Who Stole All the Runs in Major League Baseball? (Ep. 8)

Listen now:

Who Stole All the Runs in Major League Baseball?: Was it the batters, the pitchers or someone else entirely?

Last night, Roy Halladay of the Philadelphia Phillies pitched just the second post-season no-hitter in Major League Baseball history. Yes, it’s an amazing feat. And yes, it’s surprising. But if there were ever a season during which this feat might seem a bit less surprising, it’s this year.

Haven’t you heard? It’s the “Year of the Pitcher,” so decreed back in late spring, with a flurry of no-hitters that included two perfect games, one pitched by Halladay. A third perfect game was ruined on the final out by an umpire who later admitted he’d blown the call. So this season witnessed only the 19th, 20th and almost the 21st perfect games in baseball history – in the space of just 23 days.

Moreover, the number of runs scored in the majors this year hit an 18-year-low. There were 1,105 fewer runs than last year. That’s 1,105 guys who, instead of crossing the plate and scoring one more run to tell their grandkids about, instead had to jog back to the dugout, chin down. (On the bright side: that makes 1,105 times when the pitcher didn’t want to kick in the Gatorade cooler.)

Surely, this is all the “Year of the Pitcher” effect, right?

Maybe not. Our latest Freakonomics Radio podcast (subscribe at iTunes, get it via RSS feed, read the transcript, or listen live via the box above) asks a simple question: who stole all the runs in major league baseball?

Here’s one likely culprit: steroids. The theory goes like this. Steroids helped turn great baseball players into even greater ones, hitting previously unthinkable numbers of home runs. Then came the scandals, the grand-jury indictments, busted reputations, and new rules that banned performance-enhancing drugs and increased testing.

So it would make sense that players are backing away from steroids – becoming smaller, more human. Which would lead to fewer superhuman hitting stats.

But Steve Levitt isn’t so sure. In the podcast, you’ll hear him talk about his efforts to find steroid cheating in the baseball data (he’s done it with schoolteachers and sumo wrestlers, so you wouldn’t think this would be so hard), but he came up empty. His conclusion: “I ended up finding no evidence that steroids matter.”

Phillies celebrate after Halladay’s no-hitter in Game 1 of the NLDS against the Cincinnati Reds on October 6, 2010. (Photo by Chris Trotman/Getty Images)

Hard to believe. Just about everyone who cares at all about baseball is convinced that the Steroid Era contributed mightily to home runs in particular and runs in general. But here are a few confounding facts to chew on: the year steroids testing began, run scoring actually went up; run scoring is up in the minor leagues this year, even though they too have steroids testing; and in the majors, the recent run decline has come almost exclusively in the American League, even though testing there is the same as in the National League.

So if steroids can’t account for all the missing runs, what else might be happening?

This brings us back to the “Year of the Pitcher” theory. There are simply more pitchers who have gotten better at keeping hitters and runs at bay.

One problem with this theory: according to baseball stat hounds like Mitchel Lichtman and Hayes Davenport, pitchers aren’t gobbling up these runs on their own. Yes, it’s tricky to isolate a pitchers’ performance from the number of runs he gives up. But baseball statisticians these days have lots of tricks for doing that, stats like Fielding-Independent Pitching, Defensive Efficiency and Ultimate Zone Rating. (You can check out some of the numbers here.)

And what do those stats tell us?

That you can go ahead and call it “The Year of the Pitcher” if you really want, and we certainly wouldn’t blame Roy Halladay for saying so. But you should also be calling it “The Year of the Glove.”

In the podcast, you’ll hear from Lichtman along with Bud Black, the manager of the San Diego Padres (whose team, retooled last off-season to focus on defense, nearly made the playoffs this year after finishing next-to-last in 2009); and Doug Glanville, a longtime former major league player who was known for his defense.

Glanville was a great interview, a particularly thoughtful guy (he studied engineering at Penn, and comes from a family with a deep love of learning), who’s now written a very good book (see his excellent Times pieces too), and does some analysis for ESPN. His most interesting story in the podcast is something he calls “Milton’s paradise lost” (not the book), and concerns a moment of personal failure that I was surprised he was willing to talk about.

