It’s been feeling a little dull around here since WordPress nixed the MLB-themed blog pages, so I finally made it feel a little more like home. Now, what say we do a little “blogging?”
Game Score is simply another way to look at a starting pitcher’s performance, be it an individual start, a month, season, or career. It’s not perfect, and Tom Tango has altered the original Bill James model (used on Baseball-Reference). Tango’s new GmSc formula is listed in the pitcher game logs at Fangraphs. I have a BB-Ref play index subscription – which is not only time saving but also unbelievably deep – so I always refer to the original Game Score.
I have used GmSc to look at pitcher performance from a number of different angles. I categorized every start from 2016 as either a win, loss, or no decision, with qualifying scores set for each league. That was a serious undertaking, one which I probably won’t do again, but it did, in my opinion, make a pitcher’s win-loss record actually mean something.
Until recently, the Game Score metric I was most obsessed with was called Advantage. The idea came from a post I read by Bill James on his website, and I’ve written about it on this blog before. All you need to calculate Advantage are a pitcher’s starts and his average GmSc. Not much to it.
ADV = (Avg GmSc-50) * GS
Using BB-Ref’s Play Index, I spent the offseason logging Advantage scores for pitchers who made at least 15 starts in a season. My big hairy goal was to track those scores for every season back to 1913 (that’s as far back as BB-Ref goes for game logs), and I’m about 15 1940’s & 50’s seasons from completing the data… which is pretty darn cool!
But there’s a problem with Advantage. The greatest individual seasons in MLB history are as follows:
Grover Cleveland Alexander (1915): 920
Sandy Koufax (1965): 894
Bob Gibson (1968): 887
Walter Johnson (1913): 839
Koufax (1963): 792
Let’s not fool ourselves, those are some damn good seasons. But most of those were earned by a pitcher who made 40 starts that year. So how can we compare Clayton Kershaw’s greatness in this generation when he’s never eclipsed 600 ADV? In fact, the last guy to top 600 ADV in one year was Pedro, with 676 in 2000. Kershaw was close in 2015, mind you (594), when he DIDN’T win the Cy Young.
So, as much as I enjoy using Game Score to calculate a pitcher’s “Advantage,” it really doesn’t work to compare today’s starters to previous generations. I knew it was time to add a new wrinkle to the data.
Enter Game Score Plus, a simple number that compares a pitcher’s average GmSc to the average of his peers (starting 15 games or more) that season. It works just like ERA+ or OPS+, as a 100 score is equal to the MLB average, and each point above or below 100 is worth one percent. Unlike those other metrics, though, you don’t have the crazy discrepancies between terrible and great seasons. In any given year, the best and worst performers are separated by roughly 50 points. The highest GmSc+ in MLB history was a 151 (Pedro 2000), the worst that I can find is in the 60’s.
The five amazing seasons listed above? None of them crack the top 10 GmSc+ seasons in MLB history (though a few did make the top 15).
There’s so much we can do (and I have already been doing) using Game Score Plus. More posts and explanations to follow, but that’s all the time I have for tonight. In the meantime, here’s a little something for you to chew on. The 20 best individual seasons in MLB history, according to GmSc+. To Pedro and the Big Unit, we bow down to you…
What are your first impressions? Thanks for reading, and I’ll catch up with you again in the near future.
*Seasons highlighted in red are for AL Cy Youngs; Blue obviously for NL Cy.