who throw 90+ and 95+ than when I was a kid (1970s - early '80s) but is there data around to check such a thing?

This seems like more of an empirical question than a theoretical one (I would think the theory would come once you've established that fly balls are actually going farther in trying to determine whether that can best be explained by steroids or by other factors).

What a shocker: Knee jerk dismissals from the Kool Aid Krowd, without even reading the link or the article which none of us yet can read, but which has apparently been deemed scientific enough to be published in a reputable academic journal.

The professor's study ought to be exactly what anyone with any interest in the subject (on any side of the debate) desires. Is it the last word? No, of course not. But why not wait and see what and how the man argues before calling for his firing?

And the fact that it's "theoretical" (which I assume it is) is no knock against it, either.

He also points out that many other changes, including adjustments in ballpark dimensions, league expansions, entry of African-American athletes, and lowering of the pitcher's mound, could affect major league batting—although he says that none of those changes coincide with the sudden burst of home run production in the mid-1990s.

Really?

And the fact that it's "theoretical" (which I assume it is) is no knock against it, either.

I didn't mean to suggest that it is. Only that I'm more interested in the empirical question of how many extra home runs were actually hit. I think the theory is obviously going to be important in helping to make sense of the empirical data, but you're going to need to look at empirical data to answer the question.

You sure can swing and pitch harder when you miss too.

This seems like more of an empirical question than a theoretical one (I would think the theory would come once you've established that fly balls are actually going farther in trying to determine whether that can best be explained by steroids or by other factors).

Good point. And one of those empirical studies might consist of tracking the distance of Barry Bonds's home runs over the years. Is there enough contemporary data (newspaper accounts, tapes, etc.) for that to be possible?

I've read that nearly all of his tape measure jobs took place post-1998, which would be very interesting if true, but the question is: Is that true? And can we find that out?

And before someone jumps in, I should add that if this were demonstrated to be not true---if Bonds's average home run distance had not increased over the past nine years---then that would be interesting as well.

And I would think that this is the sort of information that anyone without an axe to grind might welcome.

KS: How could you possibly get answers to that question?

What a shocker: Knee jerk dismissals from the Kool Aid Krowd,

True. But that goes both directions. Intuitively, the claim seems to be reasonable and Tobin's qualifications seem to be good. My guess is that my thin knowledge base will leave me unable to comment intelligently on the validity of the study itself, if and when people here can read it, but I will try to check it out and listen to informed opinions.

I don't like using a pun on Bonds' name in the title, in that it isn't particularly funny, and it could be construed as indicating an agenda, as does the quote excerpted by PreservedFish in #12.

True. But that goes both directions.

Do you see Andy or I claiming anything other than this is interesting and it will be interesting to read his article?

I'm going to concede, before any objections, that the rest of this comment is going to be unfair. Because we can't read the actual study yet. But what it sounds like to me is that Mr Tobin had one good idea (the thought about standard distributions), an idea that might be accurate and is worth talking about. In order to turn this idea into a publishable work, he fleshed it out with all of this 4% bat speed increase stuff, comments about the "steroid era," etc, all of which is either taken from other studies or is totally theoretical or is just wrong. Judging from this press release, a lot of that "fluff" is probably crap. But that doesn't necessarily mean that his original idea is not.

I don't like using a pun on Bonds' name in the title,

I agree, but I bet it helped him get funding and/or attention.

But pondering Ozzie Canseco's adjusted power numbers just fried my mind.
Hey, there's your control case. Identical twins, one juiced, one didn't(?). One hit 462 home runs, one hit 0.

Jose wrote about brother Ozzie's steroid use in Juiced. Ozzie also did some jail time for steroid possession.

Andy, I'd think all of Bonds' games are somewhere on videotape. We're talking about a lot of flyballs though. Sheesh. I agree that would be an interesting study.

I wasn't criticizing his conclusion. Pseudoscience can be correct. He may be dead on, but if he doesn't publish his methods such that someone could, if they wished and had proper training, replicate it, then it isn't science. That by itself isn't enough to say he's wrong.

Do you see Andy or I claiming anything other than this is interesting and it will be interesting to read his article?

I was mostly referring to SBB. Additionally, neither you nor Andy is approaching this by suggesting possible faults in the study, as other people are. And your admonishing people and labeling them--"The Kool Aid Krowd"--is hardly news.

Almost everyone's posts--mine included--are generally fairly predictable in steroids threads.

Tobin reviewed previous studies of the effect of steroid use and concluded that muscle mass, the force exerted by those muscles and the kinetic energy of the bat could each be increased by about 10 percent through the use of steroids. According to his calculations, the speed of the bat as it strikes the pitched ball will be about 5 percent higher than without the use of steroids and the speed of the ball as it leaves the bat will be about 4 percent higher.

No matter how smart he is and how reasonable he is trying to be in his assumptions and interpretations, the fact remains that he is, in essence, pretending to have isolated a variable that he has not isolated. That's not a criticism; we all do this sort of thing all the time, and as long as we acknowledge that it's not the same as really doing a controlled study, it's fine for the sake of argument. But the question is unanswerable in any meaningful scientific sense.

KS: How could you possibly get answers to that question?