And best of all, you’ll hear Steve Levitt tell you why his hometown Chicago White Sox didn’t want to hire him to steal signs from opposing teams. But it’s okay. Levitt is still hoping to sell his larcenous methodology to another team.


please describe why glanville is the only person in this article (or from any recent freakonomics article) that comes "from a family with a deep love of learning"

for example, bud black is one of the smartest mgr's in the game, from a academic family, with a college-educated wife and daughter.

can i get the R^2 of the use of that term, by race? I'm sure it's nothing, but it might be an amusing data set, steve-o.


"...the year steroids testing began, run scoring actually went up..."

And last year, well after testing began, Manny Ramirez was caught. A better conclusion here might be that the beginning of testing was not the end of steroid use. Mr. Leavitt may also want to follow the money. Manny's punishment - losing nearly a third of his $20M annual paycheck - may have been what got players and their agents to finally reconsider the cost of doping. I'd like to hear from Mr. Leavitt in about ten years, after he's compared the data from, say, 1998 (when McGwire's record-breaking season showed players the financial upside of steroids) through Manny's nabbing, to the data from this season forward.


Note that the Giants' Andres Torres is ranked 1st or 2nd in range among all NL outfielders. He has great speed and, more significantly, an ability to anticipate the flight of the ball. The number of runs he has prevented and victories assured meant the difference between the Giants appearing in post season and sitting out the playoffs. Indeed, Fangraphs places his value at approx. $24 mil. His ability, and that of others like him, give credence to role that fielding has played in the apparent dominance of pitching.

One final note - The final play by the Phillie catcher, Ruiz, I believe, was just terrific amd again confirms the analysis under discussion.


Not steroids, but Human Growth Hormone (HGH) which is not tested for in MLB. Look at the advanced age of many successful MLB pitchers, whom years ago, would be considered old farts.


"That you can go ahead and call it "The Year of the Pitcher" if you really want, and we certainly wouldn't blame Roy Halladay for saying so. But you should also be calling it "The Year of the Glove." "

In my opinion, this is just an argument in favor of lack of steroids. Defensive players could be shining more because, now that steroids policy is more intense, batters are keeping their hits inside the field, instead of outside. If that is the case (I can?t support this hypothesis with statistics, sorry), then all the hits are getting outside the diamond in a lesser proportion compared to the past, which favors infield defensive players.


With regards to steroid use, I don't think it has diminished much at all. The risk of detection is quite small and the punishment is quite light. That is, in comparison to professional cycling.

Despite the strictest standards, most frequent testing and severest penalties for all Performance Enhancing Drugs (PEDs), professional cyclists, which earn a pittance compared to MLB ball players, still continue to use PEDs.

I wish Levitt followed up on his brief comment in the book Freakonomics about trying to detect cheats using data detective methods. What came of it?

Cyril Morong

If steroids don't help, how did Barry Bonds have his 4 best hitting seasons, by far, from ages 36-39? I have looked and I have not been able to find any other player like that in baseball history.

I think frankenduf in #1 made a good point that pitchers might use steroids. If steroid use by both pitchers and hitters fell with stricter testing, why would run scoring be down? it could be that more hitters were using steroids or it helped them more. I don't think we really know.

Sid pools

I think Halladay got a first-pitch strike on 23 of the 28 batters he faced. I wouldn't be suprised if the stats across major league baseball this year show a higher incidence of first pitch strikes than in previous years - and this puts the hitter at a distinct disadvantage.

It goes in cycles...and we are in the part of the cycle where pitchers think hitters are taking the first pitch - so they're throwing strikes. Hitters will adjust.

Adam Katz

It is unfair to compare seasonal run totals, especially with teams like the Tampa Bay Rays bringing us back to an era of small-ball and with defensive skills at (arguably) their sharpest to date. Instead, compare only home runs as they do not involve fielders or base-running skills.

However, even this appears to be limited in its use: I've heard many a discussion about the recently increased value placed on off-speed pitches (this is a major focal point in the "year of the pitcher" argument). A poorly thrown changeup is synonymous with "meatball," which often results in a home run or very long foul; extra power isn't needed if the ball is just lobbed at the strike zone.

Fielding is definitely the area to watch. We currently lack (to my knowledge) stats based on merit and instead make assumptions based on a lack of blunders (Fielding Percentage). Errors are seen as black marks mostly because we have no metric to offset them; why run down the ball if you're more likely to earn an Error than a Putout or an Assist?

I argue for "Awesomes" (putouts above and beyond the expected) to counterbalance the Errors (missed routine fielding plays) which sit on the other extreme of this spectrum. This assigns value to the players that go the extra mile and take risks, even though the players with the most Awesomes will likely earn an above-average number of Errors. The problem with my proposal is that awarding Awesomes would be pretty subjective (in the same way that awarding Errors is already, though likely more dramatically) and would have to be done in a very standardized manner.

