The 2026 Cy Young Races, Updated: Both Favorites Got Stronger
Updated August 22, 2026. This is a rerun of the July 9 break edition, which is kept unchanged so the regression calls graded below can be checked against what it actually said. Companion to the MVP piece. Stat lines are through games of August 21; projections and Cy Young odds are from the August 22 simulator run over the 481 games still to be played.
BBMI Research DeskUpdated August 22, 2026Model run: August 22 · 500 simulated seasons
At the All-Star break the model had Cam Schlittler at 61% in the American League and Jacob Misiorowski at 47% in the National. Six weeks later both are further ahead, at 72.6% and 61.6%, and both got there the same way: by pitching well enough that the regression the model built into their projections never arrived in the size it expected.
One note on how we rank them, unchanged from July. BBMI grades Cy Young on runs saved, meaning a pitcher's projected run prevention measured against the league's average, scaled by the innings he is projected to throw. That rewards both the rate and the volume to deliver it, and it is why a 1.60 ERA in 90 innings loses to a 2.16 ERA in 158. The league is allowing 4.54 runs per nine this season, a 4.17 ERA. Everything below is measured against those two numbers.
Two favorites, both pulling away
BBMI Cy Young-model probability: the share of 500 simulated finishes in which he leads his league in runs saved. August 22 run; the July 9 figure is in grey.
American League
Schlittler
72.6%
61.0%
Messick
10.8%
n/a
Cease
6.4%
3.0%
Melton
3.8%
n/a
Gray
3.2%
n/a
Rasmussen
0.8%
12.0%
National League
Misiorowski
61.6%
47.0%
Yamamoto
18.0%
19.0%
Sánchez
11.4%
16.0%
Sale
6.6%
6.0%
E. Rodriguez
2.0%
n/a
Burns
0.4%
0.0%
Bars share one scale across both leagues, so Schlittler's 72.6% is the strongest favorite's position on the board. In the grey column, "n/a" means the pitcher was not on the July board at all. Tarik Skubal, the AL's number three in July at 7%, was traded to the Dodgers and now appears in the National League. BBMI season simulator, runs-saved basis.
American League: the favorite widened it, and the field turned over
Of the six AL arms the July piece named, four have either slipped, changed leagues, or gotten hurt. Schlittler is not one of them.
In July the model projected him to finish 14-8 with a 2.74 ERA and 219 strikeouts, and called that "regression to an award season." He has beaten it. His ERA has moved from 2.05 to 2.16 across 53 more innings, which is a rounding error rather than a correction, while the strikeouts went from 122 to 193 and the FIP gap actually narrowed, from +0.47 to +0.45. The new projection is 13-8, a 2.53 ERA and 234 strikeouts over 196 innings, with a 1.02 WHIP; 80% of sims land his strikeout total between 227 and 242. Nobody in the American League is close on runs saved, at +38.7 against a field whose next-best is +30.3. He is also pitching leverage every time out. The Yankees are 73-55, in the playoffs in 100% of sims but only 32% to win an East that Tampa Bay still leads.
The biggest riser in either league, from nowhere on the July board to the model's clear AL number two at 10.8%. The case is volume plus a good-not-elite rate: a 2.54 ERA over 149 innings, 9.2 strikeouts per nine, and a projection of 181 innings with a 2.78 ERA and 185 strikeouts. The caution is the same one the model applies everywhere, since a +0.57 FIP gap says something under a half-run of that ERA is being done for him, and a 9-8 record on a 63-66 Cleveland team (33% playoff) gives voters nothing to hang a narrative on. But on runs saved he is genuinely second in the league, and it isn't close between him and third.
In July we wrote that his 2.76 ERA "actually understates the arm," pointing at a 2.27 FIP nearly half a run below it and a bloated .316 BABIP. The ERA has since come to him, down to 2.42, with 201 strikeouts and a 13.1 K/9 that leads the league. He is the model's projected AL strikeout leader at 251, and his odds tripled from 3% to 6.4%. The ceiling is still the walks, at 3.9 per nine and a 1.13 projected WHIP, and the team, since Toronto is 63-67 and 20% to reach October. But of every "the peripherals disagree with the ERA" call we made in July, this is the one that paid in full.
In July we called Gray "the field's loudest AL regression warning" and wrote, bluntly, "the 9-1 is not coming with him." He has since gone 7-2, and his ERA has moved a tenth of a run, from 2.69 to 2.79. On the surface, we were wrong. Underneath, the profile we objected to hasn't changed at all: a 3.60 FIP still sits three-quarters of a run above the ERA, a 15.5% strikeout-minus-walk rate is still middling, and the model projects a 3.02 ERA the rest of the way. Which is to say it still doesn't believe him, now at 3.2% and sixth in the league in runs saved. The honest version: our mechanism looks right and our conclusion has been useless for six weeks. A 16-3 pitcher on a 69-59 Boston club that is 98% to reach October is going to get Cy Young votes whatever our number says, and the model has no way to price that.