Well, it wouldn't be easy and I don't know if the data even exists to do it, but thinking off the top of my head: The key would be to put together a distribution of fly balls hit by distance and see how that distribution has changed over time.

Has the mean increased over time? If so, how does that translate to an increase in home runs (that is, this should help you to control for the impact of smaller ballparks; it might also address the possibility of a juiced ball, as that will affect steroid users and non-steroid users alike)?

If the assumption is that steroid use wasn't ubiquitous, then you'd expect the use of steroids by some players to also affect the distribution itself - I think it should skew to the right. I'm not sure how one could do something with this, but if you think of two distributions, a non-steroid user distribution w/ mean x and stdev s, and a user distribution w/ mean x+i and stdev s, if you have some best guess for i, might empirical data help you estimate what % of hitters were using (or, alternately, if you have some best guess for % of users, you could maybe try to tease out i).

None of this would be easy, of course.

Alternately, Andy's idea would be interesting: look at results for specific players. For those (very few) that we think we know when they started using, would that tell us anything? I'm uncomfortable morally with doing something like looking at changes in fly-ball distances for player Z and concluding that there's a P% chance that he used steroids, although I think that's the sort of conclusions that one might be able to draw. Which would be interesting from an academic perspective, although I wouldn't want to see this sort of information used to punish players based on mere probabilities.

Additionally, neither you nor Andy are approaching this by suggesting possible faults in the study, as other people are.

We're suggesting that we all will benefit from reading the study. Why would we suggest that the study may have possible faults - that's true of every study and doesn't require mention. We can play "moral equivalence" game, but if you find our statements equivalent to guys calling this pseudoscience and disparaging the researcher, I'll suggest you're off base.

Why would we suggest that the study may have possible faults - that's true of every study and doesn't require mention.

Take that up with the numerous posters who have done so throughout the thread then.

but if you find our statements equivalent to guys calling this pseudoscience and disparaging the researcher, I'll suggest you're off base.

Fair enough. But, "pseudoscience" and "Kool Aid Krowd" seem relatively similar to me. And it is not a moral issue, nor is it a game.

According to Tobin, the explosion in home runs coincides with the dawn of the "steroid era" in sports in the mid-1990s, and that surge quickly dropped to historic levels in 2003, when Major League Baseball instituted steroid testing.

That sounds good, but HRs didn't really drop since 2003 -- they've been within the same amounts, by and large, as they were during 1998-2002.

I also find it curious that, IMO, the most vital question of how just how much increased muscle mass, etc. affects the speed of the bat, pitch, etc. as well as whether it affects things such as flexibility. Tobin, a physicist, seems to be taking these factors as givens (10% in force and energy on the bat, for instance), a question I don't believe is a matter of physics so much as medicine.

OTOH, if indeed PEDs can be mesured to have these effects (and I strongly suspect that they vary from one PED to another), than I would certainly think that Tobin is well qualified to calculate the back half of the equation -- namely, if a pitch has X energy and a bat has Y, what affect with that have on a batted ball.

Even here, though, he could only make general assessments, of course. Pitchers do throw breaking balls, of course, and there are matters such as barometric pressure, wind conditions, and other variables at play.

I'm not really sure what measuring Barry Bonds flyballs is going to prove or disprove. This isn't like firing an artillery shell out of a cannon where we can control for angle and amount of powder. It is a human being hitting a moving target with a wooden stick. Unless you can zero in on where exactly the ball meets the bat, at what angle the ball approaches, at what angle the bat approaches, speed of the ball, and liveliness of the ball (among other things) whatever you find is going to be BS.

I'm not really sure what measuring Barry Bonds flyballs is going to prove or disprove. This isn't like firing an artillery shell out of a cannon where we can control for angle and amount of powder. It is a human being hitting a moving target with a wooden stick. Unless you can zero in on where exactly the ball meets the bat, at what angle the ball approaches, at what angle the bat approaches, speed of the ball, and liveliness of the ball (among other things) whatever you find is going to be BS.

I'm a social scientist by trade, so I'm more used to looking at the data that you can get and seeing if it tells you anything, moreso than forming a rigorous hypothesis and then doing an experiment that controls for all other possible factors. And oftentimes, the data tells you nothing.

I do think that looking only at Bonds's fly balls may not tell you anything that we don't already know - he hit a lot more home runs from 2000 on. Looking at major-league players as a whole or at groups of players about whom we know something of their usage (ideally, both known users and known non-users; in a way, the latter's the more problematic group to identify) may tell us something.

Also, as a social scientist, I think in terms of probabilities, not "proof". It is not possible to "prove" that Barry Bonds used steroids based on his baseball statistics.

We're suggesting that we all will benefit from reading the study. Why would we suggest that the study may have possible faults - that's true of every study and doesn't require mention. We can play "moral equivalence" game, but if you find our statements equivalent to guys calling this pseudoscience and disparaging the researcher, I'll suggest you're off base.

One guy labelled it "pseudoscience." Not "guys." Not "Kool Aid Krowd."

Tufts put out a press release in advance of a study's actual publication. They pimped for coverage with quotes that raise immediate questions in the absence of that study. Despite your attempt to indict everyone here who asked those questions, there's nothing wrong with asking them.