The closest we have to Awesomes is something ridiculous, like "Golden Gloves per year," which is completely unfair since it's a popularity contest with limited awards per year. In case you're wondering, Ichiro Suzuki has 1.000 GG/y and Greg Maddux had .818, while Jason Varitek, arguably among the best defensive catchers in the history of the game (he caught far more no-hitters than any other catcher even despite not participating in this recent pitching era), has .091 and most other great defensemen have zeros).



It is all defence. To me the most valuable players in the majors are the guys who hit for a high average. The Blue Jays are a good example. They led the majors with home runs, but had such a hard time getting runners aboard that the vast majority of those dingers were solo shots. I would like to see some stats on the total base runners per run scored or errors per run scored.


1105 runs less is a 4.93% change since the last year. Getting your trousers all in a twist over less than 5 percent difference seems pretty asinine. And to frame it as guys that don't go home to regale their progeny? An excerpt from that family's evening meal:
"Honey! I found a $1 on the ground on my way home from work!"
"What? last week you found a $1.05 you horrible, horrible husband!"
"Hmmmm, it must be the year of the pitcher."
"Quite right. Here are the divorce papers."


the year steroid testing went up so did runs

Are you talking about pre-modern steroid testing before mandatory suspensions? If I recall, because of the powerful players union the previous iteration of steroid testing kept testing results under wraps and would not even disclose whether a player had tested positive or not.

Erik Engquist

Didn't I see a stat on TV the other day about the total ERA in the National League declining in each of the last four seasons? I think it was down a half a run over that time. So I don't get why the article says offense decreased only in the A.L. Re the other points, steroid testing could not have been expected to end steroid use immediately. In the first couple of seasons, the testing routines were fairly predictable and players could juice without getting caught. The statistical measurement of defense also played a role. I also think injured pitchers are being DL'd sooner than ever, which keeps their ERAs down (sore-armed pitchers are rarely effective). Questech cameras have enlarged umpires' strike zones; above the belt is a strike for the first time since the 1970s.

Erik Engquist

Here are the numbers. Indeed, offense is down in both leagues -- the mystery is how the Freakonomics guys could mischaracterize this as an American League-only phenomenon.

2006 AL: 789 runs allowed (4.56 ERA)
2006 NL: 785 runs allowed (4.49 ERA)

2010 AL: 716 runs (4.14 ERA)
2010 NL: 705 runs (4.02 ERA)


You should also have talked to Hittrackeronline, they have been tracking home runs by distance and deciphering the physics of home run trajectories.


Steroids might not have had a major effect on the total number of runs scored in the 90s, but there can be no argument that the number of Home Runs hit by individual steroid users was WAY up.

Jeremy K.

What I want to see is an analysis of the impact of banning AMPHETAMINES versus steroids.

That will put the Steroid Era and baseball purists in perspective, since 'greenies' had been widely used since at least the 70's.

Here's what I know as a baseball player who is familiar with some very accomplished baseball players.

1. There are relatively few home runs that just barely clear the fence, relative to the number that are out by 20 feet or more.

2. Hitting a home run has little to do with strength when bat speed and resonance are included in the discussion.

To explain #2 in justice would take a long time, and strength does contribute to bat speed but it's far from a linear relationship.

The jist of it is that a home run is hit when someone moves that barrel of the bat fast enough to provide the ball with enough momentum to get over the fence. Me and anybody else who has played the game can stand at the plate and take several hundred swings that are capable of generating home runs - the lion's share of the difficulty does not come from being strong enough, but from being alert enough, with quick enough reflexes to get the proper part of the bat to hit the proper part of the ball. Hence, I firmly believe that amphetamines were always the major drug of abuse in MLB and it explains why studies find limited contribution of steroids to hitters' performance.


Tom Laiso

My contention is that you raise a player's average by around 25 percent overall by playing the 1st and 3rd baseman towards the line and the OF 5 steps back. so .270 hitter is a .325 hitter. A .300 hitter is a .360 hitter. Therefore, with zero or 1 outs, you should play it normal or you are -ev.

If that is true based upon more singles being given up with zero outs, then more runs will be scored and you will lose the game more often.

[WORDPRESS HASHCASH] The poster sent us '0 which is not a hashcash value.