well, oh my

after all david, think of all those thousands and thousands and thousands of kids killed by roids. as opposed to the few hundred in iraq.

who needs to deal with stuff like criminals likerove and libby and boring stuff like iraq and medicare and unimportant stuff like the deficit when babe ruth's legacy of hitting baseballs is at stake!!!

and barry bonds is walking the streets a free man!!!

I think the 10 people suspended sort of proves their point, rather than invalidating it. I'd guess they believe that more than 10 people in the major leagues were taking steroids this year. I'm inclined to agree, in which case the testing is failing either in its methodology or its execution.

But the legislation does nothing to change the methodolgy, which is the real problem. The standards for a positive are very high, so people who should be marked positive get by. But you have to do this so the chance of suspending someone who is innocent is low. If you are going to suspend someone for a first offense for a long time, you better make sure you're right.

Ahhhh, I see what you mean. I guess there's potentially a math behind this, where you can set your alpha level to whatever percent of the population you're willing to get a false positive for.

Two years does seem strong for a first offense - both as a punishment for a player and as a length of time a team might be out a star player due to what may just amount to the crapshoot of who gets the false positive first.

But, even with 2 years as a penalty, I'm of the belief that there were more than 10 people in the MLB this year who were clearly on steroids to the point of me thinking 2 years is a good punishment. I guess it's just an opinion thing though, and I'm sure everyone has different notions about what punishments are appropriate for certain test failures.

well, at least they have nothing better to do. While they're at it, they should step in and legislate away the DH! (I kid.)

I wouldn't have a problem getting rid of records created by steroids (though I agree it isn't the business of congress), but how are we ever going to know? We can suspect Barry Bonds all we want to, but right now, we have absolutely no proof that would hold up in court. The same goes for Mark McGwire. The same will go for almost everyone. We simply will never know how many players have been using steroids or who they are, so there is basically nothing we can do to remove their records, even if we want to.

Mike, for what it's worth, it's also called the Theory of Signal Detection.

so mike,

if you believe that there were more than 10 people on drugs

what's the point of testing? you want a daily test? maybe 3 times a day. maybe they gotta live with a permanent needle in them so they can get maybe blood taken after every at bat.

or we can just use your "belief" - and the minute any guy goes above what YOUR statistical guess is for that guy, let's just assume he doing drugs and throw him in maximum security prison for 10 years.

because you don't believe the tests.

BAH

Not my statistical guess, the statistical guess of the test.

For example, most psychology studies use an alpha level of .05. So when you test a hypothesis, you're only going to believe that you're correct if the math says there is a less than 5% chance of you being wrong. In other areas, they use smaller alpha levels - airlines, for example, might run tests with a .001 alpha level... because anything more than a .1% chance of being wrong when you think you're right is very deadly.

But the problem is, if you use a .05 alpha level, then 5% of the time you're going to think you're right but you're actually going to be wrong.

What I'm talking about is the idea of where that alpha level is for the MLB steroid tests. David seems to believe it's pretty low - let's say .001, so if someone tests positive we're 99.9% sure they were on steroids. That's pretty good. But needing to be SO strictly sure about things is also letting a lot of people off. So at 99.9% certainty, maybe we only catch 10 steroid users each year. I'm suggesting that there are more users. So at 99% certainty, perhaps, we'd see 100 positives per year. And while we would falsely accuse more people in this way, to me I think it's worth that compromise.

In some senses, I completely trust the test. There are certain chemicals they are designed to find, and it's up to the testers how "certain" they want to be when they say someone is using or not using. What I don't trust is:

1) that the test can catch all manners of steroids

2) that the test is frequent/random enough

3) that players don't know when the tests are coming (i.e., I DO believe that they players have enough time to prepare for the test and either stop taking the product or somehow mask its presence in the body)

I hope this clarifies.

Lisa,

Don't forget that urine tests can't catch HGH. I forsee Giambi getting another "parasite" soon...

There are very few things that are improved by the government getting involved. When fans have had enough, baseball will clean itself up. Until then, call your congressman and tell him to take a year off.

Having congressmen take a year off would almost always be a good idea.

mike,

a false positive is a REALLY bad thing. because of all the people who feel righteous leaking confidential stuff like medical records and supposed grand jury testimony. you SHOULD know that guilty headlines are giant and oops we goofed so and so innocent in teensy weensy printing hidden away somewhere. you gonna destroy a man's life and reputation for WHAT?

and a lot of these drugs - them and their by products stay detectable for a long long time

and NO ONE has proven anything bout hgh helping baseball performance. so i don't care about it