It's exactly the kind of reaction Tufts P.R. office wanted anyway.

I came up with what I think are reasonable numbers and I don't see anything close to a 50% increase, if we are dealing with someone who is already a slugger.

Here's my assumptions: Normal distribution, average flyball distance = 250 feet, HR distance = 380 feet. I assume 1 standard deviation = 130 feet, as that would give a HR% of 16% on flyballs, not an uncommon percent for a guy who is already a power hitter.

Now give him steroids and assume that he can hit the ball 4% farther. His average distance is now 260, one SD is 395, so a homerun is anything hit over 0.9 SD's instead of one. He know hits homers on 18.4% of his flyballs.

Given 150 flyballs, he hits 28 instead of 24.

To get a 50% increase steroids would have to increase your average distance by more than 4%, or homeruns would have to be a much more rare occurrence than they are - perhaps a weaker player could go from 6 homer to 9 in this model.

What the study suggests is that adding muscle mass may increase the number of home runs.

In other words, bigger guys will hit more home runs.

Since I have never heard anyone suggest this before, it is indeed ground-breaking.

Of course, since steroids are the only way to add muscle mass, all the increase must be due to steroids. Now if there was some other way to add muscle mass, like exercise or legal supplements, or such. Too bad nobody ever thought of that.

It will be interesting to see if he can demonstrate the difference between adding 2 pounds or 5 pounds or 10 pounds or 15 pounds. And where the placement of the muscles is -- arms, back, legs, etc.

Those will be very interesting figures.

Also, this study would only address ONE potential bebefit to steroids and home run hitting, the extra distance that increased mass can generate.

But if increased mass and increased bat speed allow for a hitter to wait longer to recognize a pitch and hit it "squarer" that is also a benefit that can lead to increased home runs.

Imagine 10,000 balls hit toward the 385 foot sign at Yankee Stadium. No ball hit 380 feet will be a home run. Every ball hit 395 feet -- an increase of less than 4% in distance -- will be a home run.

Yes...and the infield fly rule will become the outfield fly rule.

Here's my assumptions: Normal distribution, average flyball distance = 250 feet, HR distance = 380 feet. I assume 1 standard deviation = 130 feet, as that would give a HR% of 16% on flyballs, not an uncommon percent for a guy who is already a power hitter.

I don't think flyball distance is normally distributed. It's got a high skewness.

Imagine 10,000 balls hit toward the 385 foot sign at Yankee Stadium.

Like Bobby Abreu doesn't have enough problems as it is. Now he has to try and catch 10,000 baseballs.

If strength was the dominant characteristic, Magnus from the World's Strongest Man contests would be batting cleanup for Steinbrenner.

My guess (thoroughly unscientific) is that a combination of HGH and steroids is as magic as Viagra and Pam Anderson.

I do think that looking only at Bonds's fly balls may not tell you anything that we don't already know - he hit a lot more home runs from 2000 on.

Just to make myself clear, I was specifically attempting to discover the sort of data that might correspond to increased muscle mass: home run distance. As in average distance, plus the relative numbers of his longest home runs in each year. I don't see how that could possibly not be at least somewhat related to muscle mass.

(And yes, Srul, we know that steroids aren't the only way to increase that muscle mass. What a revelation.)

It does seem that this would correlate with Bonds's home run rate, since presumably if Bonds's average home run distance increased by (top of my head number) 15 feet, then a certain percentage of warning track fly balls would get over the fence. But I wouldn't think that adding this data to the tracking of the distance of actual home runs would change the overall implications of the results of a strictly home run distance study.

I don't see anything wrong about beginning this sort of a study with Bonds, since we have had evidence of when his juicing began. And if the results show that his hyped up home run distances didn't in fact correlate to his steroid years, then that would seem to me to be a powerful piece of evidence that lots of us may have been wrong, either about Bonds or the PE effects of steroids, or both. And I'd be the first to admit it, since this has never been a crusade against any individual player on my part.

Also, as a social scientist, I think in terms of probabilities, not "proof". It is not possible to "prove" that Barry Bonds used steroids based on his baseball statistics.

As a layman, my bias is towards publishing all the facts that one can, and then letting people argue about what they mean or don't mean. Whether it hurts or helps a particular player's reputation will have more to do with what may be discovered than anything else. Whether one calls his conclusions "proof" or "probability" (with percentages added if warranted) isn't all that important, since both assertions are going to be subject to the same sort of cross-examination anyway.

Of course there are other factors at work, such as the trajectory of his swing. But certainly there are game tapes galore to take this into account.

Everything I've written about Bonds and steroids in general has been based on my interpretation of what I've seen put out there. And so far it leads me to the conclusion that Bonds juiced, and that it helped him pad his home run totals beyond what they would ever have been without the juicing. This is a pretty standard conclusion these days, although the estimates of the padding obviously vary.

But I also am more than willing to take new empirical evidence about Bonds into consideration, and if that evidence points contrary to what I've thought so far, so be it. Which is why I think that this sort of a study might be very interesting.

I don't think flyball distance is normally distributed. It's got a high skewness.

