Except expansion has the same effect, as lesser players enter the league to fill rosters.

I am so sick of people questioning whether steroids actually enhance baseball performance.


Let me ask you this:

-Is there any doubt steroids help when you're weight training?

-Is there any doubt that weight training leads to better results on the field?


The answer to both is a resounding "No."

Ah, but what about HGH? May I remind you that absence of evidence is evidence of absence.

RTFA, the research already accounts for the effects of expansion.

For those who haven't popped in HOM threads (where Dan has explained this stuff thoroughly), he projects the standard deviation of each league-season based on three independent variables: expansion, integration and average run scoring. The regression is a decent fit overal, but in 93-04, the actual standard deviations are consistently higher than the projected ones, so there might be a variable unaccounted for, and that variable might be the roids.

Let me ask you this:

-Is there any doubt steroids help when you're weight training?

-Is there any doubt that weight training leads to better results on the field?

I think your phrasing is a bit misleading. When you ask about "doubt," are you referring to scientific evidence or common sense? Common sense isn't the same thing as a proven fact, as a lot of what we believe to be true despite the lack of evidence oftentimes ends up becoming egg on our faces when the research is done. We humans, after all, are a fairly naive yet corrigible bunch.

Secondly, how strong is the correlation between better weight training and better results? I haven't seen any research on this, but I'd think that the correlation is slim, if there is any at all. For every Barry Bonds, you have an Alex Sanchez; for every Roger Clemens, you have a Ryan Franklin.

The regression is a decent fit overal, but in 93-04, the actual standard deviations are consistently higher than the projected ones, so there might be a variable unaccounted for, and that variable might be the roids

Or the ball.

Secondly, how strong is the correlation between better weight training and better results?

One of the biggest changes in MLB over the last 20 years is the increased use of weight training. As recently as the mid-1980s, conventional wisdom among old baseball men was that weight training hurt your ability to play baseball by reducing flexibility. Sparky Anderson made a comment when Lance Parrish showed up to Tigers' training camp after an offseason of weight training something like "He needs to decide if he wants to be a baseball players or a weight lifter".

In terms of correlation, weight training was relatively uncommon back in the old days (offseason training in general was less common pre-free agency because players worked offseason jobs). The few guys who I can think of that were big into weight training would suggest that it's an obvious benefit - Honus Wagner, Carl Yastrzemski, Brian Downing - but, of course, this is pretty clearly a case of selection bias (nobody cared if "Losing Pitcher" Mulcahy or Mario Mendoza lifted weights in the offseason). For the modern major leagues, on the other hand, I suspect that there aren't enough major-leaguers who DON'T weight train for any kind of correlation analysis to be meaningful.

Or the ball.

Why would the ball affect the standard deviation of player performance?

I don't know it is usually one of the answers when explaining post 1993 era baseball.

I don't know it is usually one of the answers when explaining post 1993 era baseball.

To explain higher run scoring, which makes perfect sense. What Dan R. is saying is that the spread in MLB talent increased in this time period, even when one controls for expansion and run-scoring (both of which would increase the spread of talent). He says it's a persistent thing for about 12 years and he notes that these 12 years coincide with what we think of as "the steroid era".

The article mentions 1993 - 2004, but what do Dan's numbers look like for 2005 - 2008? If standard deviations went down with the introduction of testing, I would think that would strengthen the case that this effect is "steroids". I'm also a little skeptical that the first year that he observes this, 1993, is both an expansion year and a year with a dramatic increase in run scoring. It seems likely, to me, that higher standard deviations at this time could simply be an indication that his model is inadequately accounting for these two effects.

But overall, I think that this is probably the best macro-level evidence out there that steroids did have a significant effect on major-league baseball, and I tend to agree with alskor in #2 that the notion that steroids will improve baseball performance has always struck me as something of a no-brainer.

