Saturday, November 06, 2010
Where did all the runs go?Posted by Dave Studeman
The following is an excerpt from the introduction to the Hardball Times Annual 2011:
The Hardball Times Annual proudly displays statistics you won’t find anywhere else, Batted Ball Stats. Even web sites like FanGraphs and Baseball Reference don’t provide what you can find in the Annual. I like to think of these numbers as a combination of statistics and scouting—numbers that quantify not only how well each player performed, but how he did it.
To give you a sense of how these numbers work, I’m going to review them in the context of 2010, the "year of the pitcher." Here are some facts: In 2007, major league teams scored 23,322 runs. In 2010, they scored 21,308 runs. Two thousand runs have been lost in the past three years. Let’s see where they went. First, let’s look at strikeouts and walks as a percentage of plate appearances for the last four years:
Year K% BB% 2007 17.1 9.5 2008 17.5 9.6 2009 18.0 9.7 2010 18.5 9.3The 1.4 percent increase in strikeouts in just three years is huge. The walk rate had also been climbing, helping to offset the run-dampening effect of the strikeouts, but it dropped quite a bit in 2010. Forget home runs or ballparks—this little table explains over half of the decline in offense since 2007.
Our system applies a "linear weight" to each batting event, a weight that reflects the average impact each event has on run scoring. When you multiply the quantity of each event by its weight, you get a number that is roughly equal to the total number of runs scored each year. The system works pretty well.
So when you apply the appropriate weights to the total number of strikeouts and walks the past three years, you find that the net impact has been a decrease of about 400 runs, 20 percent of our total.
But there is another consequence of an increase in strikeouts: The total number of pitches put into play has decreased. Strikeouts and walks have grown from 26.6 percent of appearances to 27.8 percent. As a result, there were about 5,000 fewer balls put into play in 2010 compared to 2007. According to our linear weights, the net impact of this drop was about 700 fewer runs. If you add the 400 and 700, you get 1,100 runs. Over half of the decrease in runs from 2007 to 2010 has been the result of more strikeouts and fewer walks.
Okay, so what happened when the ball was hit? Here’s a list of the percentage of batted ball types for all balls put into play (ground balls, line drives and fly balls):
Year GB% LD% FB% 2007 43.5 18.6 37.9 2008 43.9 20.2 36.0 2009 43.3 18.9 37.8 2010 44.3 18.2 37.5I should tell you that our statistics partner, Baseball Info Solutions, has video reviewers watch every batted ball to classify it. They even have people review the video reviewers. Still, it’s hard to tell the difference between a line drive and a fly ball, so there are certainly some judgment calls and biases in the data.
But when the overall data tell you that ground balls were up a full percentage point in 2010, you can believe the trend. And when the data tell you that line drives were down two full percentage points in just the last two years, you can believe that something was going on. As you can imagine, the average ground ball generates fewer runs than the average line drive. I estimate that the movement toward grounders and away from liners has resulted in about 250 fewer runs.
The rest of the difference, about 650 runs, is the result of the outcome of each type of batted ball. Below you will find a few factoids about batted balls: Outs per ground ball and the average number of runs generated by each grounder (GBR); the average number of runs generated by each line drive (LDR); and three different facts about outfield flies: home runs per outfield fly, outs per outfield fly and runs generated by the average outfield fly.
Year Out/GB GBR LDR HR/OF Out/OF OFR 2007 73.9 .046 .398 .103 82.9 .181 2008 74.1 .045 .386 .107 84.2 .175 2009 74.7 .041 .388 .109 83.0 .189 2010 73.9 .043 .383 .102 82.8 .179This mishmash of data accounts for 650 fewer runs during the three years under review.
- The out rate on ground balls had been trending up, and the subsequent run rate of grounders down, but they both came back to norms in 2010.
- The run impact of the average line drive has trended down. According to BIS, this is primarily due to fewer doubles and home runs off line drives. This trend could easily be a data classification issue, so we won’t dwell on it here.
- The run impact of an outfield fly jumped up in 2009 but fell back again this past year. There were fewer outfield fly home runs, but the out rate (that’s of outfield flies that aren’t home runs) also stayed down, which helped negate the run impact of fewer homers. Again, this may partially be explained by the relative classification of flies and liners. Only partially, however.
The rest of the change has been characterized by two things: a movement toward more weakly hit balls (fewer liners, more grounders) and weaker results from both line drives and fly balls.
Of course, these sorts of statistics can also be used for individual players. In our stats, you’ll find 2010 batted ball stats for all batters with at least 100 plate appearances and all pitchers who faced at least 100 batters. Plus, batted ball stats for all major league hitters and pitchers are available as an Excel download.
To get access to Batted Ball stats, please support the site and purchase the Hardball Times Annual 2011.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.