You're probably right, I'm not sure what the distribution would like like, but I don't know if it would change the results much. The professor seems to be talking about something pretty close to a normal distribution:

"In most any statistical distribution -- of people's heights, SAT scores, or how far baseballs are hit -- there's a large bump where most of the values fall, with the graph falling rapidly as you move away from that region in either direction toward the rarer values,"

Andy,
When you have an answer it is very easy to go looking for questions that "proves" the answer. Virtually anything we do in regards to Barry Bonds is going to come back as proof of steroid use. Even if for some oddball reason whatever did happen had nothing to do with steroids. Why? For me it is because we already "know" the answer. We know Bonds did steroid therefore anything unusual we find in Bonds will indicative of steroid use.


For instance this guy from Tufts took the steroid slant and even decided ahead of time to rule out all these other factors. Which was truly bizarre as already pointed out. This guy is not the first to do it we have had many many studies in which the conclusion was steroids and usually all the other variables get brushed aside into one wide sweeping statement in which the researcher recognizes that these other variables play a role but couldn't possibly play this large of role therefore it is steroids.

I really don't see how any precision can be achieved with the mind boggling amount of variables involved. Now, it seems clear to me that steroids aid athletic endeavor, but to what degree I don't think we'll ever have any idea. And, at this point, I don't think PEDs can be stopped. We'll each one of us have to decide whether we're still going to care about sports in the future knowing that what we're seeing isn't "natural" or "un-aided" or as "God intended" or any other way you want to phrase it.

Speaking of which--hey all you history geeks, has sports also been a big deal in Western culture? Were there ever times when nobody gave a crap because they were too busy going to plays or arguing about angels on the head of a pin? Is the exaltation of sports a recent phenomenon rising with the rise of the middle class and industrialization? Huh huh huh?

I wonder if Tobin talked woth Andy Andres at all. He's another guy on the Tufts faculty and teaches a course in sabermetrics. THis sounds similar to a presentation that Andres made at a SABR regional in Boston a around MLK Day in '05.

Andy,
When you have an answer it is very easy to go looking for questions that "proves" the answer. Virtually anything we do in regards to Barry Bonds is going to come back as proof of steroid use.

But if we're interested in testing our hypotheses, then the more data we can come up with, the better.

And what if my proposed HR distance study showed that Bonds had not increased his home run distance in his steroid years?

To me that would lead to one of two possibly inescapable conclusions.

Either he really hadn't been juicing, or....

The juicing had no effect.

Which would certainly force me to reconsider my opinions about both Bonds and steroids.

And if anyone had the time, money and inclination, he could also do similar studies of any other player he wanted. I mention Bonds first because of his high profile both as a juicer and as an already established Grade A all-time hitter, but I'm not saying to restrict a study like this to Bonds alone.

WRT the Tufts guy, I'm not attacking or defending his study until I've seen it, beyond defending his right to publish it and my curiosity to see what it might say. But my idea for looking at Bonds's home run distances was completely independent of that.

Speaking of which--hey all you history geeks, has sports also been a big deal in Western culture? Were there ever times when nobody gave a crap because they were too busy going to plays or arguing about angels on the head of a pin? Is the exaltation of sports a recent phenomenon rising with the rise of the middle class and industrialization? Huh huh huh?

Depends what you mean by Sports.

Chariot racing was HUGE in Rome, and later in the Byzantine Empire. Gladitorial games were pretty big too. During the Dark and Middle Ages, I thing people were too busy dodging Huns and Vikings. Jousting and combat games were big in the late Middle Ages, part sport, part training.
I think it has to do with prosperous urban societies, though I'm no expert.

He also points out that many other changes, including adjustments in ballpark dimensions, league expansions, entry of African-American athletes, and lowering of the pitcher's mound, could affect major league batting—although he says that none of those changes coincide with the sudden burst of home run production in the mid-1990s.

Really?

Lets look at this one at a time.

Ballpark dimensions: Have gotten deeper an avg of 4-5' over the past 40 years. Especially when the last wave of ballpark construction kicked in. Mostly deeper in alleys and poles, shorter in center.

League expansion: Two expansions in this period. Obviously a help.

Black athletes: Why did he bring this up? While the talent pool has been growing, Japan didn't just open up at once.

Mound: That was what, 1969?

This article was not too convincing to me, but worth looking at when the data is available. I think there are too may factors to just point out one that caused the scoring jump of the 90's.

No matter how smart he is and how reasonable he is trying to be in his assumptions and interpretations, the fact remains that he is, in essence, pretending to have isolated a variable that he has not isolated. That's not a criticism; we all do this sort of thing all the time, and as long as we acknowledge that it's not the same as really doing a controlled study, it's fine for the sake of argument. But the question is unanswerable in any meaningful scientific sense.

Let's assume for the sake of argument that Tobin isn't just talking out of his butt on the muscle mass to distance phenomenon, and let's also assume he hasn't spent years doing secret detailed studies with real major leaguers.

Wouldn't the muscle mass business have to be based on some study of major leaguers? He'd have to take their weight, infer from that (and perhaps body type, and/or body fat percentages, assuming those figures are available or can be usefully generalized) the amount of muscle in a body, estimate muscle mass, then do a comparative study: existing players with more muscle mass and the distance of their hits/home runs versus existing players with less muscle mass and the distance of their hits/home runs.

