Amongst the comments of an article I recently read, I encountered claims that Clay Buchholz’s 2010 season was a product of his burgeoning talent and completely luck-free. These claims were fueled by a belief that the sabermetric statistics that cause many to label Buchholz lucky are flawed and by extension useless.
I never was of the belief that Clay Buchholz is a bad pitcher. I think he’s actually quite good, but not as good as his 2.33 ERA from 2010 might lead one to believe. I sort of thought it was common knowledge that his 2010 season’s stats were luck-inflated. I’m finding, both among published baseball writings and my daily encounters with fellow fans of the game, that there is still a widespread resistance to sabermetrics. I constantly encounter an unwillingness to look past wins and ERAs to assess a pitcher’s performance. I encounter claims that these advanced metrics don’t really mean much, or that they don’t make it much easier to measure a player’s talent.
Often times, it is true that these stats don’t tell us everything we need to know. Often times, they are just indicators of future performance, rather than outright predictors. I agree that it is hard to quantify the effect of many sabermetrically-approved statistics, but there is compelling evidence that they are, in fact, valuable in understanding baseball statistics.
The arguments in favor indicators of Buchholz’s luck, as I see them are as follows.
1) A low K/BB rate
Clay Buchholz struck out 6.22 hitters per nine innings while walking 3.47 per nine for a K/BB ratio of 1.79. I have heard that a general rule of thumb is that you want pitchers to post a K/BB ratio of 2.00 or above. A pitcher certainly can be successful with a lower ratio, but the fact remains that the ratio tells us something about how much or little the pitcher helps himself with defense-free outs or harms himself by giving a free pass.
It’s a fair to point to make that pitchers in every group can be the product of luck, either good or bad. However, we can see the general trend clearly supports K/BB rates being linearly related to ERA. ERA is a perfect retrospective measure of how effective the pitcher was in the past, since the goal of the pitcher is always to allow as few runs as possible, whether by control, power, or luck. It can be a lousy predictor of how a pitcher will perform in the future and is often a lousy measure of how good a pitcher actually is, especially when you’re only looking at an ERA over a short period.
I am not denying Buchholz’s production in 2010. He gave up only 2.33 runs per nine innings, which is exceptional. There is no debate about his effectiveness in 2010. Without Buchholz, the Red Sox would have been much worse off, and they would be worse off if he weren’t returning in 2011. He was profoundly effective. I don’t think he has the ability to be that effective again (barring luck).
I looked at all 115 MLB pitchers who threw 130 or more innings in 2010. I divided them into three groups: those with K/BB rates below 2.00, those with rates ranging from 2.00 to 2.99, and those with rates of 3.00 or higher.
There are some cases of pitchers who seemed to defy the general trend. In these cases, the performance was influenced by other things.
For instance, James Shields is an exceptional control pitcher who posted a K/BB rate of 3.67, but he allowed a ton of home runs en route to a 5.18 ERA. The pitchers who seem to be exceptions to the rule can be viewed as outliers, but they do not disprove the general trend that more strikeouts per walk leads to fewer runs allowed.
Here’s what I noticed:
Pitchers with K/BB rates below 2.00 (37 pitchers, 6556.0 innings total):
Combined ERA: 4.32
ERA range: 2.33 to 5.94.
Median ERA: 4.47
Three of these pitchers wound up with an ERA below 3.00. Other than Buchholz, Tim Hudson spun a 2.83 ERA, owing a lot to his ability to generate ground balls. Trevor Cahill spun a 2.97 ERA, also showing himself to be a ground ball pitcher but pitching to a very low BABIP-against. Seven of these pitchers had ERAs in the threes. Some of these guys are great ground ball pitchers, some of them played in very favorable ballparks, and some were just lucky. The vast majority posted ERAs in the fours, with the median ERA being 4.47. Nine guys were above 5.00.
Pitchers with K/BB rates ranging from 2.00 to 2.99 (51 pitchers, 9691.33 innings total):
Combined ERA: 3.80
ERA range: 2.62 to 5.34.
Median ERA: 3.84
Of this group, seven pitchers had ERAs below 3.00. Brian Duensing led the pack with his 2.62 in 130.2 innings. He is a textbook case of lucky pitching, but I cannot argue with the seasons David Price and Ubaldo Jimenez put together. Three guys posted ERAs of 5.00 or above: Rodrigo Lopez, Kevin Millwood, and Tim Wakefield, a very homer-prone trio. 24 guys were in the threes, and 17 were in the fours.
Pitchers with K/BB rates of 3.00 and above (26 pitchers, 2060.0 innings total):
Combined ERA: 3.50
ERA range: 2.27 to 5.18
Median ERA: 3.53
This group was obviously the best, with their median ERA of 3.53 and combined ERA of 3.50. Only seven of these guys posted ERAs of 4.00 or higher, including James Shield’s 5.18. These included 12 pitchers below 3.40, half of whom were under 3.00.
2) The lower the strikeout rate, the more difficult it is to live with a K/BB rate under 2.00.
We’ve established that pitchers can be successful with a K/BB rate under 2.00, but the pitchers who do this tend to be the guys who strike out a lot of batters. That way, when they get themselves into jams, they are more likely to get out of it without having to involve the defense and the possibility of runners advancing or scoring.