So assuming I have the current version of Dan's spreadsheet, here are the RMSE for Dan's estimator of leaguewide standard deviations:

NL 1893-2005: .170

NL 1961-2005: .154

NL 1993-2004: .175

AL 1893-2005: .220

AL 1961-2005: .181

AL 1993-2004: .134

Just looking at those, it does not look like Dan's estimator is having a noticeable problem with 1993-2004. Of course, standard deviations were up, but it looks like most of the net increase was explicable in terms of the factors Dan used for his regression analysis. This is, by the way, consistent with an analysis Nate Silver did on BP a few years ago.

Assuming that pitchers and hitters used performance-enhancing drugs equally, there is no reason to expect that overall run-scoring levels would change.

OK, first of all, this is a giant assumption that they used equally. I mean, do we have any reason to assume this? Even a common sense reason? Most people actually assume that only hitters really use, based on articles by various sportswriters. I'm sure that's wrong, but it points out that there's no common sense reason to assume that both used equally and no hard data, either. There's just no way to know either way.

Secondly, why would this assumption lead to no change to run-scoring levels? Is that another common sense argument? Because I don't see one. To be fair, I don't see an argument that they should change one way or another, either.

That brings me to another question, which actually goes along with alksor's post #2. I don't see anyone arguing that steroids don't help you get stronger (with proper training, workouts, etc.) I don't see anyone arguing that being stronger doesn't make you a better hitter (yes, it won't give you the ability to hit the ball, but if you already have that, stronger is better, even for Alex Sanchez). It's common sense. Yes, common sense can be wrong, but that's not my point here.

Here's my question. What are steroids supposed to do for pitchers? Or to put it another way, what is weight training supposed to do for pitchers? Is it just the "extra 3 MPH on the fastball"? Or is working out supposed to give you that extra bite on a curve ball or extra dip on the slider? Because I can see that if it doesn't make breaking pitches or off speed pitches any better, but the fastball's better, then scoring could take on a different shape. You're not just making pitchers better, you're improving certain areas.

Or do pitchers take steroids to recover faster, not from workouts, but from the rigors of pitching? In that case you'd probably see pitchers throwing more complete games, not less.

Or to put it another way, what is weight training supposed to do for pitchers? Is it just the "extra 3 MPH on the fastball"?

To quote Bill Mazer from a couple of years ago..."How come pitchers today don't throw as hard as pitchers did in the 1950's?

The juicers’ production should, on the whole, exceed the league average, while the clean players should lag.

This is exactly what happened between 1993 and 2004.

Boy is that poorly phrased. The SD may have increased (I assume that is correct) but that tells you nothing about the effect of steroids.

-Is there any doubt steroids help when you're weight training?

-Is there any doubt that weight training leads to better results on the field?

Let us assume that, in fact, weight training leads to better results on the field. That's certainly a reasonable assumption and is presumably true for the "right" form of training whatever it might be.

Let's also assume that steroids help you when you're weight training.

But what is pretty much indisputable is that players have done a lot more weight training over the last 15 years -- but I see no reason to see why that's not true of both steroid users and non-users. So your two questions are not really on point.

So how much do steroids improve weight training benefits? And how much effect does that benefit have on the on-field results? To my knowledge, nobody knows the answers to those although you could perhaps broadly generalize from some studies to take a WAG at #1.

There are probably diminishing returns to weight training. How much does lifting 7 days instead of 4-5 per week help? Does it help at all in the long run or is it just a matter of being able to bulk up more quickly?

Frank Thomas is as big and strong as a house and generally considered to be clean and contributed as much to the increased SD as probably anybody this side of Bonds. Pujols? Griffey? Maddux? Pedro? Chipper? Berkman? Larry Walker? Edgar? There are a lot of outliers, a lot of guys putting up almost unheard-of numbers (those last 4 are in the rare group of OBP>400, SLG>500, something almost unheard of for nearly 60 years). There are simply too many factors impacting on SD to ascribe it all (or even mostly) to steroids.