How else would you do it?

I don't see anything wrong about beginning this sort of a study with Bonds, since we have had evidence of when his juicing began. And if the results show that his hyped up home run distances didn't in fact correlate to his steroid years, then that would seem to me to be a powerful piece of evidence that lots of us may have been wrong, either about Bonds or the PE effects of steroids, or both. And I'd be the first to admit it, since this has never been a crusade against any individual player on my part.

There's nothing wrong with the study per se, Andy. But as others have noted, there are simply too many variables. You can tell us that Bonds started hitting fly balls x% farther on average in 2000, but you'll never be able to tell us why with any scientific certainty. You would first need to go back in time and have Bonds spend the 1999/2000 off-season working his tail off in the wight room, without using any supplements, and then play a season to measure the effect of that. Next, crank up the time machine again and add creatine and any other legal supplements you like to the off-season regimen and replay the season to measure the incremental benefit of those changes. Finally, rewind once again and add the cream and clear this time. And even though my little thought experiment "controls" for age, there are still about a bazillion other confounding factors that I've left out.

Andy, I'd think all of Bonds' games are somewhere on videotape. We're talking about a lot of flyballs though. Sheesh. I agree that would be an interesting study.

I wasn't criticizing his conclusion. Pseudoscience can be correct. He may be dead on, but if he doesn't publish his methods such that someone could, if they wished and had proper training, replicate it, then it isn't science. That by itself isn't enough to say he's wrong.

I'll briefly toot my own horn by noting that I've asked the Bonds "distance through videotape" question several times. It seems eminently doable.

I'll add the observation that not only could you get distances from videotape, but you could also work backwards to bat speed by going frame by frame and getting velocity off the bat -- essentially the same thing.

Golf ball distance was limited by the USGA at one time (I'm not sure if it is anymore) by essentially these means -- a robot hit the ball and if its velocity off the club was too high, it was an illegal ball.

Anybody know if a 4% increase in speed off bat = 4% increase in flyball distance? I'll have to play around with hittracker.

Also, I wonder if his "50%" headline and the much smaller calculation I came up with above has to do with this:

This disproportionate effect arises because home runs are relatively rare events that occur on the "tail of the range distribution" of batted balls.

You have to look at flyballs, not batted balls, and by doing so homeruns become much less of a rare event. Even Conan isn't going to hit a groundball hard enough to turn it into a homerun. Willie Wilson might on astroturf, but we aren't talking about inside-the-parkers here.

Wouldn't the muscle mass business have to be based on some study of major leaguers? ... How else would you do it?

I doubt very much that his study involved major leaguers in any way. The press release says that he reviewed previously published studies of steroid benefits to come up with those percentages. I'm not aware of any published studies of steroid effects that used major league baseball players as test subjects. Presumably he used whatever studies he could find of healthy, young, well-conditioned males to estimate the average benefits in terms of muscle mass and efficiency.

I'm not really sure what measuring Barry Bonds flyballs is going to prove or disprove. This isn't like firing an artillery shell out of a cannon where we can control for angle and amount of powder. It is a human being hitting a moving target with a wooden stick. Unless you can zero in on where exactly the ball meets the bat, at what angle the ball approaches, at what angle the bat approaches, speed of the ball, and liveliness of the ball (among other things) whatever you find is going to be BS.

There's no reason you couldn't do that with video. If you can get all that info with the Zapruder tape, you can get it with a baseball swing.

It must be pointed out, in fairness to those who disagree with me on the matter, that there's a helluva lot more to bat speed than strength.

I have just read the study and here is a brief summary: it's possible for steroid use to explain the rise in home runs. He also explains why steroids asymmetrically affect hitters and pitchers. This is useful information, and it's a study that needed to be done, but I already believed this.

If Bonds changed his approach to hit homers, how much was swing v.s. muscle?
What does his GB/FB graph look like over time?

What does his GB/FB graph look like over time?

He definitely increased his FB percentage over time:

Year---GB/FB
1987---1.13
1988---0.85
1989---1.03
1990---0.83
1991---0.87
1992---0.74
1993---0.75
1994---0.72
1995---0.72
1996---0.71
1997---0.76
1998---0.63
1999---0.62
2000---0.57
2001---0.56
2002---0.65
2003---0.65
2004---0.75
2005---0.50
2006---0.66
2007---0.97

Add to what I've posted a couple times before by adjusting post and pre 1998 for walk/AB ratio, HR ratio, and GB/FB ratio. Without GB/FB, Bonds is at somewhere north of 670; he's probably near 700 if you add the GB/FB change (which the objective among us agree is interesting).

Imagine 10,000 balls hit toward the 385 foot sign at Yankee Stadium.

Like Bobby Abreu doesn't have enough problems as it is. Now he has to try and catch 10,000 baseballs.

RDF Mr. Murphy.

Best Regards

John

i think he under-estimates the effects steroids and HGH can have on a pitchers performance.

