Thursday, July 4, 2013

Fun With Sabermetrics Part I: Average Game Score

Travis Wood is having a very memorable season for the Cubs, seemingly dominating opposing hitters. The team around him might be faltering every step of the way, but Wood is blowing his peripherals clean out of the water. He isn't getting the wins that he deserves (only five), and without delving much into statistics, his season can be defined as merely "good". This is not the case.

Wood is tied with division rival Adam Wainwright, as well as Philadelphia Phillies' ace Cliff Lee for the league lead in quality starts, and is tied with Lee in quality start percentage. Coming into Thursday, both had 15 quality starts; both had quality start percentages of 88%. Those numbers are achieved by staff aces, something that Wood is not considered to be. Wainwright, being a former Cy Young winner and ace of the Cardinals, is sure to get accolades (or at least extensive media coverage) for his remarkable season; Lee will undoubtedly get the same treatment, and will be talked about on Baseball Tonight. Wood likely will not hear much about it outside of the Friendly Confines (but thats another story for another time).

One would expect that Wood and Wainwright would be atop most of the player efficiency ratings, as they have far out-pitched so much of the league. If you go off of one of Bill James' formulas, though, Wood and Wainwright aren't even on the map.

Average Game Score is a fairly simple formula that measures the strength of any given game by a pitcher. The formula is listed below:
1. Start with 50
2. Add 1 for every recorded out, and add two for each inning pitched after the fourth.
3. Add 1 for every strikeout, and take away one for every walk.
4. Subtract two for each hit, and subtract two for each unearned run
5. Subtract four for every earned run

For example, looking at Wood's game against the Oakland Athletics on Thursday: 50+22 (recorded outs)+4 (5 strikeouts and 1 walk)-3 (hits). That leaves Wood with a very solid score of 73. This start for Wood was excellent, but something that seems to be common place this year for Wood.

Adam Wainwright, facing off against the Angels in Los Angeles, did not have the same success as Wood tonight. His score would be: 50+32 (outs recorded)+2 (3 strikeouts and 1 walk)-14 (hits)-16 (earned runs). Wainwright did not earn a quality start, and ended up with a score of 54, which is still above average. The numbers make it seem like he struggled a bit, when in reality, this was still a solid performance.

Seeing as Wood has pitched as well as he has, and an off night for Wainwright still netted him an above average score, it should come as a surprise to no one that Wood's Average Game Score (before his Thursday start) was a 30.9, and Wainwright's (also before his Thursday Start) was a 32. How in the world does that make any sense.

Wood and Wainwright vary greatly in several categories, each of them strong in a category that the other has difficulty with.

Wainwright is really earning the AGS (Average Game Score) points when it comes to innings. Going into Thursday, Wainwright had pitched just over 133 innings, three more than AGS leader Clayton Kershaw. Pitching into later innings earns more points, and Wainwright's 133 are a real tribute to his stamina. Wood, on the other hand, has pitched 110 innings, and has not gone into particularly late innings. He pitches late enough to get quality starts, and does not go much longer.

Over his 133 innings, Wainwright has allowed 120 hits, a number that would be very troublesome if not for his low walk rate (only 13 thus far). Overall, Wainwright is allowing 9 base runners per nine innings. Kershaw has allowed far less hits, though he has walked more batters, and his WHIP (walks and hits per innings pitched) sits at an outstanding .93. Like Kershaw, Wood has not allowed nearly the amount of hits as Wainwright (75), yet has walked more batters, and has a WHIP of .98.

The category that has done the most damage to Wood is strikeouts, where he isn't even in the same ballpark as Wainwright and Kershaw. Wood has 80 strikeouts, which pales in comparison to Kershaw and Wainwright (126 and 117, respectively). Strikeouts are a large part of the reason that Wood sits nearly 40 averages points behind Kershaw.

To put it lightly, strikeouts have a massive impact on AGS. The highest achievable score for a pitcher would be 114. That would mean: 50+37 (outs recorded)+ 27 (strikeouts)+ 0 (hits, walks, or runs). In other words, a perfect AGS would mean that a pitcher threw a perfect game, while striking out all 27 batters. Cy Young threw a no hitter in the early 20th century, yet he earned a score of 88 due to the fact that he had amassed only two strikeouts. Kerry Wood threw a complete game shutout in 1998 in which he tossed 20 strikeouts (any baseball fan should remember that game). He allowed only two hits, and earned the highest ever AGS for a nine inning game. His score for that game was 105, only nine points away from perfection. Matt Harvey came close to that earlier this year, posting a 97.

When putting it into context, one has to question the validity of the AGS metric. There is no question that Kershaw has been better than Wainwright and Wood this season, but is it signifiant enough to separate them by nearly 40 points? Bill James' formula for AGS measures a quality start as a score of exactly 50. If Wood has averaged six or more innings and three or fewer runs (the dictionary definition of a quality start) 88% of the time, is it fair to say that he has been well below the AGS quality start number?

To throw another curveball into the mix, Cubs' frontline starter Jeff Samardzija's AGS is a 57.9. Scott Feldman, now considered a former Cub, has an AGS of 54.2. Have both of these pitchers been significantly better than Wood? This, quite frankly is mind-boggling, considering how much better Wood has seemed. He has a lower ERA than both of them; he has a much lower WHIP than both of them; he has more quality starts than both of them. If this doesn't confuse you, you're on a different wave-length than I am.

If AGS is the best measurement of a pitcher's efficiency rating, do quality starts even carry much weight? After all, a pitcher could conceivably have an ERA of 4.50 and have a quality start in every appearance.

No comments:

Post a Comment