Also, if user/non-user is a big part of the increase in the SD, what you'd really expect to find is a bi-modal distribution (controlling for other factors which wouldn't be easy) -- i.e. it's a sample of 2 (or more) populations, not one population. Latent mixture models could potentially be used to find this. Note, you still couldn't ascribe it to steroids but it would at least be more consistent with that than a uni-modal distribution with wider variance.

Kiko Sakata, I have only calculated stdevs through 2005. (The '05 numbers are substantially lower than the '04 ones, but it's only one year, and Bonds's absence makes a significant difference). I'll see if I can get around to doing the last three years sometime soon.

Blackadder, that estimation includes all sorts of factors, including the league HR rate, which in turn are likely to be influenced by steroids. I'll see if I can run a regression with just run scoring and expansion and see what the errors look like then.

Greg Pope, what % of players suspended for PED's have been pitchers? Yes, my understanding is that pitchers juice primarily to improve their recovery time...

Walt Davis, I say explicitly in the final sentence that there is no proof that steroids are the cause of the increased SD's; I merely note that it is a pattern consistent with what we would theoretically expect would be a consequence of steroid use.

Also, if user/non-user is a big part of the increase in the SD, what you'd really expect to find is a bi-modal distribution (controlling for other factors which wouldn't be easy) -- i.e. it's a sample of 2 (or more) populations, not one population

Only if players either don't use, or use consistently. But I can well imagine there is a continuum between those who never use, those who have used rarely, those who have used frequently but not consistently, and those who have used regularly.

Forget about whether AAS help with weight lifting. The question is whether they help with baseball. Of course they do. I'm sure most of you can remember back to when you were in high school and college when your testosterone levels were shooting through the roof. You were young, powerful and alive, confident, carefree and cocky. You could go all night and all day the next day without a care. This is what introducing exogenous testosterone does for someone whose levels have already peaked and begun to decline. This reintroduction of youthful abilities and mental attitude/sense of well being is the HUGE benefit to an athlete playing baseball and traveling for 6 or 7 months out of the year and training for the other 5 or 6 months. Increased energy, increased ability to recover from intense physical work, increased sense of well being. Yes, AAS help with baseball and other sports as well, absolutely. No, they won't turn Alex Sanchez into Barry Bonds, or Ryan Franklin into Roger Clemens, but they will help an athlete perform better, for longer periods of time and into a later age range.

Only if players either don't use, or use consistently. But I can well imagine there is a continuum between those who never use, those who have used rarely, those who have used frequently but not consistently, and those who have used regularly.

Also only if players use in consistent doses, and only if players train in consistent manners/levels. As Walt notes, it'd be damn hard to properly control for other factors, like those we've mentioned, as well as a ton of others.

Fortunately, Walt Davis is involved in this thread, so he can math-check this: Is it not possible that the increase in Standard Deviation (SD) is primarily the result of specialization? What happened in 1993 onwards is not just that runs scored went up. It is specifically that home runs went up, taking runs scored with them. This causes one group of hitters - power guys - to gain much more value than other types of hitter. Ever since Babe Ruth, power has had a disproportionate effect on runs scored, but this makes matters just much worse. Having your most potent hitters gain the most from an effect is exactly the type of thing, as I best remember from my ancient math degree, that causes SDs to rise. You don't need steroids. As Dan says, his SD evidence does not prove anything about steroids. It does say things about performance distribution, but not about the reasons for any changes in the distribution. So, Walt, do I have the gist of SD math correct here?

Also, there is evidence that steroids do NOT enhance baseball performance. I don't mean that there is little evidence that they do; I mean that there is evidence that they don't. It's not conclusive or anything, but it's there. One place to find it (out of many but this is a post, not a book) is in the wonderful world of professional bodybuilding. These guys take more steroids than anyone, and do more weight training, and have been doing that since at least the late 1960s, when I first started following the activity to learn how to do weight training so I could swordfight better. Now, the absolute top end of bodybuilders - the regular Mr. Olympia winners - make plenty of money. But when you get just one class below those guys - to the guys who regularly finish in the quarterfinals - well, those guys still mostly have day jobs to supplement their bodybuilding incomes. This was certainly true in the 1960s through 1990s, if not still today. Now compare that to the Major League minimum salary. If heavy steroids and weight training could turn a normally athletic bodybuilder into a ML baseball player, where are the converted bodybuilders? Where are the guys who take their steroid-fueled abilities and turn them into million-dollar contracts instead of jobs as bar bouncers? There aren't any. If steroids do what their opponents claim, there would be.

