Starting Pitcher Ratings: Additional Thoughts on the Top 100

Over the last three posts, we’ve gone through the top 100 pitching careers by GSDev combo rating (100-51, 50-11, and 10-1). This time, we’ll further explore the list as a whole and see what additional conclusions can be drawn.

Let’s start with a subject I have specifically avoided during the main posts about the top 100, because I’d like those to be comparatively timeless and not specifically focused on the occasionally bumpy development process for the ratings. As such, I did not reference the mathematical error that led to mid-publication changes to the GSDev system. Now, it’s time. How many new pitchers joined the top 100 due to the correction of that error?

None.

Yes, really. It’s exactly the same 100 pitchers as before. This is not universal throughout the rankings; new pitchers joined the top 10, 20, 40, 50, 60, 70, and 90, and likely almost every other multiple of 10 past 100. But the gap between pitchers 100 and 101 was unusually large, and the original list didn’t have anyone who pitched in an extreme-deviation era (deadball or early ‘80s) who either barely missed or barely qualified.

However, while the list of pitchers is the same, the order of the list underwent some shuffling. Here are the particularly significant changes – gains first, then drops:

Pitcher	Old Rnk	New Rnk	Change
Grover Alexander	11	7	+4
Curt Schilling	10	8	+2
Christy Mathewson	22	20	+2
Cy Young	47	39	+8
Ed Walsh	53	44	+9
Johan Santana	49	46	+3
Eddie Plank	60	54	+6
Nolan Ryan	7	10	-3
Steve Carlton	8	11	-3
Dave Stieb	41	47	-6
Steve Rogers	44	53	-9
Ron Guidry	48	55	-7
Jack Morris	58	63	-5
Tommy John	74	79	-5

That’s everyone who moved by 2 spots or more in the top 20, 3 or more in the top 50, or more than 3 from 51-100. The results are almost exactly what you’d expect from previous discussion of the changes: deadball pitchers move up, early ‘80s stars move down. The pitchers who don’t fall into one of those categories were affected indirectly by them; Schilling passes Carlton and Ryan, and Santana passes Stieb, Rogers, and Guidry. The shifts are not insignificant – replacing a member of the top 10 is at least somewhat of a big deal. But they also aren’t overwhelming.

If nobody new joined the top 100, who was left just outside of the arbitrary cutoff? As mentioned above, this particular endpoint happens to be well-placed; the difference between #100 and #101 shrank somewhat after the changes, but it’s still about the same as the difference between #100 and #96, or #101 and #108. That being said, here are the pitchers who came closest to making it, ending up with combined scores between 29 and 30:

Rank	Pitcher	Years	RSS	WtSum	Combo	bWAR Rnk
101	Bobo Newsom	1929-53	29.58	76.8	29.76	78
102	Jamie Moyer	1986-2012	29.12	77.7	29.71	85
103	Jake Peavy	2002-16	29.36	77.0	29.69	169
104	Curt Simmons	1947-67	28.90	76.8	29.44	132
105	Urban Shocker	1916-27	29.33	75.5	29.39	64
106	Burleigh Grimes	1916-34	29.00	76.4	29.39	104
107	Dolf Luque	1914-32	29.64	74.6	29.37	120
108	Chris Carpenter	1997-2012	29.88	73.9	29.34	180
109	Charles Bender	1903-17	28.87	75.7	29.19	133
110	Sam McDowell	1961-75	30.06	72.1	29.09	129
111	Dennis Eckersley	1975-87	29.19	74.3	29.08	42

For only including 11 pitchers, that list is spread out pretty well; it includes at least one active pitcher throughout the period from 1903 to 2016.

Eckersley’s career is significantly longer than is reflected above; the listed seasons are the ones captured by the GSDev system, which is to say every year in which he had at least one start. He of course would go on to another decade of often excellent relief pitching, and that’s why he’s in the Hall of Fame. That relief work also explains his position on the bWAR list; if you remove his seasons from 1987 (in which he had 2 starts and then fully moved to the bullpen) onward, he drops to approximately #111 in bWAR from 1901-2022, which is a pretty remarkable match between the systems.

