Last weekend over dinner, I engaged in a spirited discussion with Mickey Lichtman and Jon Sciambi about a singularly vexing question: What's happened to the hitters?

Hitting went way up in 2006 and stayed (almost) there in 2007.

Hitting was down a bunch in 2008 and stayed there in 2009.

Hitting was really down a bunch in 2010 and dropped a bit more in 2011.

Now, there's an easy and obvious explanation: STEROIDS.

The problem with that explanation is-- well, actually, there are a couple of problems.

One is that the current drug policy was instituted in 2006; that season, three players were suspended for 50 games. But offense was (again) up that season and held steady the next. Why the dramatic decline in 2008, and again in 2010? Do the numbers support the notion that

The other problem is that nobody's yet found a clear relationship between hitting statistics and steroid use. Yes, I know. Mark McGwire. Sammy Sosa. Et cetera, & ad infinitum. I'm just saying that if you actually try to nail down the actual impact of steroids on hitting statistics, it's a slippery thing. Especially when you consider the fact that pitchers were using drugs, too.

Anyway, my personal opinion is that the drug policy has made a difference, but perhaps not in the obvious ways that would precisely explain what's happened to hitting stats these last few years. Other factors might include weather, a conspiracy to deaden the baseballs, new pitcher-friendly ballparks, just your usual random variation, etc.

Those are the things we talked about last week, over dinner. Without coming to a satisfactory conclusion.

And then, as if magically summoned by our little troika of baseballites last Friday night, here comes Dave Golebiewski at BaseballAnalytics.org, with some stunning data that somehow I had never seen. Here, for the last four seasons, are the percentages of taken pitches in the strike zone -- as measured by PITCHf/x -- that were actually called for strikes by the umpire:

2008 74.5
2009 76.3
2010 78.9
2011 79.1

That's a massive increase.

Nut graf from Golebiewski:

Those extra strikes have massive effect over the course of a full season. The difference between 2008's called strike rate on in-zone pitches taken and 2011's rate amounted to an extra 5,732 called strikes on hitters. The difference between a called ball and a called strike is about 0.15 runs. Multiply that by 5,732 and you have a nearly 860 run decrease compared to 2008's rate. That's 0.35 runs per game, or nearly the entire difference between 2008 and 2011 run scoring.

And there are some nifty graphics with heat maps showing precisely where all those extra strikes have been getting called: just above the knees. What's more, the umpires are still missing a bunch of them down there.

Or maybe they're not. The PITCHf/x strike zone can't be perfect, due to what defines the batter's natural stance. But it's pretty damned close. And this strongly suggests something that I've suspected for a few years: As much as we piss and moan about the umpires, they have been improving, presumably because a) PITCHf/x gives MLB a great teaching tool, and b) PITCHf/x exposes the umpires' missed calls, so they're trying harder to miss fewer calls.

I'm not saying this explains everything. But it sure goes a long way, doesn't it? Maybe it's really as simple as the umpires calling the strike zone that's in the freaking rule book. More often, anyway.

Update/Corrections: Via the comments at The Book Blog, we find that a) it's actually 0.17 runs per game rather than 0.35, and b) there's a lot of play in the measured strike zone, as defined by Golebiewski. So it's hard to say exactly what impact, if any, the umpires have had on scoring. As always, the search continues...