Rough and quick estimate here:

500 ABs with 140 Ks is 360 balls put in play. In his lowest GB ratio year, 2001, 64.1 out of every hundred balls in play were fly balls, giving him on the assumptive numbers used here, 231 fly balls.

In a year like 1997, only 57% of the balls in play were fly balls which, out of 360, would be 205 fly balls.

So I'm thinking roughly 26 extra fly balls in the high fly ball years and in a 40HR year around 1 HR per 5 fly balls, so the fly ball effect alone would give Bonds somewhere from 5-7 extra HRs. The 73 would be 66-68 otherwise ... still pretty suspicious and still an outlier.

I thought the numbers would be a little better for Bonds, but they are what they are.

Make of it what you will.

i think he under-estimates the effects steroids and HGH can have on a pitchers performance.

Steroid use by pitchers enhances home runs. Everything else equal a ball thrown harder will fly further when hit.

The stats/features of the steroid era reflect exactly what one would predict -- harder pitches, more strikeouts, more home runs.

500 ABs with 140 Ks is 360 balls put in play. In his lowest GB ratio year, 2001, 64.1 out of every hundred balls in play were fly balls, giving him on the assumptive numbers used here, 231 fly balls.

In a year like 1997, only 57% of the balls in play were fly balls which, out of 360, would be 205 fly balls.

So I'm thinking roughly 26 extra fly balls in the high fly ball years. Make of it what you will.

We had a thread maybe 3-4 weeks ago that got into this. Bonds actually didn't increase the raw number of fly balls that he hit very much at all in his peak years of 2001 - 05 because he started drawing those absurdly high numbers of walks, so he hit a higher percentage of a smaller number of balls-in-play. In 1997, he had 445 non-K ABs, whereas in 2001, he only had 383 non-K ABs. ESPN lists total FB and GB numbers for players by season.

I don't see anything wrong about beginning this sort of a study with Bonds, since we have had evidence of when his juicing began. And if the results show that his hyped up home run distances didn't in fact correlate to his steroid years, then that would seem to me to be a powerful piece of evidence that lots of us may have been wrong, either about Bonds or the PE effects of steroids, or both. And I'd be the first to admit it, since this has never been a crusade against any individual player on my part.

There's nothing wrong with the study per se, Andy. But as others have noted, there are simply too many variables. You can tell us that Bonds started hitting fly balls x% farther on average in 2000, but you'll never be able to tell us why with any scientific certainty. You would first need to go back in time and have Bonds spend the 1999/2000 off-season working his tail off in the wight room, without using any supplements, and then play a season to measure the effect of that. Next, crank up the time machine again and add creatine and any other legal supplements you like to the off-season regimen and replay the season to measure the incremental benefit of those changes. Finally, rewind once again and add the cream and clear this time. And even though my little thought experiment "controls" for age, there are still about a bazillion other confounding factors that I've left out.

I'd have two responses to this:

Depends on how much the differences were (if in fact there are any at all), and

Depends on what level of conclusiveness you're trying to demonstrate.

To take the extremes here, a tiny or nonexistent difference in those home runs distances would suggest the strong possibility that either Bonds or steroids (or both) have been unfairly maligned, whereas a large jump in average distance combined with a finding that nearly all of his longest home runs occurred during his steroid years would suggest the strong possibility that not only was Bonds juicing, but that the steroids added to his home run power.

Note my emphasis in both cases on "possibility." I do this both because of the inherent imprecision of any such findings, and because one person's "strong possibility" is another person's "proof." And we can go back and forth on that for years without getting anywhere. Much as everyone here hates to admit it, there's always going to be a fair amount of subjectivity entering into our judgments, based not only on what we may think of particular players, or even of particular drugs and our differing concepts of sportsmanship, but also because some of us think more like lawyers (with strict rules of evidence) or scientists (with high standards of "proof"), while others among us are simply trying to make a good educated guess about probabilities, based on whatever credible evidence we might find. Obviously I'd put myself in the latter category, but the former is certainly an equally valid way of looking at things, as long as it's made clear what is and is not being claimed.

RDF Mr. Murphy.

A little bit of you rubs off every now and then, John. ;-)

There's also a lot more to fly ball length than bat speed and strength. You also have to take into account the quality and angle of the contact. This isn't an argument against the effectiveness of steroids, it is an argument against flyball length as a proxy for the effectiveness of steroids.

Steroid use by pitchers enhances home runs. Everything else equal a ball thrown harder will fly further when hit.

The stats/features of the steroid era reflect exactly what one would predict -- harder pitches, more strikeouts, more home runs.

maybe, but it still means better performance by the pitchers. possibly averaging more strikesouts per game.

To take the extremes here, a tiny or nonexistent difference in those home runs distances would suggest the strong possibility that either Bonds or steroids (or both) have been unfairly maligned, whereas a large jump in average distance combined with a finding that nearly all of his longest home runs occurred during his steroid years would suggest the strong possibility that not only was Bonds juicing, but that the steroids added to his home run power.

I don't see how this gets us anywhere. We already know he hit more home runs. Let's assume the distances were longer (and I don't know how you plan to get accurate distance measurements). So what? The question is still what caused these things, and we're no closer to showing that steroids did.

