Last night, I ran across an article on Yahoo saying that home runs are up this year (and not just in Jim Thome's stat line).

Then I moseyed to ESPN and was told that an AP-AOL Sports poll found that 53 percent of baseball fans feel that "Baseball has fallen short on keeping the sport drug-free." They also found that two-thirds of fans are down on Barry Bonds though more than half would still put him in the Hall of Fame.

To quote my dad, "What the hex!" What's going on here?

The fans are correct in being suspicious of baseball's intentions where it comes to a drug policy. Both the players union and the owners profited by the inflated offensive numbers of the past dozen or so years. Both were slow to address the issue, but they now have a policy that appears effective. I mean, at least there have been a handful of minor leaguers already suspended for almost a full season (100 games).

But some will question why home runs are up if the drug policy is working so well? Oddly, those are the same people rejoicing over Barry Bonds' painfully slow start this year. I have trouble with this sort of cognitive dissonance myself.

Then again, there's the next level of cynics who say that the drug testing is academic since major-league players have switched to untraceable types of enhancements like HGH (Human Growth Hormone) or prolonged sessions on Howard Stern's sybean [sp?]. Minor-leaguers get caught using steroids because they can't afford the good stuff.

Finally, your trek through the baseball cognoscenti caste system, you will reach nirvana. That is, you will realize that there are many, many other factors that have conspired to prolong the home run inflation.

Expansion and tinkering with the strike zone and the ball have all been factors. However, in my opinion, the largest factor is the high turnover in baseball stadiums, usually from the old standard tin can-type parks to today's revenue-generating bandboxes. The new stadium boom, you might notice, coincides almost perfectly with the current offensive boom.

OK, so it's just a theory, but maybe we can test it…

First, let's verify that there is an issue. The numbers look impressive in the Yahoo article, but what do they mean? Let's look at the numbers for Aprils in the last twenty years and see what they mean as compared to the season as a whole (current stats through last night's games, for all years stats are through April and may include March stats):

Balls are flying out of the ballpark more quickly (i.e., in 3.52% of all at-bats) than in any other full season. However, April 2006 ranks behind 2000 and is just ahead of 2001. In those seasons, homers fell sharply after April.

However, on average April home run rates are slightly below the season average (i.e., at 99.10%). One would expect if this holds true that then homers will comprise 3.55% of all at-bats by season's end, the highest percentage ever.

Ok, so there seems to be a potential issue here, but could this season just be an aberration? Homers had been falling in previous seasons, right?

Well, yes, home runs were slightly down last season, but it wasn't 1968 or anything. Besides slight downturns happen, how does the current decade compare with the past? Let's see…

So where's the change coming from? St. Louis has a new ballpark. Could that be skewing the data?

Here is a comparison over the last two seasons of ballpark factors using ESPN.com data (based on PF = ((homeRS + homeRA)/(homeG)) / ((roadRS + roadRA)/(roadG))). The table is split to fit in the column:

Now last year's data with the difference between the two:

What happened to Coors? And moving the left field back two feet in Philly really mattered? Hmm… And what happened to PETCO and the BOB, former pitcher's parks?

Could it be this is what happens when you use a small sample of data, like what we have so far? Sure, but I'm too impatient to wait for the season's end.

Do stadiums really change that much from year to year? Well, for historic ballpark data we are left with Baseball-Reference.com/Sean lahman data, which unfortunately uses a different format from ESPN. However it is useful for looking at annual changes to data.

Here are the greatest increasex in ballpark factor (BPF) from a previous season. You'll notice new ballparks are a factor:

Now, here are the greatest decreases:

Let's just look at stadium changes (i.e., not just stadium name changes). I'm including all new stadiums since the dawn of time since I love this stuff:

Now, here are stadiums for new ballclubs since 1901:

You'll note that almost all of the new stadiums in the last 15 seasons have had ballparks that have BPFs greater than the league average. You may also notice that some parks have more ups and downs than Jason Giambi on a rollercoaster. I thought it might be fun to look at the standard deviation and variance for BPFs:

I think a couple of things are happening: As more homers are hit, the variation among team stadiums as well as for a given stadium from year to year goes up just because we're dealing with bigger numbers.

The other thing is that teams are building a number of idiosyncratic stadiums, and as each is built a cookie-cutter tin can stadium bids adieu. With the old Busch gone, Shea's the last one left, right? These throwback stadiums create throwback instability. This decade's BPF changes are closer to the 1890s than the 1990s.

The end result? Get used to the homers, and don't expect that a hitter's park one year will necessarily be one the next. Is it steroid use? Doubtful. Oh, and get off Barry Bonds' back.