There is more of this kind of stuff. If Barry Bonds' homer explosion in 2001 was due to steroids, why doesn't Luis Gonzalez, who had the same explosion but who is thin as a rail, look like a bulked up steroid user? And why did the phenomenon only last one year for Luis and Barry? is it not a lot more reasonable to think that, since MLB specifically said in the offseason that they were going to ask the umpires to start calling high strikes in 2001, it turned out the Bonds and Gonzalez had sweet spots high and inside that they could exploit for homers, but that they had not known about because they always laid off that pitch because it had been a ball? ESPN taped all of Barry's homers into one tape at the end of that year. Take a look at it if you can find a clip somewhere. Homer after homer hit off of high inside pitches that were strikes in 2001 but not before. Into the previously-not-famous McCovey Cove. I've looked at that tape a lot. There is a lot of evidence for Bonds having a sweet spot high and tight, but none for steroids. And, since the league's pitchers would figure that out and adjust in 2002 to not throwing him that strike, the one-year aspect makes sense that it does not make if you assume steroids.

Sometimes, asking the right question is a lot more important than doing complicated math. "Did they do steroids?" is a very limited question that leads to very limited analysis. "Would steroids have helped if they had done them?" is a much better one leading to many more good questions to ask and answers to get.

- Brock Hanke

Why would the ball affect the standard deviation of player performance?

Because players have unequal ability to take advantage of changed conditions.

One way to test that hypothesis would be to look at specifically the NL in 1931. We know they made a major change to the ball. Though it was to make it substantially less lively. At which point you've got two factors in play potentially working in opposite direcitons. A radical change in overall playing conditions could potentially reward the early adapters while the lowering of offensive levels should serve to lower the SD.

Interesting, and timely, reading: http://www.t-nation.com/free_online_article/sports_body_training_performance_steroids_drugs/the_steroid_interviews

Also, there is evidence that steroids do NOT enhance baseball performance. I don't mean that there is little evidence that they do; I mean that there is evidence that they don't. It's not conclusive or anything, but it's there. One place to find it (out of many but this is a post, not a book) is in the wonderful world of professional bodybuilding. These guys take more steroids than anyone, and do more weight training, and have been doing that since at least the late 1960s, when I first started following the activity to learn how to do weight training so I could swordfight better. Now, the absolute top end of bodybuilders - the regular Mr. Olympia winners - make plenty of money. But when you get just one class below those guys - to the guys who regularly finish in the quarterfinals - well, those guys still mostly have day jobs to supplement their bodybuilding incomes. This was certainly true in the 1960s through 1990s, if not still today. Now compare that to the Major League minimum salary. If heavy steroids and weight training could turn a normally athletic bodybuilder into a ML baseball player, where are the converted bodybuilders? Where are the guys who take their steroid-fueled abilities and turn them into million-dollar contracts instead of jobs as bar bouncers? There aren't any. If steroids do what their opponents claim, there would be.

I hope this is a joke. You don't honestly believe an abscence of former pro bodybuilders playing professional baseball = evidence AAS don't help athletes like baseball players, do you?

First, although it's hard to know where to begin, bodybuilders are not athletes; in fact, they don't do anything which could be considered athletic. They take enormous amounts of AAS and other drugs, consume huge numbers of calories and perform a tremendous amount of volume single joint, isolation movements which are about the furthest type of exercise you can get from athletic.

