Lies, Damned Lies and Statistics
Team Clemens' latest defense doesn't prove anything
Posted: Tuesday January 29, 2008 11:57AM; Updated: Tuesday January 29, 2008 2:21PM
Roger Clemens, master of the high, hard one, reared back and threw another 18,000 words at his credibility problem this week. "An Analysis of the Career of Roger Clemens," prepared by his agents as if bringing his entire career to an arbitration hearing, follows The Denial by Lawyer, The Online Video Denial, the 60 Minutes Appearance, The Lawsuit, The Angry Press Conference and The Taped Phone Call in Team Clemens' multi-platformed defense of the Rocket. And yet the greatest success of his statistical manifesto, however thorough it may be, is as modest as were his other maneuvers: It proves that Clemens was right when he told Mike Wallace how difficult it is to "prove a negative."
To read the manifesto is akin to listening to that awkward phone call between Clemens and Brian McNamee. It gets your attention, but in the end it doesn't move the meter on whether you think Clemens is clean or dirty.
The casual fan -- or, say, your garden variety congressional representative -- could never make it through the 45 pages of charts, graphs and year-by-year numerical deconstruction of Clemens' career. And yet the fan rabid enough to read through the Clemens Report should find its "conclusions" patently obvious: that Clemens' career, just like all careers great and small, had peaks and valleys and that he was at his very best when he added the split-fingered fastball at a time when he still had his premium velocity.
I asked Randy Hendricks, one of Clemens' agents, via e-mail what was the purpose of the report, seeing that it stated what already was evident to most baseball observers.
Hendricks wrote, "What is evident to you and most baseball observers, as you say ... is obviously not evident in the general world, based upon the things that I have read and heard in the public domain."
Translation: We have to stop the perception that steroids shaped the second half of Clemens' career. What was missing in Hendricks' reply, and in the report itself, is any direct mention of steroids. Here's the closest the report comes to referencing the real issue at hand: "It has been suggested that Clemens' performance during July and August of 1998 was unusual."
"Unusual," of course, is code for the Mitchell Report's claiming that McNamee was sticking needles full of steroids into Clemens' butt back then.
Do all the numbers prove that Clemens did not use steroids? Of course not, not any more than the sudden spikes in any career prove something fishy was going on. Steroids aren't magic beans. Do they work? Sure. I've heard from enough players who have used steroids to know that they enhance strength, confidence, recovery and, yes, performance. But the effect is more nuanced than we like to think, subject to the drugs, dosage, baseline of ability, age, health, etc. Alex Sanchez and Ken Caminiti were not enhanced in the same measurable manner.
Team Clemens' blizzard of statistics is impressive unto itself, but I don't think a congressman wants to depose Clemens regarding how many pitches per start he averaged in 1989. At issue is whether McNamee told the truth about injecting Clemens up to 21 times in 1998, 2000 and 2001. And any good fan of statistics knows that with some selective editing you can make the numbers line up and behave as you like.