Well, Ray, if the hypothetical increased distances correlated with the seasons that we know he was juicing, then some people might read a bit more into that than you would.

I find it somewhat droll that amidst all the charges of creationism, I'm the one who seems interested in discovering new empirical evidence, and the anti-creationists seem to want to dismiss it before it's even gathered. I'm certainly not saying that whatever findings we might come up with would constitute 100% conclusive proof, but to say in advance "I don't see how this gets us anywhere" doesn't strike me as particularly "scientific." It sounds to me more as if you've arrived at your conclusion and don't want anything to disturb it.

And your conclusion? I might paraphrase it along the lines of "You can't prove he juiced, and even if he did, so what?" Hard to argue with that sort of a moving target.

And if this is a distortion of what you believe, then please explain in specific ways how it's a distortion.

Perhaps you do think that he juiced, and perhaps it does matter to you if he did. Perhaps you're merely waiting for better evidence to emerge. In that case, then, how would you suggest going about getting such evidence, other than waiting for Bonds to confess and admit that he couldn't have done it without the juice?

I just noticed that Yahoo has linked to a story about this article on its front page. Here is a quote from it:

He calculated that a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent, or 4 to 5 mph for a pitcher who throws a 90-mph fastball.

That could translate into one fewer earned run every other game.

"That is enough to have a meaningful effect on the success of a pitcher, but it is not nearly as dramatic as the effects on home run production," Tobin said.

Where on earth does he come up with this equation??? 10 % increase in "muscle mass" = 4 mph faster = one fewer earned run every other game. Every single step of that is pure garbage.

Not to mention how terribly myopic the conclusion is. Take, as an example, Dontrelle Willis. If you increase his muscle mass by 10% and follow the equation, you've changed him from a 3.79 ERA pitcher, ranked 23rd among active pitchers, into a 3.08 ERA pitcher, ranking 2nd. If you increase Roy Oswalt's muscle mass by 10%, he becomes the greatest starting pitcher in the history of baseball. One wonders why he doesn't spend more time in the weight room! This simple cluelessness about baseball suggests that the rest of the study will be rife with errors.

Where on earth does he come up with this equation??? 10 % increase in "muscle mass" = 4 mph faster = one fewer earned run every other game. Every single step of that is pure garbage.

That's not fair. How can you draw any conclusions at all from an excerpt? You're reading a one-page blurb that summarizes the conclusions of a scientific study in layman's terms. You could be right, but how can you possibly conclude that this is "pure garbage" when as you yourself acknowledge, you have no idea how he came up with the equation?

And how are you drawing your conclusions for Dontrelle Willis and Roy Oswalt? The excerpt you quoted doesn't give you any ability to do so.

And how are you drawing your conclusions for Dontrelle Willis and Roy Oswalt? The excerpt you quoted doesn't give you any ability to do so.

He calculated that a 10 percent increase in muscle mass should increase the speed of a thrown ball by about 5 percent, or 4 to 5 mph for a pitcher who throws a 90-mph fastball.

That could translate into one fewer earned run every other game.

Oswalt has started 208 games. Remove 104 ER (one every other game) from his record and his ERA and ERA+ go to 2.42 and 181 respectively.

I find it somewhat droll that amidst all the charges of creationism, I'm the one who seems interested in discovering new empirical evidence, and the anti-creationists seem to want to dismiss it before it's even gathered. I'm certainly not saying that whatever findings we might come up with would constitute 100% conclusive proof, but to say in advance "I don't see how this gets us anywhere" doesn't strike me as particularly "scientific." It sounds to me more as if you've arrived at your conclusion and don't want anything to disturb it.

I find it somewhat droller that the more you go down this path, the more you sound like a creationist. This comment, for example, is exactly the sort of response you get from a Discovery Institute type ("these nasty closeminded paleontologists won't investigate my 'theories'") when it's not-so-gently pointed out that their "evidence" of "signs of intelligence" can't actually be tested scientifically.

Oswalt has started 208 games. Remove 104 ER (one every other game) from his record and his ERA and ERA+ go to 2.42 and 181 respectively.

That could be a sloppy quote from Yahoo. From the article linked here, which I quote in #3 above, "That translates to a reduction in earned run average of about 0.5 runs per game." Lower Oswalt's ERA by 0.5 and his ERA+ is 169. Sure that's high, but is it really inconceivable for an elite pitcher who hasn't sniffed his decline phase yet? And besides, who's to say that Oswalt's ERA isn't already lower by 0.5 or that Tobin's equation would suggest that Oswalt's ERA would go down that much.

This just reinforces my point - there's absolutely no way to draw any conclusions about his conclusions without knowing exactly what they are and how he arrived at them.

The claim that a pitcher's ERA would drop by 0.5 run due to 10% incr in muscle mass is utter non-sense. Yes, I can dump this entire study due to that absurd claim.
still absurd at 0.5 run.

Since when is velocity so easy to isolate over location and change of speeds??

bye bye.

The claim that a pitcher's ERA would drop by 1 run due to 10% incr in muscle mass is utter non-sense. Yes, I can dump this entire study due to that absurd claim.

Except that the claim is that a pitcher's ERA would drop by half that.