Second, a baseball player is a speed strength athlete; he does not need bulk to do his job. Taking huge amounts of AAS and bulking up is not the goal of a baseball player, but that is not the only effect AAS may have on a user. Introducing exogenous testosterone, in reasonable yet still efficacious amounts, will enhance recuperative ability, increase protein synthesis, improve metabolic function and enhance feelings of well being. If a baseball player, an athlete, already has the skills to succeed in his sport, the ability to train harder, longer and recover quicker is a huge boost. Why do players decline in their mid to late 30's? Because their bodies are aging, testosterone production is flagging and they can't recover from the 162 game grind like they did in their 20's and early 30's. Why did Barry Bonds set an otherworldy level of performance in his late 30's and early 40's (don't look at one season of extraordinarily high HR's, look at his HR rates, per PA or per AB, and you'll see they dropped dramatically in his late seasons)? Because he introduced artificial testosterone into his system turning back the clock.

AAS won't give you the skills to be a baseball player. If you're a baseball player already, they absolutely will allow you to improve and maintain your skills.

If Barry Bonds' homer explosion in 2001 was due to steroids, why doesn't Luis Gonzalez, who had the same explosion but who is thin as a rail, look like a bulked up steroid user?

Barry Bonds = Luis Gonzalez? Really? Seriously? Oh, and the vascularity Gonzalez displays in his forearms is a classic sign of AAS use, but that is neither here nor there. He may have used, he may not have. But to try to make him analogous to Bonds is absurd.

The SD is correlated with the mean; therefore, as mean HR rate increases, so will the SD. The coefficient of variation (SD/mean) is an adjustment that helps correct for this bias.

Brock Hanke, power has a disproportionate effect on run scoring? Disproportionate compared to what? It's not as important as OBP...and Bonds' power spike lasted for three years after 2001; it's just that he was walked so much that the HR totals went down. He was still cranking out a HR every 8 AB in those years.

JCB, my stdevs are calculated using wins above average; the mean is always 0.

This is an interesting corrollary to the debate about Koufax and Martinez and whether it's easier to put up a higher ERA+ or OPS+ in today's game. Maybe it is easier because there's a wider gap in the talent from top to bottom, and those at the top can pad their numbers by feasting on those at the bottom. I have no idea if this is true...just spitballing.

If you're a baseball player already, they absolutely will allow you to improve and maintain your skills.

And if that is so, they are GOOD.

Dan, have you changed your league standard deviation estimator since creating your WARP? The formula you list on the WARP thread is:

.00366*Year + .1254*Runs per game - .028777 * NL Teams - .00567 * MLB Teams - .00256 * Season Length - .932 * Win% of worst team in league - .0278 * Years since expansion (Max 12) + .15 * World War (1 or 0) + .00158 * Estimate of player population (14.8M in 1893, 60.1M in 2005) - .2466* Integration (1 or 0) + 2.789

of which the only factor that seems potentially related to steroid use is runs per game. If so, would you mind posting your new formula?

If heavy steroids and weight training could turn a normally athletic bodybuilder into a ML baseball player, where are the converted bodybuilders? Where are the guys who take their steroid-fueled abilities and turn them into million-dollar contracts instead of jobs as bar bouncers? There aren't any. If steroids do what their opponents claim, there would be.

As alilisd points out: AAS won't give you the skills to be a baseball player. If you're a baseball player already, they absolutely will allow you to improve and maintain your skills. No "opponents" of steroids claim what you seem to think that they claim. None (at least not on BTF). I have yet to see anyone here say that steroids turned an average joe into a baseball superstar. What is claimed is this:

Barry Bonds was a superstar player, one of the best in the game. Then he took steroids and broke records.

Mark McGwire was a very good home run hitter, but somewhat one-dimensional. Then he took steroids and broke records.

Alex Sanchez was a mediocre baseball player, but probably a player who would max out at Triple-A. Then he took steroids and was able to hold a job in MLB for a few years.

That may be oversimplifying a bit, but if you're going to argue with someone, at least understand their points, even if you disagree.

I have yet to see anyone here say that steroids turned an average joe into a baseball superstar

There is, however, Mr. Applegate ...

