Pitching Run Differential

Ok, so, this is a stat I came up with last October at Braves Nation, and decided to take a second look at it this season. For anyone who missed it the first time around, the original thread can be found here. There's a short description of it there, but for those of us who are too lazy to check out a link, here it is again:

Honestly, it's a very simple analysis, with a lot of room for error and misinterpretation, but here goes...So, what I did was take the number of runs that a pitcher per out, and compared that to the average number of team runs scored over that same period of time.

Think of it as an adjusted run differential for pitchers.

Granted, I could have taken this a step further and calculated the actual run support a pitcher received instead of using the overall team average, but, honestly, I think run support is about a luck-oriented stat as there is. The purpose of the game is to score runs, regardless of who is on the mound, so I just went by what the team could do on any given day.

So, what you're gonna see is a positive or negative numerical value which represents, on average, where a team stands when Pitcher X departs the game.

There you have it. Average runs scored minus average runs allowed. Simple enough.

Hell, I'm in a giving mood for those of you who may still not be getting it, so I'll throw out an example:

Lance Harbor is a starting pitcher for the West Canaan Coyotes. The West Canaan offense, lead by leadoff hitter Wendell Brown, is scoring runs at a pace of 4.76 per game -- or 119 in 25 games, over 224 innings (672 outs). This total comes to 0.1771 runs/out for the Coyotes lineup. Over Harbor's six starts, he's worked 32 innings (96 outs) and allowed 12 runs. So, he's getting 16 outs per game and letting up two runs per game. So, in his time in the game, the offense is scoring 2.8336 runs and he's allowing two runs per game, this gives him a run differential of +0.8366 R/G. So, obviously, he's doing his part, it's in the hands of the bullpen and offense as to what the outcome of the game is.

Pretty simple: A positive number says the pitcher is doing his job and he's usually leaving the game with the lead, so, the higher the better. A negative number says he's not carrying his weight and is more times than not leaving the game with his team trailing, further away from zero, the worse he is.

Just to take the explanation a step further, here is how the Braves starting pitchers faired last season, their number of starts in parenthesis:

Hudson: +1.11 (34)
Smoltz: +1.12 (32)
James: +0.41 (30)
Carlyle: -0.44 (20)
Davies: -0.77 (17)
Reyes: -1.16 (10)
Cormier: -1.25 (9)
Redman: -1.59 (5)
Bennett: +1.21 (2)
Lerew: -1.21 (3)

Alright, got that? Smoltz and Huddy = Awesome. Cormier and Redman = Crap.

Let's take it to 2008.

Hudson: +1.19 (12)
Jurrjens: +1.00 (11)
Glavine: -0.10 (9)
Reyes: -0.36 (6)
Smoltz: +1.67 (5)
James: -1.76 (5)
Bennett: +0.93 (3)
Campillo: +2.66 (2)

Starters: +0.56
Bullpen: +0.35
Total: +0.91

Last season, the Braves staff had a total run differential of +0.47, so that lets you see just how much better the pitching has been thus far. The starters were at +0.17 (a decent jump) and the bullpen was right around the same, at +0.30.

The truly shocking stat from last year to this season is that last season, the Braves trotted out a starting pitcher with a positive run differential only 41.98% of the time, compared to a 62.26% for this season. The best in the NL East last year was the Mets at 85.19%, while the division champ Phillies were at 57.41%.

Then again, it's still very early. We're just under a third of the way into the season thus far, so there's still a lot of baseball left to be played. I'll keep tracking this and updating as the season goes on. I'll also take a look at some other teams around the league when I get a chance.

Note: I promised Luke a look at the differences of pitch count on offense comparing home and road splits, which I thought I'd get up tonight, but that's not gonna happen. Look for it sometime tomorrow afternoon.