I know a lot of you are concerned about the Braves’ starting pitchers. Let me start by saying that every fan of ever team ought to be concerned about their starting pitchers every March. Even when the Braves had Maddux, Glavine and Smoltz, there were still concerns about the other two starters. And every pitcher is an ulnar collateral ligament away from an extremely well-paid vacation. And all of our top 4 pitchers have injury question marks of some sort, as do most of those in the wings.

You know who doesn’t? Bryce Elder. Instead, of course, the criticism directed at Elder is whether or not he sucks. He doesn’t. But before I get into that, a few preliminaries.

He’s the fifth starter, for Pete’s sake

That’s not Pete Alonso, by the way… it’s some saint. People spend an immense amount of time evaluating contenders for fifth starter, Obviously, should there be a bunch of injuries, fifth starters have a way of becoming de facto higher starters. And his All-Star year notwithstanding, a pitching staff led by Bryce Elder is a barely competitive MLB pitching staff. But that’s not his fault — it’s the fault of the guys who got hurt. Evaluating who the fifth starter ought to be is just not all that important to becoming a playoff team, and it’s irrelevant during the playoffs, as the fifth starter is the first guy relegated to long duty in the bullpen.

He’s been an average pitcher, and there’s nothing wrong with average

I’ve been playing with evaluating starting pitcher by mean and standard deviation of game score. There are a lot of things I like about it. But in his 79 career starts, he has an average Game Score of 49.4. 50 is average. I’ll discuss this more below, but Elder is not an objectively bad pitcher. He generates an average of just over 1 WAR per year. His ERA+ (92) is uninspiring, but not bad for a fifth starter.

Looking at him, people want him to be something he’s not

Elder is listed at 6’2″ 220. I actually think he looks bigger than that. People want him to throw like Roger Clemens and instead he throws like Mark Buehrle on a day when Buehrle’s control wasn’t quite there. That is yet another thing that’s not his fault. It is damned difficult to be good enough to be a starting pitcher in MLB; doing that while matching what your physique ought to do is dramatically unfair.

He is being compared to people who aren’t on the team

The focus on Elder is really a focus on Alex Anthopoulos. He is the visible symbol of deals AA hasn’t made. (The existence of such abandoned deals are simply assumed.) Yet again: not Elder’s fault.

Game Score

In a post last August, I discussed Game Score. I’ve been looking at since then and I’ve decided it’s a really good summary stat for starting pitchers. Bill James, who invented the idea, was not really enamored with it — and I’m using Tom Tango’s version (the one FanGraphs and MLB.com use) that James likes even less. But it has a lot of great features:

  • It isn’t defense-independent, but it’s offense-independent, at least in an age where pitchers don’t have to be substituted out for pinch hitters.
  • It is really simple to calculate, unlike WAR
  • It is a counting stat at the game level, but a rate stat at the career and seasonal level
  • The only other common stat used to evaluate the start, the Quality Start, is a very crude binary. The other binary stat, Win, is brutally offense-dependent, and gives you no extra credit for extra outs, giving no credit for saving the team bullpen innings.
  • It lines up with the history of the importance of starting pitching. Here are the average Game Scores by decade, along with the correlations between pitcher Game Score and both the pitcher’s own win and the team’s win
DecadeMean ScorePitcher Win CorrelationTeam Win Correlation
189052.70.600.54
190060.20.560.49
191058.30.590.48
192051.60.610.50
193050.50.620.49
194053.30.620.49
195051.00.630.49
196053.80.590.47
197053.00.580.47
198051.50.550.45
199049.80.520.44
200048.80.500.45
201051.20.500.45
202050.30.490.42

Great Pitchers, Great Average Game Scores

As everyone points out when they discuss Game Scores, Kerry Wood’s 20 strikeout game is the highest game score ever. High individual game scores aren’t that critical to me. But when we look at pitchers who made over 200 starts (subsetting it to pretty good pitchers) the highest 5 average game scores belong to Ed Walsh, Addie Joss, Walter Johnson, Christy Mathewson and Cy Young. Getting past the ancient immortals, the top 5 since I was born are Bob Gibson, Sandy Koufax, Jacob deGrom, Pedro Martinez and Clayton Kershaw. (There are surprises: Smoltz and Maddux are just below the top 100, and Glavine is just inside #300.)

