For some reason I got the bright idea to tinker with Bill James' Game Score statistic.
I don't remember exactly why, but my original idea was to average out game scores for the pitching staff to have yet another angle in which to evaluate the Astros rotation. And that in itself has worked out well. I've got both a Major League database and a Minor League database and now with one quick look at the spreadsheets I can tell you who has been the better pitcher according to Game Score.
One of the things I like about Game Score is that it gives you a singular number for one outing. This allows us then to compare starts without having to look at a bunch of numbers like innings, walks, strikeouts, hits, etc.
However, this isn't a new statistic and it's actually been around for over two decades now, its just never caught on. So I had an idea that might make it more appealing. What's evolved from playing around with Game Score, I think, is even better than that original idea of averaging out a players game score.
For an explanation of Game Score check out my article explaining the statistic.
I am by no means the first one to tinker with it to try and make it more appealing, Both Jeff Angus and Tango Tiger have had their go with it. And now I'm throwing my hat into the ring with those individuals whose brain power far exceeds my own.
My first proposal for making Game Score more appealing is to simply double it. This moves the average line from 50 to 100 and aligns it with other statistics that use a similar scale. OPS+, ERA+ and wRC+ all use a the scale where a 100 is average. To a certain extent FIP- and ERA- use the same scale, except in reverse order.
I prefer to call it Pitcher Rating or PR, but I guess it could be called something like GSc+.*
My thinking for not preferring the "+" at the end of the statistic is that it usually means theirs a more accurate and sophisticated calculation to the statistic. My method does neither, but instead make it more appealing, which is why I prefer PR.
My belief as to why Game Score has never caught on is that with a Game Score typically being on a 1-100 scale people tend to associate it with the school grading system: 50 is a F, 60 is a D, 70 is a C, 80 is a B, 90 is an A and 100 is an A+.
Over the course of a season the top pitchers in the league are going to have an average game score between 60-70. That's a "D" in school and probably results in a meeting between you, your parents and your teacher. Yet, in baseball that's an outstanding year that may result in All-Star appearances and end of the year awards.
In 2000 Pedro Martinez pitched quite possibly the best season in baseball history. His Game Score for that year 73.28, which will keep your parents and teachers off your back but won't win you any points with your teachers or parents. Doubling that average makes it 146.55 aligning it with similar statistics. It's not in line with Pedro's ERA+ of 291 for that year in the context of Game Score, however, it makes the statistic more appealing to our psyche.
Diving deeper into Pedro Martinez's Game Score's for the 2000 season, and looking at it by a per start basis, the two 98's he posted in his seventh and 28th start become a pitcher rating of 196 for each start. His worst start of the season was his 14th start in which he posted a 53 Game Score. Converted to PR that's a 106, which is still above average and further highlights just how good of a season he had. In either case simply doubling the Game Score appears to make it more appealing and in-line with other statistics that use the 100 baseline as an average designation.
Traveling further down the rabbit hole I then came up with the idea to put Pitcher Rating inline with the NFL's Passer Rating statistc.
Anyone that has watched football is already somewhat familiar with Passer Rating. Essentially it calculates a quarter backs game statistics and turns it into a single number with which to base a quarterbacks performance. It's much more complicated than Game Score and has it's own detractors. In fact ESPN decided to create its own statistic to rate quarterbacks called, Total Quarterback Rating (QBR). Here's ESPN's reasoning for coming up with it's own statistic to rate quarterbacks:
Do we all really need another uber-stat? Well, yes, because there are two basic problems with traditional passer ratings: what they measure and what they don't measure. The official formula for passer rating is actually less complicated than its reputation. It takes completions, passing yards, touchdown passes and interceptions, all on a per-attempt basis, compares each to a league-average figure, and mashes them into one number. But passer rating doesn't attempt to weight its categories by their importance to winning football games. It just averages them together, which tends to bias scores heavily in favor of QBs who complete a lot of short passes, driving up completion percentage without necessarily generating more yards or points. It's even possible, absurdly enough, to improve your rating by throwing passes for negative yards.
Very sabermetricsie
The focus here isn't on the new QBR statistic and so I'm going to avoid talking about the new statistic and move on. If you are curious thought and want to read more you can read the article here.
While Passer Rating like, Game Score, is a flawed statistic it is one of the more well known stats and something I thought could be used as a connection between baseball fans and football fans. Finding a way to draw other sporting fans into a baseball discussion would only help to promote the sport.
So how do we accomplish that? Again -- and I hope I don't blown any minds with this -- simply multiple Game Score by 1.6 to get a Passer Rating like, Pitcher Rating.
Is it exact? No, however, the average passer rating from 2000-2009 is 75-85. 50 multiplied by 1.6 is 80.
The maximum Passer Rating a quarterback can get is a 158.2.
Phil Humber's perfect game has a Game Score of 96 attached to it. If we multiply that by 1.6, to make it similar to Passer Rating, it comes out to a 153.6 Pitcher Rating, which is very close to that 158.2 mark. If Humber had struck out three more batters his Pitcher Rating would of been a 158.4, which is even closer to Passer Ratings max score.
Pedro Martinez's 2000 season we discussed earlier comes out to a 117.24 Pitcher Rating which at the time would of topped any quarterback's passer rating up to that point. Up to 2000 Steve Young had the highest Passer Rating of any quarterback posting a 112.8 Passer Rating in 1984.
Since then, and with some of the new NFL rules favoring offense, Tom Brady, Peyton Manning and Aaron Rodgers have all surpassed Steve Young.
Rodgers and Manning are the only two to beat Martinez's 117.24, with 122.5 and 121.1 respectively. Tom Brady's 117.2 Passer Rating in 2007 is pretty much the exact same as Martinez's Pitcher Rating.
After mulling over these ideas and particular the last one it seems even to me a little bit ridiculous, however, I think making Game Score more appealing could allow for casual fans to get more involved in the game, it would give analytically inclined fans another method in which to evaluate a pitcher and it would get football fans involved with the sport using a statistic they're already familiar with.
Getting more people involved in the game even if it's based on the simplest of statistics can only be a good thing for baseball. In my next post I'll apply these to the Astros pitching rotations in the Major and Minor Leagues.
Any thoughts or feedback?
Reference:
Does "Game Score" Still Work In Today's High-Offense Game? by Jeff Angus
Game Score (and Crowd-Sourcing) by tangotiger
New formula will redefine QBs by Peter Keating
Breaking Down Average Passer Rating Performances by Jason Lisk