The great myth of the "steroid era"
11 Comments Published by Colin Wyers on Friday, February 15, 2008 at 12:14 PM.
Must-read article documenting how the ball was juiced during the height of the "steroid era."
Labels: Baseball
That was a good read. I've long believed that other factors than steroids have had a larger impact on the home run totals and it seems that some people are beginning to quantify this, which the media will promptly ignore because it doesn't sell newspapers.
Colin,
I already did a far better job on this subject at my blog back in July:
http://badnewscubs.wordpress.com/2007/07/26/the-juiced-baseball-of-1987-endless-home-run-stats/
One of my most viewed posts to this day.
The ball was juiced in the 80's. In the 90's, it was modern fitness + steroids. (I don't know why that article went into the ballpark thing so in depth. It obviously wasn't ballparks and that's easily proved. Just look at the home run numbers at the stadiums that DIDN'T change like Wrigley, Yankee Stadium, Fenway, Shea, etc.)
As for juice, the Oakland A's of the late 80's looked demonstratively bigger than other teams. The joke was that they looked like a softball team. By the early to mid '90's, every team had guys like that. The numbers soared.
Meanwhile, I've really outdone myself this time. If you read the current article at my site, you'll find a far better version of ZORP 2.2 with far fewer calculations. A player's offensive worth is easily summarized in the amount of runs he's scored + the amount he's batted in. It's more accurate than WARP3, less fictional, and a far better barometer of every players' offensive contributions. (No replacement level players necessary!)
Also, I challenge any stathead to answer the RBI question. Why do players have about the same amount no matter where they play or how good their team is? Read my piece, and you'll see that the commonly held stathead notion that RBI's are 'team dependent' is really quite stupid, unfounded... and shouldn't be going unchallenged for so long... especially by numbers-crazed statheads.
This is sort of like how Bill James is revisiting clutch hitting these days, huh?
Dare take the challenge? Anonymous, where are you? Runs+RBI's truly is a minimalist, relevant, accurate Voros McCracken. If you use this stat along with obp and slugging... you need little else.
Uh... your "new stat" is Runs Produced. Bill James talked about it in his 1983 Baseball Abstract; I'm sure it's even older than that. Your inability to interact with the large body of work since then that has improved upon it is seriously not my problem.
As for replacement level - I can turn Runs Produced into a stat baselined at replacement level; I can take VORP and recast it as MLV and do away with the "replacement level fiction" that you needlessly obsess over. I could use average as a baseline instead of replacement. But what we're talking about now is an issue of data presentation, not of actual accuracy.
I'll go into more detail about this later, which you will then ignore because you don't seem to understand selective sampling and sample size issues and you think that throwing out thirty or so seasons spread between a half-dozen or so players proves your point.
Why do you continue to ignore the RBI question?
I'm happy to see that my conclusion is one of Bill James' from his early work. Personally, I wouldn't subtract home runs from the equation.
Beyond these numbers, these guys are simply taking your money. Show me anything, literally one single thing that ZORP3 shows me about a player's offensive contributions that runs+rbi's doesn't.
Don't give me this sample size bullshit unless you're going to prove something otherwise. Show me the players whose runs/rbi's numbers differentiated greatly based upon what team they played for.
It's all just time wasted and accomplishes nothing.
Name any two players that you have a better grasp on because of ZORP2.2 than I do with runs produced.
Show me anything, literally one single thing that ZORP3 shows me about a player's offensive contributions that runs+rbi's doesn't.
Why don't you try it as a rate stat? As in Runs+RBI/PA? That would allow you to better compare it to a lot of other statistics to compare its accuracy. I don't think that you will find it to be incredibly conclusive in its superiority as you seem to think. Part of the reason for lower Run+RBI totals could be due to batting at the bottom of an NL lineup and having one nearly automatic out immediately following you, mitigating your chances at Runs. I'm not sure as to its usefulness in valuing players in a non-fantasy setting as well.
A very good way to test the veracity of runs created is to compare current players and then also compare other statistics, such as VORP (since it includes any offensive contributions a player has over the course of a season) or wOBA.
The only real problem is that runs created doesn't account for defense at all, whereas stats such as WARP3 take defense into account pretty much as runs prevented above what a replacement level defender would provide. The replacement level moniker is relatively inaccurate as in order to play a position in the majors you need to be better than 'replacement' level at it unless you are a Prince Fielder or Ryan Braun caliber bat.
Samael,
You could use it as a rate stat, but then you're not factoring in health. I want to know the total amount of runs, period. Over the years, that will tell you more than the rate. I'm guessing Cliff Floyd probably hate a pretty darn good rate last year, but that certainly wouldn't tell the whole story.