Goodbye, old girl
My old girl
When you awaken I'll be gone
Can't tell you where I go
It isn't fair, I know
But trust in me and carry on

No "opponents" of steroids claim what you seem to think that they claim. None (at least not on BTF). I have yet to see anyone here say that steroids turned an average joe into a baseball superstar. What is claimed is this:

Fortunately for you, then, you've never read the late unlamented Kevin's screeds. He claimed that steroids were the difference between Jose and Ozzie Canseco, and further claimed that steroids were the difference between Jason Giambi's 2004 OPS (721) and his 2003/2005 OPS (939/975).

(In fact, he went further and claimed that they not only had this effect, but had it instantaneously. In 2005, Giambi had a ~.720 OPS for the first two months of the season, and there was talk of demoting him; Kevin claims that this talk led him to start taking steroids, and allowed him to immediately pull up his OPS to the point where he finished at .975.)

Sometimes, asking the right question is a lot more important than doing complicated math. "Did they do steroids?" is a very limited question that leads to very limited analysis. "Would steroids have helped if they had done them?" is a much better one leading to many more good questions to ask and answers to get.

I could not agree more with this. So in that spirit, instead of asking "How would lifting a car help make me a better ballplayer?" why not ask something relevant like "How helpful would it be if I could get back in the batting cage for another three hours after lunch?" Or maybe "Do you suppose increased leg strength and stamina could help me carry my best mechanics further into games? Do you think that would help my pitching?"

Let's start from there and see what happens.

Whoever alilisd's is, I agree. I think I'll just cut and paste those posts in every steroid thread I'm tempted to enter.

Alex Sanchez was a mediocre baseball player, but probably a player who would max out at Triple-A. Then he took steroids and was able to hold a job in MLB for a few years.

How about:

Alex Sanchez was a mediocre baseball player, probably a player who would be able to hold a job in MLB for a few years. Then he took steroids and was able to hold a job in MLB for a few years.

A player good enough to "max out at AAA" is certainly good enough to be a crappy MLBer for a few years.

Dan R: are your standard deviations posted somewhere?

I agree this is a very interesting way to look at the issue. But there are a couple of complicating factors that might have also increased SDs in this period:

1) Denver. Both pitcher and hitter stats in Denver are so extreme that I would think they could have a measurable impact on SDs. It would be interesting to calculate SDs without CO players (assuming you haven't already).

2) Changes in pitcher usage: in the early 90s there was a dramatic increase in the use of short relievers. And these relievers have a big advantage over starters (about 1 R/G). So if your pitcher SDs include relievers, that shift in usage would produce larger SDs.

Blackadder, I am using a new equation--I'll send it to you when I get home.

Guy M--Yes, they're posted in the StDevs and Rep Levels.xls file in the Rosenheck WARP.zip archive in the Hall of Merit Yahoo group.

Park effects definitely don't have a role, since the stdevs are calculated using wins above average which are already park-adjusted.

There's no doubt that lower seasonal IP totals for starters are correlated to higher stdevs of effectiveness. That's one of the main reasons I tend to focus on offensive stdevs to evaluate the ease of domination of a league.

Fortunately for you, then, you've never read the late unlamented Kevin's screeds. He claimed that steroids were the difference between Jose and Ozzie Canseco, and further claimed that steroids were the difference between Jason Giambi's 2004 OPS (721) and his 2003/2005 OPS (939/975).

(In fact, he went further and claimed that they not only had this effect, but had it instantaneously. In 2005, Giambi had a ~.720 OPS for the first two months of the season, and there was talk of demoting him; Kevin claims that this talk led him to start taking steroids, and allowed him to immediately pull up his OPS to the point where he finished at .975.)

Fair enough. I certainly have read my share of the threads and I don't recall kevin making those kind of statements. But I was never directly involved so I can't say, and kevin certainly came up with his share of out-there claims. Can I change my statement to "anyone sane"?