Speaking of bWAR rankings, I provided that information for each pitcher in the GSDev top 100 as we went through the list. There were 20 pitchers in the GSDev top 100 who fall short of the same standard according to bWAR. That in turn means the reverse is true of bWAR’s top 100 list from GSDev’s perspective. Here are the two lists of 20 next to each other, for the sake of comparison. (The WAR and Dev columns are rankings, not totals.)

Pitcher	Years	WAR	Dev	Pitcher	Years	WAR	Dev
Dennis Eckersley	1975-87	42	111	Steve Rogers	1973-85	114	53
Jack Quinn	1909-31	52	156	Jon Lester	2006-21	124	61
Eddie Cicotte	1905-20	56	117	Cliff Lee	2002-14	134	62
Urban Shocker	1916-27	64	105	Jack Morris	1977-94	123	63
Rube Waddell	1901-10	65	133	Chris Sale	2012-22	111	65
Waite Hoyt	1919-38	67	141	Adam Wainwright	2007-22	131	69
Larry Jackson	1955-68	72	118	Vida Blue	1969-86	117	71
Wilbur Wood	1961-78	73	286	Jacob deGrom	2014-22	140	75
Dutch Leonard	1933-51	74	129	Lefty Gomez	1930-43	128	76
Bobo Newsom	1929-53	78	101	Javier Vazquez	1998-2011	126	78
Kenny Rogers	1990-2008	80	152	David Price	2008-21	143	82
Vic Willis	1901-10	81	220	Gerrit Cole	2013-22	205	84
Eddie Rommel	1920-31	82	316	Bucky Walters	1934-48	106	86
Jamie Moyer	1986-2012	85	102	Fernando Valenzuela	1981-97	168	87
Wes Ferrell	1928-41	90	162	Catfish Hunter	1965-79	174	94
Wilbur Cooper	1912-26	91	123	John Lackey	2002-17	158	95
Mel Harder	1928-47	92	144	Josh Beckett	2001-14	179	97
Addie Joss	1902-10	97	113	Corey Kluber	2012-22	187	98
Nap Rucker	1907-16	98	198	Paul Derringer	1931-45	149	99
Hippo Vaughn	1910-21	100	143	Madison Bumgarner	2009-22	206	100

The GSDev-favored guys we’ve covered to some extent already; most of them pitched either in the late ‘70s and early ‘80s or within the last 20 years. The exceptions to that rule (Gomez, Hunter, Walters and Derringer) are generally guys who had their peaks on teams with good defenses (so bWAR downgrades them a bit), and also had at least reasonable playoff success.

The bWAR list is the one we haven’t explored at all yet. Some of the biggest differences there are deadball pitchers (Cicotte, Waddell, Willis, Cooper, Rucker, and Vaughn), for the same reasons we’ve talked about already. Some are pitchers with long careers but few big seasons (Quinn, Leonard, Rogers and Moyer), who benefit from the lack of peak adjustment in the bWAR list. But by far the most disproportionate drops are from Eddie Rommel and Wilbur Wood, both of whom have the same issue as Eckersley: over half of their career appearances were in relief. They went about this in very different ways. Rommel spent his career as a classic swingman. His 12-year stint in the majors featured at least 9 relief appearances every year, with only one season of fewer than 5 starts in the mix. Wood, meanwhile, changed roles mid-career. Through age 28, he had only 21 starts and 344 relief outings; starting at age 29, he ran off a five-year stretch of 224 starts, which was extremely impressive and also marked the end of his effectiveness as a pitcher. But however they went about it, both pitchers have over 20% of their career innings in a form that GSDev does not account for.

That said, Rommel’s best years aren’t wildly impressive even if you account for his relief work, and Wood’s peak is quite good but is more geared toward high volume than high efficiency, and we know GSDev’s inclinations run in the other direction. Even with their relief work factored in, I don’t know that either man would crack the top 200, let alone the top 100.

