De-steroidi-fying players?

[petrel]

Has anyone actually tried to remove the influence of steroids from players through a judicious use of the player editor? For example, if one wanted to "transform" Barry Bonds or Mark McGwire (or Roger Clemens) back to "normal" players, how would one do that with the player editor?

Any ideas? Or theories as to how to do that?

--Pet

[ohms_law]

It all depends on how you define what effect steroids actually have.

[RickD]

And how would you determine what would be "normal?"

[Imgran]

And do you then add the extra handful of years to the end of their careers that the health problems associated with steroids took away?

[JayC]

The BM guys themselves did this last year, adjusting statistics for all players instead of judging who may or may not have used steroids (and awarded the single-season home run mark to McGwire with 64), so you could use their data or adopt it however you'd like.

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

[ohms_law]

The "steroid removal" in Mogul isn't specific to any single player or group of players though (and, it's still in in Mogul 2008). The main aspect which creates a "steroid free" environment in Mogul is that the model (intentionally) doe snot account for things such as "power spikes" in a players career. In other words, just because a player hits 70 homers in one season that doesn't necessarily affect the power rating/predicted home runs that that player will receive in Mogul. On import, the game basically smooths players stats somewhat so that it doesn't consider any stat that is outside of the players career mean for that stat very heavily. Of course, if players who average 40-50 home runs over several seasons, that still heavily influences the players ratings so that there are still several really good players through the years.

[beerchaser]

And do you then add the extra handful of years to the end of their careers that the health problems associated with steroids took away?

Given that Bonds is still playing at age 40-whatever, and many other "suspected" roiders were around until age 40 or so, I'd say no.

[aj58078]

The BM guys themselves did this last year, adjusting statistics for all players instead of judging who may or may not have used steroids (and awarded the single-season home run mark to McGwire with 64), so you could use their data or adopt it however you'd like.

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

I really wish you wouldn't have said that. That takes my view of Mogul and totally flips it. I have used Mogul since '02 and now I am not sure if I will buy '09 or not...

[ohms_law]

Why is that?

[damonrusst]

Those BM adjustments were just an experiment,right? Their ratings aren't adjusted in the game itself, are they?

[ohms_law]

Well, the example on the webpage is just a game that Clay ran specifically for that purpose. He may have adjusted something to exaggerate the effects, but any statistical model will give those sorts of results... of course, you'll still get a player to pop in one statistical category or another occasionally. Mogul gives slightly more statistical deviations than could be accounted for with a "pure" mathematical model though, in my experience. If the concern is that players won't have unusual seasons, then you can lay that concern to rest. However, the primary point is that players will "regress to the mean" over time, which is a statistical term meaning that despite the fact that players will be able to have unusually good (or bad) seasons they will always end up returning to the average performance level which their ratings give them.

[CatKnight]

I may be misunderstanding ohms' explanation - but I think the steroids effect will still be there simply because the OVERALL numbers will still be so high. I ran a sim where Barry Bonds did indeed beat Aaron's record. Then lost it in 2009. To a guy who lost it in 2011 or 2012. While I agree that a player's career will be 'averaged' by his ratings, what those ratings mean change from era to era. A 1920s player with 99 power is NOT hitting 70 homers. I think BM accurately (depressingly) reflects the steroid era just fine.

petrel: What I might do as a quick and dirty fix is to make an assumption about the first year steroids have a serious effect on MLB. Clay's article changed all data from 1985 up. I suspect that might be a little early, but I'm fine with it.

Reduce all settings to 1985 levels. For example I think homers is something like .83 - use that instead of 1.00.

It won't completely erase the effect of steroids - but it should get the overall numbers back where they should be.

[ohms_law]

I may be misunderstanding ohms' explanation - but I think the steroids effect will still be there simply because the OVERALL numbers will still be so high. I ran a sim where Barry Bonds did indeed beat Aaron's record. Then lost it in 2009. To a guy who lost it in 2011 or 2012. While I agree that a player's career will be 'averaged' by his ratings, what those ratings mean change from era to era. A 1920s player with 99 power is NOT hitting 70 homers. I think BM accurately (depressingly) reflects the steroid era just fine.

Very true. Also, regarding the "steroids era" itself, there is quite a bit of debate going on right now as to the exact effects of steroids. Many people feel that steroids assist pitchers more than hitters, and there is a significant portion of the sabermetric community who beleave that there are probably more pitchers who are currently using steroids than there are pitchers. People see things such as Home Runs much more easily than they can see better pitching however. I tend to thing that the effects of steroids are vastly overstated, myself.

One other issue that people don't seem to be discussing much these days is "juicing" the ball. MLB doesn't allow people to measure official baseballs in any way, so there's no way of knowing for sure, but I would be willing to bet that they've been having the balls wound tighter since the mid-90's in an attempt to get more people to the ballparks. It would hardly be the first time that something like that had happened.

In the 70's, MLB adjusted the height and distance of the pitching mound of a few years. You can see the effects that the new rules had in the statistics of the time, and the experiment was abandoned a few years later. The point is, you can't really point to one thing and say "that's the problem" without complete information. Even without changes, baseball has undergone several "eras" where either pitching, fielding, or hitting has had a disproportionate effect. Just take what the media has to say about these things with a grain of salt is all. Remember, their trying to sell your attention.

petrel: What I might do as a quick and dirty fix is to make an assumption about the first year steroids have a serious effect on MLB. Clay's article changed all data from 1985 up. I suspect that might be a little early, but I'm fine with it.

Reduce all settings to 1985 levels. For example I think homers is something like .83 - use that instead of 1.00.

It won't completely erase the effect of steroids - but it should get the overall numbers back where they should be.
Doing this does work, but keep in mind that doing this will give you "unrealistic" modern day numbers. I've done several statistical evaluations, and I can tell you for certain that the offensive game picked up way before the "steroid era" supposedly began. Higher offensive numbers actually clearly started in 1986 (after the first strike. I might have the year wrong here, as I'm writing this off of the top of my head, but it's right around there), which is long before people figured that steroid use became widespread. That, in and of itself, leads me personally to believe that MLB changed the baseball rather than there being some steroid induced offensive power spike. According to all accounts, steroids can't make you hit for a higher average or on base percentage anyway, which is what has happened in recent years.

[JustinM]

And then you have years like the mid-60s where pitchers completely dominated batters. In 1968, Carl Yastrzemski won the AL batting title with a .301 average. Tony Conigliaro won the 1965 HR title with 32. That would be unheard of today, but there's nothing saying that sort of thing won't return someday.