Alex Sanchez was a mediocre baseball player, probably a player who would be able to hold a job in MLB for a few years. Then he took steroids and was able to hold a job in MLB for a few years.

A player good enough to "max out at AAA" is certainly good enough to be a crappy MLBer for a few years.

Do you seriously not get my point? It's possible that both "Alex Sanchez wasn't very good and didn't hit home runs" and "Alex Sanchez was helped by steroids" are true statements. And most of the people on this site who are "steroid opponents" agree with both of those.

I agree with JCB: in principle, you should be able to calculate directly the effect of increased RPG on the SD of player ability rather than resorting to regression. Maybe if I have time later I'll run through a quick calculation and see whether it agrees with what Blackadder posted in #28.

Do you seriously not get my point? It's possible that both "Alex Sanchez wasn't very good and didn't hit home runs" and "Alex Sanchez was helped by steroids" are true statements. And most of the people on this site who are "steroid opponents" agree with both of those.

In essence you're arguing that Sanchez was perhaps marginally helped by PEDs. But that's not what the witch hunters (I'll continue to use the term) are screaming about; they're screaming that steroids turn bad players into average players (and average players into stars, etc.). They're asserting significant increases in performance, not marginal ones.

Also, there is evidence that steroids do NOT enhance baseball performance. I don't mean that there is little evidence that they do; I mean that there is evidence that they don't. It's not conclusive or anything, but it's there. One place to find it (out of many but this is a post, not a book) is in the wonderful world of professional bodybuilding. These guys take more steroids than anyone, and do more weight training, and have been doing that since at least the late 1960s, when I first started following the activity to learn how to do weight training so I could swordfight better. Now, the absolute top end of bodybuilders - the regular Mr. Olympia winners - make plenty of money. But when you get just one class below those guys - to the guys who regularly finish in the quarterfinals - well, those guys still mostly have day jobs to supplement their bodybuilding incomes. This was certainly true in the 1960s through 1990s, if not still today. Now compare that to the Major League minimum salary. If heavy steroids and weight training could turn a normally athletic bodybuilder into a ML baseball player, where are the converted bodybuilders? Where are the guys who take their steroid-fueled abilities and turn them into million-dollar contracts instead of jobs as bar bouncers? There aren't any. If steroids do what their opponents claim, there would be.


This is a lot words to lead to such a nonsensical conclusion. For a body builder to "convert" to a major league ball player, he/she would actually have to you know, be able to play baseball at the same level as AAA players.

If Michael Jordan couldn't "learn" to play baseball, I sort of doubt that Joe Bodybuilder would have much chance.

Ray DiPerna Posted: February 16, 2009 at 02:16 PM (#3078291)

A player good enough to "max out at AAA" is certainly good enough to be a crappy MLBer for a few years.

Sorry, by definition if you "maxed out at AAA", you "were not" a crappy MLBer for a few years. Although PED may help one achieve this goal.

Sorry, by definition if you "maxed out at AAA", you "were not" a crappy MLBer for a few years. Although PED may help one achieve this goal.

I didn't say you "were" a crappy MLBer; I said you were good enough to be a crappy MLBer.

In essence you're arguing that Sanchez was perhaps marginally helped by PEDs. But that's not what the witch hunters (I'll continue to use the term) are screaming about; they're screaming that steroids turn bad players into average players (and average players into stars, etc.). They're asserting significant increases in performance, not marginal ones.

How big a change in an underlying skill does there have to be to render a significant change in end value? If a player can shave a tenth of a second off of his time to first, how many extra infield hits is that likely to gain him over a year, or how many extra balls can he get to in the field? If he improves his reflexes by a tiny amount, how many extra steals can he pick up on the basepaths, how much better a jump can he get in the field, or how much of an advantage can he get from having that tiny little increment of extra time to pick up the spin on a pitch? If he's a little bit stronger, then how much does he likely gain from that ball carrying an extra couple feet, or from changing that fly ball into a soft liner, or being able to get around a bit more quickly on that fastball?