Mentioning the pitchers who GSDev ranks significantly higher than WAR brings us naturally to our next topic: active pitchers. Here are the pitchers who were ranked in the top 100 by GSDev who were also active in 2022:

Rank	Pitcher	Years	Combo
13	Clayton Kershaw	2008-22	48.37
15	Justin Verlander	2005-22	47.66
19	Max Scherzer	2008-22	44.57
29	Zack Greinke	2004-22	40.76
65	Chris Sale	2012-22	33.54
69	Adam Wainwright	2007-22	33.42
75	Jacob deGrom	2014-22	32.79
84	Gerrit Cole	2013-22	31.87
98	Corey Kluber	2012-22	30.43
100	Madison Bumgarner	2009-22	30.20

Of those ten pitchers, four of them (Greinke, Wainwright, Kluber and Bumgarner) either haven’t pitched since 2022 or have pitched ineffectively enough that they likely haven’t moved up in the rankings at all. As far as the other six are concerned, we have preliminary numbers in 2023-24 to work with. deGrom had limited but effective duty in those years and appears to have moved up roughly 10 positions in the rankings as a result. Kershaw, Verlander, and Scherzer all had pretty productive 2023 seasons, then did little in ’24. If they had started out in deGrom’s vicinity, they would have moved up farther than he did over those two years. But the higher you go on the list, the more distance there is between the pitchers and therefore the harder it is to make progress. My best estimate is that ’23-’24 will move the three future Hall of Famers up by at most two spots, and Kershaw in particular may not move at all; his deficit behind Bob Gibson is respectably large.

That leaves two more pitchers, who are both in similar positions. As was mentioned during the top-100 posts, Gerrit Cole won his long-awaited Cy Young award in 2023, and Chris Sale made a remarkable comeback to do the same in 2024. Both pitchers were also reasonably productive in their non-Cy Young seasons. As such, Sale looks to have moved up nearly 30 spots in the rankings, cracking the top 40; Cole appears to have jumped by about 40 positions, breaking into the top 50.

For 2025, I don’t even have especially reliable preliminary numbers yet (since multi-year park factors won’t be final until the end of the 2026 regular season), but Sale and deGrom are the only pitchers listed above who look to have moved much at all.

Of course, the ten pitchers in the top 100 aren’t the only notable hurlers who were active in 2022. There were several others ranked between 101-300:

Rank	Pitcher	Years	Combo
112	Stephen Strasburg	2010-22	28.98
120	Johnny Cueto	2008-22	28.63
148	Yu Darvish	2012-22	27.00
188	Aaron Nola	2015-22	24.62
202	Anibal Sanchez	2006-22	23.81
218	Carlos Carrasco	2009-22	23.11
222	Lance Lynn	2011-22	23.04
245	Sonny Gray	2013-22	22.44
268	Charlie Morton	2008-22	21.83
271	Dallas Keuchel	2012-22	21.78
272	Jose Quintana	2012-22	21.78
279	Zack Wheeler	2013-22	21.69

Much like the top 100 group, several of these guys were either completely or almost completely washed up as of ’22; Strasburg, Cueto, Sanchez, Carrasco, and Keuchel have not moved at all. Lynn, Quintana, and Morton have all made some progress up the list, but also look likely to fall far short of the top 100 (particularly with Lynn and Morton now having retired).

The other four (Darvish, Nola, Gray and Wheeler) have all made significant jumps in the last three years, with all but Gray likely ranking among the top 130-ish by the end of 2024 (Gray lags the others by 20 spots or so). None appear to have joined the top 100 at that point, but Wheeler’s 2025 has a good chance to have pushed him over the hump.

