Interesting Adjustment For Steriod Use

[Ben E Lou]

http://www.sportsmogul.com/content/steroids.htm

Baseball Mogul Rewrites The Record Books

March 24, 2005

Using the Baseball Mogul physics engine, we have adjusted the statistics of the last 20 years to those that we believe would have occurred without the use of steroids by major league hitters. In making these adjustments we have followed the general consensus of media and researchers who believe that steroid use became a measurable concern in the mid-1980s, and that the effects of steroids reached a peak between 1998 and 2004.




Adjusted players shown in bold (click on player name to see adjusted career stats)

Source: Baseball Mogul 2006 NOTES

We did not assume steroid use on the part of any individual player. Because we are not judging the guilt or innocence of individual players, we have instead adjusted all player statistics since 1985. Greater adjustments have been applied:

1. In seasons where a large number of home runs were hit, implying a larger steroid-related effect.
2. In seasons where a player's home run rate rose greatly from his previously established career average.

There is no general agreement on whether steroids increase injuries or promote healing (or both). Therefore, games played and plate appearances were unchanged.

If it becomes clear that certain players did use steroids while others did not, the home run numbers for steroid-users would need to be adjusted further downward, while those for the clean players would be reverted to their historical figures.

Steroid-Adjustment FAQ

Q: If Barry Bonds holds the real-life single-season home run record (with 73 in 2001), why does Mark McGwire hold the "steroid-free" single-season home run record?

A: Mark McGwire had a greater established level of "home runs per at bat" before his record-breaking 1998 season, including a 49-homer season in his rookie year. By contrast, Barry Bonds' 73-homer year was a greater deviation from his established performance level. The Baseball Mogul engine separates each player's ability into natural "power hitting" and that which may have been augmented by performance-enhancing drugs.

Again, we are not surmising the guilt or innocence of Mr. McGwire or Mr. Bonds. We are simply using the Baseball Mogul statistical and physics engines to "replay" the last 20 years without the estimated effects of steroids on baseball as a whole. However, within that context, it is reasonable to assume that the effects of steroids would be most salient in seasons with high slugging totals, and in seasons where a player's power-hitting deviated significantly from his career average to date.

Q: Were stats other than home runs affected by eliminating steroids from baseball?

A: Yes, all stats since 1985 were adjusted as part of this analysis. To see a player's entire career without steroids, click on a name in the "steroid-free" list.

[Franklinnoble]

Very interesting. Thanks for sharing.

[cartman]

So, based on this, it appears that juicing didn't really have all that great of an impact.

[Klinglerware]

Very questionable. As is suggested in other threads, steroids may have helped pitchers who used even more. The steroid effect would therefore be a wash, and other factors (ballpark, strike zone, Colorado, etc) would matter more to the home run rate.

Statistics can only be judged in the context of the conditions they were produced, so cross era comparisons of raw statistics are essentially meaningless. If Babe Ruth played his home games in the Astrodome and faced a better talent pool post-integration, I am sure his home run total would be different too...

[Ben E Lou]

Originally Posted by cartman

So, based on this, it appears that juicing didn't really have all that great of an impact.

Ummm...those are some pretty big drops...8-15% homer increase in a season. That's a significant impact on homers.

[Franklinnoble]

One thing I don't get is the implication that Griffey Jr. used the juice... is he really under suspicion?

[Ben E Lou]

Originally Posted by Franklinnoble

One thing I don't get is the implication that Griffey Jr. used the juice... is he really under suspicion?

I don't see any implication there:

We did not assume steroid use on the part of any individual player. Because we are not judging the guilt or innocence of individual players, we have instead adjusted all player statistics since 1985. Greater adjustments have been applied:

1. In seasons where a large number of home runs were hit, implying a larger steroid-related effect.
2. In seasons where a player's home run rate rose greatly from his previously established career average.

There is no general agreement on whether steroids increase injuries or promote healing (or both).

[bosshogg23]

Originally Posted by Franklinnoble

One thing I don't get is the implication that Griffey Jr. used the juice... is he really under suspicion?

He is the only home run hitter I havent heard associated with it. He weighs nothing and comes back from injuries slow as hell. I hope for his sake he isnt on it. He will look like Chris Rock with a walker if he got off the juice.

