K/BB revisited: Switch to Run Impact

In my article from a few weeks ago, entitled “K/BB ratios: Does it matter how a pitcher does it?,” I hinted at my displeasure with the K/BB statistic. It’s intention is fantastic, but its implementation isn’t as good as it could be. It is misleading.

Strikeouts and walks are crucial statistics for a pitcher. You know that much by now. But they do not have an identical impact on a pitcher’s ability to prevent runs. They have different measures of importance, which K/BB does not account for. K/BB compares the two at face value, which simply is not an accurate depiction of a pitcher’s ability to maximize run prevention by accumulating strikeouts and limiting walks.

Look at the formula for FIP, one of the most basic ERA estimators:

13 x HR + 3 x BB – 2 x K+ 3.2
IP

As you can see, the weighting is different for strikeouts and walks. Now consider that this formula was developed in 2001. Why are we still using K/BB as one of our primary evaluators of pitcher skill when its flaw has so long been understood?

Quite frankly, I’m not sure, which is why I plan on using a different statistic from this point forward. In place of K/BB ratio, I’ll now be referring to a pitcher’s run impact derived from strikeouts and walks (K-BB RI seems like a simple abbreviation, but if you guys can think of a better one, feel free to let me know!). What this does is measures the relative impact a pitcher’s strikeouts and walks — combined — have on his runs allowed. In other words, it quantifies how much a pitcher gains (in terms of runs) from his combined strikeout and walk rates.

To calculate it, you first determine how many strikeouts, walks, and batted balls occurred per major league game. Next, multiply each event by it’s corresponding relative run value, giving you runs per game. For batted balls, you need to back-calculate runs per game using total league average runs per game and runs per game on strikeouts and walks. You also back-calculate batted balls per game by subtracting strikeouts and walks per game from total batters faced per game. Divide runs per game (on batted balls) by batted balls per game to get a relative run value on batted balls.

For each individual pitcher, you then multiply his strikeout, walk, and batted ball per game figures by each event’s relative run value. After doing this, you get a runs per game figure for each event. Add them up. Subtract this number from league average runs per game, and you arrive at the impact strikeouts and walks have on a pitcher’s runs allowed.

Here are a couple of lists from 2007, measuring the leaders and trailers in K-BB RI:

Leaders

LAST FIRST G GS IP K-BB RI K/9 BB/9
Bedard Erik 28 28 182.0000 1.24 10.93 2.82
Santana Johan 33 33 219.0000 1.04 9.66 2.14
Peavy Jake 34 34 223.3333 0.92 9.67 2.74
Kazmir Scott E 34 34 206.6666 0.88 10.41 3.88
Beckett Josh 30 30 200.6666 0.86 8.70 1.79
Vazquez Javier 32 32 216.6666 0.84 8.85 2.08
Hamels Cole 28 28 183.3333 0.79 8.69 2.11
Smoltz John 32 32 205.6666 0.78 8.62 2.06
Harang Aaron 34 34 231.6666 0.75 8.47 2.02
Burnett A.J. 25 25 165.6666 0.72 9.56 3.59
Sabathia C.C. 34 34 241.0000 0.70 7.80 1.38
Shields James A 31 31 215.0000 0.65 7.70 1.51
Hill Rich 32 32 195.0000 0.56 8.45 2.91
Lincecum Tim 24 24 146.3333 0.55 9.23 4.00
Matsuzaka Daisuke 32 32 204.6666 0.54 8.84 3.52
Haren Dan 34 34 222.6666 0.52 7.76 2.22
Gallardo Yovani 20 17 110.3333 0.49 8.24 3.02
Hernandez Felix A 30 30 190.3333 0.48 7.80 2.46
Young Chris 30 30 173.0000 0.46 8.69 3.75
Verlander Justin B 32 32 201.6666 0.46 8.12 2.99

Trailers

LAST FIRST G GS IP K-BB RI K/9 BB/9
Trachsel Steve 25 25 140.6666 -1.22 2.88 4.41
Maroth Mike 13 13 78.3333 -1.00 3.22 3.79
Kennedy Joe 27 16 101.0000 -0.96 3.74 4.28
Ramirez Horacio 20 20 98.0000 -0.89 3.67 3.86
Sowers Jeremy B 13 13 67.3333 -0.80 3.21 2.81
Lowry Noah 26 26 156.0000 -0.77 5.02 5.02
Hernandez Livan 33 33 204.3333 -0.74 3.96 3.48
DiNardo Lenny E 35 20 131.3333 -0.71 4.04 3.43
Cook Aaron 25 25 166.0000 -0.69 3.31 2.39
Thompson Brad 44 17 129.3333 -0.67 3.69 2.78
Perez Odalis 26 26 137.3333 -0.64 4.19 3.28
Bacsik Mike 29 20 118.0000 -0.62 3.43 2.21
Glavine Tom 34 34 200.3333 -0.61 4.00 2.88
Duke Zach 20 19 107.3333 -0.60 3.44 2.10
Litsch Jesse 20 20 111.0000 -0.60 4.05 2.92
Pelfrey Mike A 15 13 72.6666 -0.59 5.57 4.83
Redding Tim 15 15 84.0000 -0.58 5.04 4.07
McCarthy Brandon P 23 22 101.6666 -0.56 5.22 4.25
Durbin Chad 36 19 127.6666 -0.55 4.65 3.45
Chico Matthew B 31 31 167.0000 -0.55 5.07 3.99

Concluding thoughts

As we would expect, those with good strikeout and walk rates are at the top of the list, and those with poor strikeout and walk rates are at the bottom. This simply gives us a better way of quantifying exactly how valuable the combination of strikeouts and walks are. One day soon, we’ll check out some guys who are looked at in a better light with K-BB RI than they were using K/BB and visa-versa.

If you have any questions, feel free to let me know!


Comments are closed.