Todd Jones, defender of the poor, oppressed, white, straight male has ventured his opinion on whether he thinks steriod testing has had an affect on baseball:

"Unfortunately I do. I hate it, but there has been a correction made in the system, and the numbers are going to suffer for a couple of years," he said Monday. "I hate to admit it because I didn't want to. I'm as disappointed as any fan would be that it's going to end up showing to be the truth. But it's got to be good for the game to get back to an even playing field. I just didn't realize how deep it was."

The evidence? The evidence lies in the fact that home runs are down 8.8% from last year. In 460 games this year, players have hit 1.97 homers per game as opposed to 2.16 homers per game through the first 459 games last year. Clearly this is a big difference, right? ESPN and many other news organizations have grabbed these numbers and nestled them in between comments from concerned former players. The only voice of dissent:

"I think five weeks is too short a statistical sample to draw any conclusions," said Bob DuPuy, baseball's chief operating officer.

Is Bob DuPuy right? Is this too small a statistical sample to draw a conclusion? Well, we are all scientists here, so let's find out. It is easy enough to figure out the statistic error in all of this (figuring out any other errors would be nearly impossible). Taking the square root of the number of games played to be the statistical error, we get about a 4.6% statistical error on the games played. This is pretty conservative, but better to be conservative when steriods are on the line.

So, if we factor in the statistical error, we get: 1.97 +/- 0.09 hr/gm in 2005 as opposed to 2.16 +/- 0.1 hr/gm in 2004. In case anyone at ESPN.com is reading this...you shouldn't quote a number without its error or we will all call sample size on you.

So those are the new values we are considering. But the question is still out there, is this significant? Is this likely a statistical fluctuation or is it something real? Well, if we subtract the two numbers we get 0.19 hr/game difference and the new error becomes 0.13 hr/game. The significance is given by just dividing these two numbers and we get 1.46. This is known as a 1.46 sigma significance. In science, we don't really consider this anything more than a statistic fluctuation until we hit about 3 sigma. So, Todd et al., there is a long way to go before you or anyone else can call this anything more than a random blip in the home run continuum. Call me at the end of the season and we'll see if there is anything real going on.

What I also find sort of funny is that their little sidebar graphic also shows that from 2001-2002 the number of homers per game dropped from 2.31 to 1.93 when supposedly everyone was still juicing. Where is the explaination for that? 71% of ESPN Nation disagree with me and agree with Todd Jones, according to the poll on their page. A few well placed numbers can be a dangerous thing.

So, in conclusion, there is no evidence that steriod testing has had any effect and I am a huge nerd. I'm glad we settled that.