[Franklinnoble]

Originally Posted by Klinglerware

Very questionable. As is suggested in other threads, steroids may have helped pitchers who used even more. The steroid effect would therefore be a wash, and other factors (ballpark, strike zone, Colorado, etc) would matter more to the home run rate.

Statistics can only be judged in the context of the conditions they were produced, so cross era comparisons of raw statistics are essentially meaningless. If Babe Ruth played his home games in the Astrodome and faced a better talent pool post-integration, I am sure his home run total would be different too...

The cross-era effect on Ruth could also be argued as a wash.

Yes, integration might have made a difference, but expansion almost completely negates integration, and Ruth didn't play in an era where superstars trained year-round with high-tech equipment, diets, and personal trainers.

[cartman]

Originally Posted by SkyDog

Ummm...those are some pretty big drops...8-15% homer increase in a season. That's a significant impact on homers.

For a single season, that's really not a lot. That seems to fall in natural variation season to season. Take for example a 15% drop in a 40 home run season. That would net 34 homers. So you go from roughly 1 homer every 4 games to 1 homer every 4.7 games.

I would have thought the effect would have been more pronounced, especially for Bonds. His "non-juiced" career projection was only 40 or so less than his current total, so only the addition of around 3 more homers a year.

[Klinglerware]

Originally Posted by Franklinnoble

The cross-era effect on Ruth could also be argued as a wash.

Yes, integration might have made a difference, but expansion almost completely negates integration, and Ruth didn't play in an era where superstars trained year-round with high-tech equipment, diets, and personal trainers.

Yes, I would suspect Ruth would also perform extremely well post '93, the parks are much smaller today plus the other factors you mention. My point was that Ruth's statistical performance would differ depending on the era. I did mention the astrodome, Ruth's HR rate would probably go down if he played there and the other caverns we called baseball stadiums in the late 70s early 80s instead of the cozy Yankee stadium of the late 20s early 30s (at the time 295 ft to RF).

[Klinglerware]

Greater adjustments have been applied:

1. In seasons where a large number of home runs were hit, implying a larger steroid-related effect.
2. In seasons where a player's home run rate rose greatly from his previously established career average.

I do quite a bit of statistical modeling in my line of work and this seems a bit questionable methodolgically to apply seasonal adjustments in this manner. It is a bit of a leap to assume that steroid effects were larger in seasons where more home runs were hit. If the rate of home runs was high in 87 but went back to normal in 88, did players suddenly stop using steroids? It is better to assume that steroid use as being a constant or having an upward trend. It is also better to leave outliers and unexplained variation alone--massaging the effect is unnecessary if a general effect really does exist.

[Klinglerware]

dola--

Another weakness of this exercise is that the "steroid effect" was applied uniformly. The assumption is that every batter used steroids, which is probably not a reasonable assumption and is an unknowable unless a player is found out or admits to it. Another effect of this macro adjustment is that the resulting adjustment is really an average adjustment applied across the board. This may be okay in generating overall league effects, but it is poor in coming up with effects for individual players. If a player did use steroids, the adjustment understates the effect of steroids, if a player did not, the effect is overstated. They did use the "seasonal adjustment" to try to tweak the numbers--but as I mentioned previously, the rationale behind the adjustment is in effect cheating with the statistics--the effect is boosted artificially...

[PilotMan]

While I have a tendency to agree with the assumption that steroid use in baseball has tainted stats and had a noticible effect on the game itself, I find it total bullshit that somone feels that they can quantify what that effect actually is. There are just way to many other variables that come into play. The whole idea is speculation, and poor science and a slap in the face to the baseball fan.

Hey, if it happend we need to deal with it. We need to deal with the results and the outcomes whatever they are. All this romaticizing(sp?) is nonsense.

[dawgfan]

A fun little exercise I suppose, but highly questionable as to how accurate it is. We simply don't know how steroids have affected the game, and by how much. How do we know that steroids have helped hitting more than pitching? We don't. How do we quantify the effect steroids have had by themselves without factoring in all the other things that have changed in the game in the last 2 decades? We can't.