Sure that's high, but is it really inconceivable for an elite pitcher who hasn't sniffed his decline phase yet? And besides, who's to say that Oswalt's ERA isn't already lower by 0.5

I agree with the latter point, but as to the former, very, very few pitchers have had a career ERA+ as high as 169 at any time in their career. Aside from Pedro and Walter Johnson, I don't there are any. OK, Fernando Valenzuela after his first 8 starts, but certainly none with over 1400 IP.

I agree with the latter point, but as to the former, very, very few pitchers have had a career ERA+ as high as 169 at any time in their career. Aside from Pedro and Walter Johnson, I don't there are any. OK, Fernando Valenzuela after his first 8 starts, but certainly none with over 1400 IP.

Yeah, I have to admit that I was surprised to note that a 169 ERA+ would be the alltime leader by 9 points. I'm skeptical that Roy Oswalt could become that good. But that quickie blurb doesn't say that Tobin really thinks he could either.

I have just read the study

Where did you get it?

And, further, note that the claim that it COULD translate into 1 fewer earned run every other game is not a quote from the good Dr. Let's call off the wolves until his actual report is published.

some of us think more ... scientists (with high standards of "proof")

I yam what I yam.

Where did you get it?

I asked the author for a copy.

You could be right, but how can you possibly conclude that this is "pure garbage" when as you yourself acknowledge, you have no idea how he came up with the equation?

Because I cannot imagine that he did the research to support those claims.

If someone published a study that asserted, convincingly, that a 10% increase in muscle mass would add 4-5 mph to a pitcher's fastball, that would be a landmark study far more important than the conclusions presented in this one.

If someone published a study that asserted, convincingly, that a 4-5 mph increase in velocity would drop your ERA by 0.50 or more, that would be a landmark study far more important than the conclusions presented in this one.

Either one of those would have to have years of experiments with control groups of professional baseball players, massive amounts of aid from physiologists, access to statistics of unprecendented depth (pitch speed velocities for every game going back a generation), etc.

And yet, those two equations are just baby steps, almost tangental ones at that, in a study that presents the not exactly earth-shattering thesis that steroids might help hitters hit a lot more homeruns. I just don't buy it. It seems totally impossible to me.

I find it somewhat droll that amidst all the charges of creationism, I'm the one who seems interested in discovering new empirical evidence, and the anti-creationists seem to want to dismiss it before it's even gathered. I'm certainly not saying that whatever findings we might come up with would constitute 100% conclusive proof, but to say in advance "I don't see how this gets us anywhere" doesn't strike me as particularly "scientific." It sounds to me more as if you've arrived at your conclusion and don't want anything to disturb it.

I find it somewhat droller that the more you go down this path, the more you sound like a creationist. This comment, for example, is exactly the sort of response you get from a Discovery Institute type ("these nasty closeminded paleontologists won't investigate my 'theories'") when it's not-so-gently pointed out that their "evidence" of "signs of intelligence" can't actually be tested scientifically.

Well, given that I've been referred to as a "steroid McCarthyite" who supports congressional "Stalinist show trials" (i.e. the steroid hearings), I guess "creationist" is a relatively mild mannered putdown. Chip is clearly the kinder and gentler version of Primate we're now getting in these steroid wars.

OK, Chip, I won't insist that any findings of these distance measurements be taught in the public schools of Kansas. Your children will be safe.

But if ever such a study is made, it still might be interesting to see the results, even if Chip and his friends will greet them with cotton in their ears and duct tape over their eyes, screaming at the top of their lungs that they don't mean anything. And the rather hysterical reaction to even a suggestion of such a study might possibly suggest that certain people might not like what the results of that study might be.

Of course, the one factor that he fails to mention is that this period coincides with the time that many players decided to build up muscle mass. It doesn't follow from this that they all used steroids.

The underlying assumption that all that was new was steroid use, rather than the general acceptance of the idea that building up muscles would make you a better hitter, is just not true.

But if ever such a study is made, it still might be interesting to see the results, even if Chip and his friends will greet them with cotton in their ears and duct tape over their eyes, screaming at the top of their lungs that they don't mean anything. And the rather hysterical reaction to even a suggestion of such a study might possibly suggest that certain people might not like what the results of that study might be.

Speaking of hysterical reaction.

Thanks for continuing to reinforce my analogy.

But if ever such a study is made, it still might be interesting to see the results, even if Chip and his friends will greet them with cotton in their ears and duct tape over their eyes, screaming at the top of their lungs that they don't mean anything. And the rather hysterical reaction to even a suggestion of such a study might possibly suggest that certain people might not like what the results of that study might be.

If you did such a study, you might come up with evidence that Bonds was stronger post 1998. Is that in dispute?

If restrained amusement equals hysteria, you certainly can send the butterfly nets after me.

If you did such a study, you might come up with evidence that Bonds was stronger post 1998. Is that in dispute?

Probably not, but you never know what such a study might discover, which is why I think it would be an interesting one to undertake. And BTW I am not saying that the results would "prove" anything, only that they might suggest one interpretation over another. Steroids might well be a strong factor in any demonstrated gain of strength, but I'm not saying that they would be the only factor.