Next, you have no idea how funny it is that explained to me that I should consider defense. As is well on record, no one considers defense as much as I do. I'm looking at this solely from an offensive standpoint, and not a player's overall value.
Next, it shouldn't matter where a guy hits in the batting order for this stat because simply, if a guy is producing a lot of runs in the back of the order, he isn't going to be there long. A guy's usually back there for a reason. And then still, my argument is that if you produce runs, you'll still produce plenty from no matter where you hit and what team you play for. Bill Mueller's 2003 season is an example of this. He produced just about as many runs as Jason Varitek, Kevin Millar, Todd Walker, Johnny Damon, Trot Nixon, and David Ortiz.
Finally, regarding VORP... my whole argument is against VORP and that it's a waste of time. If you look at runs produced, obp, and slugging, you really don't need the other metrics to tell you anything new. You'll have a stong sense of that player's offensive worth.
Meanwhile, I'm still waiting on godot for evidence that RBI's are 'team dependent' when all the numbers in the world suggest otherwise. This dogma... it's the one area that statheads really shy away from.
"commonly held stathead notion that RBI's are 'team dependent' is really quite stupid"
Yep. The idea that it takes teammates, or runners on base, to accrue RBI is quite stupid.
Do you understand what team-dependent means? It means that a player hitting in the 3rd spot in the lineup behind 2 guys who have a combined OBP of .300 will generate fewer RBI than if he had 2 guys who had a combined OBP of .350. It's really rather simple and, quite honestly, nothing more than common sense.
But, what you said, yeah! I agree. The stupid stathead notion that lineups have anything to do with run scoring is quite annoying. If you watch baseball like me and understand it as well as I do, you know that Player A is equally capable of driving in 100 runs while hitting 3rd behind Neifi and Corey Patterson than if he was hitting behind Pedroia and Youkilis. Because, you know, the number of times those guys are on base has nothing to do with how many runs the 3rd man in the lineup will drive in. Nothing at all.
You've convinced me. The small number of players you've more than likely taken a look at proves you are right. Runners on base has nothing to do with RBI. Spot in the lineup has nothing to do with it. A-Rod would have driven in 156 if he was leading off easily because runners on base don't matter. And he'd drive in the same number last season if he was on the White Sox rather than the Yankees. The Sox OBP of .318 compared to the Yankees .366 would have nothing to do with RBI.
It's also a fact that he'd have driven in 156 if he played during the Deadball Era as well because RBI is obviously not team dependent. I mean, RBI is so not team-dependent that when the cork centered ball was invented we didn't see any changes in players RBI totals. It's a fact. And when the mound was lowered, no change in RBI totals. Another fact.
The Hawk, I really hope you continue to post around here. You provide a great deal of joy to someone who really does enjoy the little things in life. Thank you for providing my Saturday morning funnies.
Oh yeah, save your 17-paragraph maniacal rant in response to my last comment. I assure you I will not read it as I have absolutely no intention whatsoever of coming back to this thread at any point in the future. I do not care what you say.
I'm sure you'll feel the need to impart your true genius upon me with thousands of swear words anyway, but I'm just trying to save you some time.
To the contrary Maddog, you read every single thing I ever write. It's obvious to everyone.
You, like all other statheads... cannot name a single player whose RBI stats show any team-dependent relationship. Show a single player whose numbers significantly go up or down depending on which team they've played for. Go as Corey/Neifi as you'd like.
And then, to use this 9th in the lineup garbage as an arguing point... if a guy is racking up loads of RBI's in the 9 spot, he won't be there for long. He'll move up in the lineup. It's irrelevant. Guys with low runs+rbi's will find themselves in the 7th and 8th spots and that's why they're there.
Lineup construction matters insofar as having a good leadoff man and a balance of lefties and righties. That stuff matters, the stuff you don't care for. But, the numbers clearly indicate that the RBI totals tend to stay with players no matter where they play.
Name a single player whose numbers illustrate team dependency for rbi's. You can't. At my site, as you're well aware, I showed A-Rod, Soriano, Carlos Lee, and more... who no matter whom they play for, the RBI numbers remain about the same. Your Neifi/Corey vs. Pedroia/Youkilis thing holds no water.
The STATS bare that out.
Why do you continue to contact me... over and over again? Your 8 paragraphs above do a great job of demonstrating how little you care what I have to say.
Please disregard everything I say. I am a mental midget.
Sincerely,
TheHawk