Outside the top 300, we’re getting deep into “way too soon” territory. Guys like Kevin Gausman, Luis Castillo, Blake Snell, and Max Fried were all among the top 500 and have made progress, as have some pitchers who were even lower ranked three years ago (Nathan Eovaldi, Corbin Burnes, Zac Gallen, Logan Webb), and of course some pitchers who weren’t of interest at all in 2022 (particularly Tarik Skubal and Paul Skenes). Hopefully, that crop of younger pitchers will go on to produce its share of all-time greats as well.

As it happens, active players are not the only group who stand a good chance to gain ground in these rankings in the near future. After several years of focusing on other projects, Retrosheet (which is the source Baseball Reference uses for its game logs in older seasons) has released box score data for the 1898, 1899 and 1900 seasons this year, and one might reasonably hope that more 1890’s seasons are on their way. As such, here are the highest-ranked pitchers who were active in 1901 and pitched a significant amount before that season (“Debut” means first year with multiple starts):

Rank	Pitcher	Years	Combo	Debut
39	Cy Young	1901-11	36.99	1890
133	Rube Waddell	1901-10	27.80	1899
220	Vic Willis	1901-10	23.07	1898
238	Joe McGinnity	1901-08	22.66	1899
246	Jack Chesbro	1901-09	22.37	1899
273	Jack Powell	1901-12	21.78	1897
331	Deacon Phillippe	1901-10	20.39	1899
338	Bill Dinneen	1901-09	20.19	1898
377	Noodles Hahn	1901-06	19.16	1899
381	Harry Howell	1901-09	19.14	1898

The difference between these guys and the active pitchers is that we already know what they accomplished in their careers, we just don’t yet know what GSDev thinks of it. Waddell only had 22 starts in 1900 (and even fewer in 1899), but led the 1900 NL (the year’s only major league) in both ERA and FIP. Clearing the distance to the top 100 is a big ask for one 22-start season, but the FIP/ERA combo could at least make things interesting. By contrast, Willis and McGinnity are adding multiple seasons of workhorse performance, with good if unspectacular game-to-game results. It wouldn’t surprise me if either or both of them move up 100-plus spots once those seasons are included. This trio may still end up lower than you’d expect for Hall of Famers, but there should be notable improvements on their current standing.

And then, of course, there’s the first half of Cy Young’s career. By bWAR, Young was the best pitcher in baseball four times before 1901; if you ask fWAR, it’s a full half-dozen league leading efforts in that time. If we ever end up with anything close to his full career on file, Cy Young has a pretty good chance to crack the all-time top 10.

That covers the pre-AL period as well as we can for now; the other top pitchers of the 1890’s (Kid Nichols in particular) will exist in the realm of pure speculation until such time as we know how GSDev will handle that era.

And on that note, let’s close with a look at the GSDev top 100 through the lens of eras, once again comparing to total bWAR. We’ll use decade of median career start to classify pitchers:

Decade	bWAR	GSDev
1901-09	8	5
1910-19	6	3
1920-29	9	4
1930-39	7	7
1940-49	4	3
1950-59	4	4
1960-69	7	6
1970-79	14	16
1980-89	7	8
1990-99	15	15
2000-09	13	15
2010-19	6	14

The difference is less stark than it was for the top 100 seasons; as noted above, 80 of the 100 pitchers are shared between both lists. But all of the decades in which bWAR has more top-100 representatives are earlier than any of the decades in which GSDev does. In particular, bWAR top 100s outnumber GSDev 23-12 before 1930, with most of that difference accounted for by pitchers from the 2010s. The shift in the median median season (super-median?) is from 1975 for bWAR to 1984 for GSDev; that may not sound overwhelming, but it represents 13 older pitchers being replaced by 13 newer pitchers. (The other seven changes are age-neutral, such as Paul Derringer and Lefty Gomez replacing Wes Ferrell and Mel Harder as representatives from the ‘30s.)

And that’s it for the top 100 careers. Next up, I plan to swing back through the seasonal stats for another comparison of GSDev against the WAR systems. This time, rather than comparing pitchers of different eras, we’ll be looking at a much simpler question: which pitcher does each system think is the best each year?