Basically, it's a conversation starter for Clay's website, and there's nothing inherently wrong with that, so long as we remember this is nothing more than guesswork at best.

[ISiddiqui]

Originally Posted by Klinglerware

Very questionable. As is suggested in other threads, steroids may have helped pitchers who used even more. The steroid effect would therefore be a wash, and other factors (ballpark, strike zone, Colorado, etc) would matter more to the home run rate.

Statistics can only be judged in the context of the conditions they were produced, so cross era comparisons of raw statistics are essentially meaningless. If Babe Ruth played his home games in the Astrodome and faced a better talent pool post-integration, I am sure his home run total would be different too...

Yep, I was just going to make the point about whether BM looked at doping by pitchers as well or made a knee jerk reaction, saying it only helped the hitters.

[Huckleberry]

The only real help pitchers would get from steroids would be recovery time.

Bulking up hurts pitchers, for the most part.

[ISiddiqui]

Originally Posted by Huckleberry

The only real help pitchers would get from steroids would be recovery time.

Bulking up hurts pitchers, for the most part.

Have you watched a baseball game in the last 5 years? I can't imagine that you wouldn't hear some announcer go on and on about the drive that a certain pitcher gets from their legs allowing them to throw harder...

More muscular legs -> more drive to the plate -> faster pitches

So yeah, it helps pitchers more than it hurts them.

[Crapshoot]

what an inanely pointless excercise- seems primarily an attempt to grab some attention.

[sterlingice]

Originally Posted by bosshogg23

He is the only home run hitter I havent heard associated with it. He weighs nothing and comes back from injuries slow as hell. I hope for his sake he isnt on it. He will look like Chris Rock with a walker if he got off the juice.

Yeah, no kidding. If Griff is on the juice, it's any wonder he even made the majors.

SI

[Ragone]

There is no general agreement on whether steroids increase injuries or promote healing (or both). Therefore, games played and plate appearances were unchanged.

Two words for you... David Boston


Honestly if you eliminate steroids from the simulation.. i think the players career's would lengthen by 1-2 years.. but the power #'s would dip a bit more dramatically.. maybe somewhere between 15-20%

[clintl]

Originally Posted by Ragone

There is no general agreement on whether steroids increase injuries or promote healing (or both). Therefore, games played and plate appearances were unchanged.

Two words for you... David Boston


Honestly if you eliminate steroids from the simulation.. i think the players career's would lengthen by 1-2 years.. but the power #'s would dip a bit more dramatically.. maybe somewhere between 15-20%

The real answer is that there were so many other different changes going on (two expansions, old large parks being replaced by smaller new parks, maple bats, different training methods, different hitting theories, different pitching theories, smaller strike zone, etc) that there is no way to tell how much steroids impacted the power numbers right now. Maybe in five years, when we have had a decent sample size of seasons under the testing program, we will. But right now, it's just a guess, and not even an educated guess.

[Arles]

A better study is to look at HR totals before strike zone changes, park changes, new expansion, ball changes that occurred after 1994-95. If you believe Canseco and the FBI investigations (two most credible arguments for steroids, combined with Balco), you would know that use started around 1990. Yet, from 1990 to 1995 there wasn't that big of a jump in HRs. Only once the other factors were added did we see that jump.

[QuikSand]

Originally Posted by dawgfan

A fun little exercise I suppose, but highly questionable as to how accurate it is. We simply don't know how steroids have affected the game, and by how much. How do we know that steroids have helped hitting more than pitching? We don't. How do we quantify the effect steroids have had by themselves without factoring in all the other things that have changed in the game in the last 2 decades? We can't.

Basically, it's a conversation starter for Clay's website, and there's nothing inherently wrong with that, so long as we remember this is nothing more than guesswork at best.

I completely agree.

In fact -- unless I'm missing it (I didn't spend a whole lot of time with this, I confess) he explains his rationale and his logic -- but underneath it all, someone has to make some assumption on exactly how much effect there is to all this... and as near as I can tell, this conjecture draws that right out of the ether. Unless there's some particular agrument why a certain year's stats are reduced by a specific amount... then why is this analysis beter than anyone else's analysis? Because of the Baseball Mogul engine?