50 pitchers threw 100 or more innings with a K/BB rate below 2.00. I used only unintentional walks to compute the group’s walk rate; however when I did this, seven pitchers then had a K/unintentional-BB rate of 2.00, but I still included this group with the understanding that an intentional walk still puts a hitter on base and there are never enough intentional walks to really distort our assessment of a pitcher’s control very much.
The group combined for a K/9 rate of 5.74 and a BB/9 rate of 3.20. Their collective ERA was 4.46. The median ERA was 4.56, showing that a comparable number of them were above and below the median.
Just looking at the guys with ERAs better than 4.56, they combined for a K/9 rate of 6.03 and a BB/9 rate of 3.30 (using only unintentional walks). The group with ERAs worse than 4.56 combined for a K/9 rate of 5.40 and a BB/9 rate of 3.07. Interestingly, the pitchers who gave up more runs allowed fewer walks, but the differences between the two groups strikeout rates are greater, which lends credence to the notion that if you strike out enough guys, you can get by with a low K/BB rate.
The pitchers with the 20 highest strikeout rates combined for an ERA of 4.14. This group had a K/9 of 6.83 to go with 3.45 walks per nine innings. The pitchers with the 20 lowest strikeout rates combined for an ERA of 4.71 to go with 4.70 K/9 and 2.95 BB/9.
Buchholz has thrown 364.1 innings in his career. Over the first 100 of them or so, he struck out over eight batters per nine innings. Since then, his strikeout rate fell to 6.65 in 2009 and 6.22 last year. During that time, his walk rate fell from about 4.7 per nine to somewhere near 3.5. I would interpret this as him learning to pitch in the big leagues and adapting as hitters got used to him.
Buchholz’s strikeout rate is towards the higher end of the group I was looking at (his was the 16th best out of the 50 pitchers). I would argue that, at this point, he actually does get enough strikeouts to survive a low K/BB rate; however, unless he either precipitously raises the K’s or lowers the BB’s, he is very unlikely to repeat his exceptional 2010 season, because while he does strike out a fair amount, he doesn’t compliment that with a low enough walk rate to be a star.
3) A lucky home run rate.
Buchholz has done a tremendous job generating grounders in his career. His grounder per fly rate is 1.64 over parts of four seasons. This is the thing I like most about him; however, I don’t think this accounts for such a low percentage of fly balls leaving the yard. Even in Fenway Park, where many a righty batter sees his home runs turn into doubles. His line drive rate was 17.7 percent, tied for 58th out of the 147 pitchers with 100 or more innings to their name in 2010.
Buchholz’s fly-ball rate was 31.5 percent, tied for 25th out of the same group of pitchers. This doesn’t explain why so few of these flies left the yard though, especially since last year, when he generated even more grounders per fly, 15.7 percent of the flies left the park. Only 5.6 percent of fly balls left the ballpark, fewer than all but seven of his peers. This is significantly lower than the average, which tends to be somewhere between nine and 10 percent.
Was Buchholz lucky in 2010?
A low HR/FB rate can be taken as a tendency to generate “short fly-balls” as someone once said to me. However, his infield fly rate is not especially high at 8.1 percent (tied for 57th). I could not find a hit chart for Buchholz, but I find it difficult to believe that a pitcher can control the depth of fly balls hit against him. This is a guy who, one year ago, saw three times as many of his fly balls go for home runs (an unlucky rate). Unless Buchholz completely reinvented himself, how does that change for any reason other than normal fluctuation? I don’t know.
4) A low BABIP-against
Sometimes the hits just fall in, and sometimes they don’t. The rate at which this happens can be alarming and can cause us to judge players more harshly or too generously. Buchholz’s .265 BABIP-against is not ridiculously low, but it does speak to a little bit of luck.
Out of the 25 pitchers who generate more grounders per fly than Buchholz, only Trevor Cahill and Tim Hudson showed lower BABIP-againsts. None showed lower HR/FB rates, so it seems clear to me that Clay Buchholz benefited from some luck. That said, ground ball pitchers tend to allow fewer hits anyway, simply because major league fielders can handle a higher proportion of grounders than flies.
I noticed that most of the ground ball pitchers I looked at tended to be somewhere around .280 or .290 in terms of BABIP-against. This often has a lot to do with defense as well, and the Red Sox played well in the field in 2010.
The conclusion I draw from all this is that Buchholz was lucky to have an ERA as low as 2.33, but he wasn’t that lucky. He looks to me like the sort of pitcher who should be capable of ERAs in the threes year in and year out. I think he’s going to continue to generate grounders and continue to strike out six to seven hitters per nine innings while walking 3.50 or so. This isn’t a recipe for brilliance, as we see guys like Tim Hudson and Derek Lowe who consistently posted decent to good seasons over the last decade.
Those two pitchers get more grounders though, and they walk fewer, strike out about as many, and face easier lineups these days. So, I consider Buchholz to be built from the same mold but a notch below talentwise. 2010 was to Buchholz what 2004 was to Jake Westbrook or what 2008 was to Paul Maholm (or what 2010 was to Jon Garland and Carl Pavano), an uncharacteristically good season that is unlikely to repeat.