What About Consistency?

Many of you will know the old joke about the three statisticians who go hunting. The first one fires three feet in front of the deer, the second fires three feet behind the deer, and the third one jumps up and say “Got it!” My new respect for the Game Score is the fact that the variance in Game Score is, I think, the best measure we have of starting pitcher consistency. This was the basis for my Elder evaluation last August, in which I produced this chart:

This was the graph of his appearances to mid-August. Now I can show his whole career:

2023 was excellent, 2024 was none-too-good, and 2025 was… up-and-down. But the point of the first chart was that Bryce Elder in 2025 had a very high standard deviation of Game Score: his highs were high and his lows were low. But after all, one year isn’t all that much data. Yet the variance is pretty high, even back in 2023 when he was good.

Before looking at that, a technical note: pitchers with low average Game Scores also tend to have higher standard deviations. This is because if you’re a bad pitcher with a low variance, you will find yourself in the minors. We can see this when we look over all pitchers: I took all the pitchers with 15 or more starts in a season since 2000. I divided them into quintiles of average Game Score and looked at the average standard deviation. (Each observation is a single pitcher-season with its own standard deviation.) Here is the result:

QuintileAverage Standard Deviation of Season’s Game Score
1: 32.3-45.819.2
2: 45.8-49.318.7
3: 49.3-52.318.6
4: 52.3-55.818.3
5: 55.8-75.6 17.8

So, Elder, who in his career has a 49.4 mean and a 21.7 standard deviation is unusually inconsistent. And that frustrates people. Heck, it probably frustrates Elder.

Whither Bryce?

But that’s the past. The current debate is whether Bryce, through consultations, has learned something. The hopeful point out that he had quality starts in six of his last seven starts in 2025. So what? I asked the question: what happens to pitchers who through their first 80 starts were within 1 of his average game score (49.4) and within one of his inconsistency (21.7). It turns out there were 49 such pitchers. Here is the complete list, ordered by mean score in their remaining starts, no matter how long.

Remaining StartsMean ScoreStd Score
id
Bobby Burke937.420.15
Alfredo Simón1337.621.95
Jerry Augustine2440.023.41
Brian Bannister3442.019.55
Jeff Robinson3743.520.85
Bill Greif1743.917.67
Paul Wilson7344.121.91
Duke Maas1144.220.21
Hal Gregg3744.421.01
Percy Jones3445.220.71
Bud Daley3645.918.68
Socks Seibold5546.222.42
Roger Pavlik4546.321.55
Hank Johnson3646.320.46
Curt Young8247.019.32
Pete Broberg5447.317.62
Cliff Chambers3347.618.04
Al Milnar4747.623.15
Matt Young8348.218.84
Derek Holland15348.221.32
Brooks Lawrence4748.620.60
Kris Benson12048.720.99
Jeremy Bonderman12348.819.85
Tim Wakefield39449.019.32
Ray Kolp9249.121.51
Kenny Rogers40349.318.77
Pat Hentgen22949.619.34
Arnie Portocarrero3749.921.58
Cal McLish12950.019.82
Matt Harrison2750.119.84
Joe Genewich8650.319.88
Harry Byrd2850.420.62
Jack Fisher18550.420.81
Tom Gordon12350.820.27
Bill Wight11851.219.72
Milt Wilcox20551.719.97
Al Jackson10451.820.63
Floyd Bannister28452.219.50
Alan Foster6852.618.24
Melido Perez12152.619.50
General Crowder21653.020.26
José Guzmán10653.118.01
Bobo Newsom41353.521.03
Bob Klinger5054.521.69
Kelvim Escobar12454.817.91
Mark Langston34954.920.20
Zack Greinke48357.118.70
Chris Short22857.121.11
Larry Jackson34957.320.34

There are things to note about this very diverse set of pitchers. Under half of them managed to produce an above-50 average game score for the rest of their careers. And most of those who did managed to reduce their inconsistency. Kelvim Escobar and Zack Greinke would be the best examples here. So for those who want to ditch Elder, you might be getting rid of an Escobar or Greinke. On the other hand, you only have a four percent chance of getting them. Much more likely is Socks Seibold or any of a number of pitchers nobody remember (although now that I’ve heard about him, I’m not forgetting Socks Seibold.)