When we're talking about major leaguers (or even AAA players), we're already talking about those who have demonstrated skills far beyond the general population. For them, it is not unreasonable to believe that a very small change in these underlying skills could potentially result in a significant change in on-field performance.

Dan asks, "Brock Hanke, power has a disproportionate effect on run scoring? Disproportionate compared to what? It's not as important as OBP...and Bonds' power spike lasted for three years after 2001; it's just that he was walked so much that the HR totals went down. He was still cranking out a HR every 8 AB in those years"

A very fair question and a good pertinent fact. In an attempt to keep the post length down, I didn't explain much there. Here's what I mean: There is a proportion to Ty Cobb and Honus Wagner's game in the deadball era. There are more singles than doubles, more doubles than triples, and more triples than homers. That is, frequency is inversely related to the effect of the event. The proportion isn't exact or anything - the effect of all the league's singles is more than that of all the league's homers - but the most powerful events have the least probability. Babe Ruth's game is based on overturning that proportion. Homers were almost instantly more frequent than triples, and they are slowly but surely creeping up on doubles. The current effect of a league's total homers is certainly more than its doubles. That's what I meant by the term.

In this analysis, here's how it applies: Between 1992 and 1996 (I chose those years to span the big jump of 93-95), the National League's homers per team go from 105 to 159, an increase of 51%. Runs went from 628 to 759, which is an increase of 21%. So, homers were driving the increase of runs scored. In fact, they were the primary driving force of runs. Now, let's look at a Honus Wagner year, let's say 1908. The average NL team hit 19 homers. If you increased that by 51%, you'd get 29 homers, or about ten more per team. That obviously is not going to have anything like a 21% impact on total runs scored. What I was thinking is that, because an increase in the one specialized stat of homers has a disproportionately large effect on runs scored compared to a similiar increase in any other base stat (OBP, which is certainly more important than homers, is a compound stat of a lot of different base events), and because the spread of homers hit by players is very large, it seems likely that the increase in SD of runs is due primarily to the increase in homers, leading to an increase in SD of homers. Does that make sense? I do, in general, trust your SD work. It's very good, as far as I can tell. But I'm not an expert in SD theory (I was much more an applied mathematician in college), so I wasn't completely sure of myself, which is why the appeal to Walt Davis, which JCD was kind enough to answer. (Just for future use, do you know the theory of SD or are you like me, much better at applying it than theorizing? I only appealed to Walt because I understand that he is a math professor and knows the theory. If you know the theory, too, I'll just ask you directly next time. All of which is to say that I wasn't trying to question your work or insult you. I was just going with the one guy who I knew would know this stuff.)

I am aware the Bonds' homer rates per AB don't go down after 2001. I attribute this to the pitchers of the league finding out that the pre-2001 book on Bonds (throw it high and tight and jam him) was no longer valid, because that pitch had become a strike, so Barry made an adjustment to keep it from jamming him. If you check out some video of 2001 games, even late in the year, the broadcasters are still obsessing over the idea that you should jam Barry, while Barry takes high tight pitch after high tight pitch and deposits them into McCovey Cove. Apparently the league's pitchers decided that, with that pitch no longer effective, they had no good way to get him out, so they started throwing the thing out of the strike zone and just living with the inevitable walks. Well, there are also the intentional ones, which are the same idea - we don't now how to get him out. I don't attribute any of this to steroids, because the new strike zone seems completely sufficient to explain the whole phenomenon.

- Brock

Assuming that pitchers and hitters used performance-enhancing drugs equally, there is no reason to expect that overall run-scoring levels would change.

Only if we also assume that pitchers and hitters benefit equally from the use of performance-enhancing drugs.

Also, if PED use was disproportionately high among less skilled players, then mightn't we expect the SDs to get smaller rather than bigger? Maybe that's a bit of a reductio ad absurdum, but I think there's just too much assumed here for us to conclude much. Of course, since the conclusion is that we can't conclude much, that's probably OK.