The lowest ERA in either league belongs to a rookie the model does not believe for a second. A 1.60 ERA over 15 starts sits on a 3.84 FIP, a +2.24 gap and the widest in the sport, built on 7.6 strikeouts per nine, 2.8 walks, and ten home runs allowed in 90 innings. The projection tells the story: a 2.48 full-season ERA, a full 88 points above where he sits, and only 127 innings because he started the year in the minors. Every ingredient of the Eduardo Rodriguez case we flagged in July is here, in a louder register. If you take one name off this page as a September sell, take this one.
What happened to the rest of the July AL board.Drew Rasmussen, our number two at 12%, actually improved his ERA from 3.26 to 3.01, with the cleanest FIP gap in the field at −0.01 and a 0.99 projected WHIP, and still fell to 0.8%, because the innings cap we warned about is real at 170 projected and the field above him got better. Shane McClanahan regressed exactly as his +0.50 FIP gap predicted, from a 2.56 ERA to 3.25, and is off the board. Joe Ryan, at 4% in July with a 211-strikeout projection, is 6-8 with a 3.65 ERA, has not pitched since August 4, and is on the 15-day injured list. Jacob deGrom, whom the model left off its top ten in July while "betting the innings don't hold up," is 9-8 with a 3.77 ERA; the innings held and the results didn't. And Tarik Skubal, the AL's number three, was traded to the Dodgers at the deadline, so he appears in the National League section below.
The regression board
FIP minus ERA, both leagues. Right of the line, the ERA flatters; left, the pitcher has been unlucky.
Melton
+2.24
Blackburn
+1.96
E. Rodriguez
+1.17
Martinez
+0.95
Gray
+0.81
Yamamoto
+0.66
Messick
+0.57
Burns
+0.56
Misiorowski
+0.46
Schlittler
+0.45
Sánchez
+0.11
Sale
+0.04
Rasmussen
−0.01
Cease
−0.05
Skubal
−0.11
Luzardo
−0.51
deGrom
−0.66
Skenes
−0.94
Red AL, blue NL. "FIP gap" is FIP minus ERA; a large positive gap flags an ERA likely to rise. Note how the July board has resolved. Eduardo Rodriguez was +2.07 and is now +1.17, Chase Burns was +1.02 and is now +0.56, Zack Wheeler was +0.80 and is now +0.29. The gaps closed, mostly by the ERAs rising. Sonny Gray is the exception, still +0.81 with a 16-3 record. Through August 21.
National League: Milwaukee's ace kept going, and the field arrived
The NL was the tighter race in July, a 47% favorite with a 19% and a 16% behind him. It has spread out at the top and thickened underneath, because four different arms are now inside seven runs of each other for second.
The July piece projected him to lead baseball in ERA and strikeouts and called that "dominance of an order that beats volume." Thirty-four more innings later, nothing has moved: a 1.68 ERA where it was 1.62, 216 strikeouts, a 0.76 WHIP, and a 33.3% strikeout-minus-walk rate that still leads the field. The projection is 15-7, a 2.27 ERA, 267 strikeouts over 182 innings and a 0.85 WHIP, which is best in baseball in all three, with the WHIP a full tenth clear of the next-best projected mark. His +41.6 runs saved is the highest figure in either league. And he is doing it for the best team in baseball by record, since Milwaukee is 80-49 and 97% to win the Central. The only argument against him is innings, and at 182 projected he now has enough of those too.
The July entry was headlined "a case study in trusting the projection over the peripherals." His 2.82 ERA sat under a skeptical 3.64 FIP and the model refused to discount it, projecting a 2.84 finish. Since then his ERA has gone down to 2.61 and his FIP gap has narrowed from +0.82 to +0.66. He is the model's NL number two at 18.0%, projecting 15-9 with a 2.62 ERA and a 0.96 WHIP, second-best in the league behind Misiorowski's, over 193 innings. Los Angeles is 78-51 with about a 37% shot at reaching the World Series, so he will be pitching Game 1 of something in October regardless of how this vote lands.
Still the field's volume king, and now with the rate to match. He has taken a 2.83 ERA down to 2.51 while his FIP tracks it almost exactly at +0.11, and he is second in baseball in innings pitched, behind only Miami's Sandy Alcantara. The model projects 19-6 over 197 innings with 225 strikeouts, the most wins anyone in either league is projected to collect, and a 2.73 ERA. He is third on the NL runs-saved board at +34.4, which is the ceiling of the volume argument: he throws nearly the most innings in the sport and still gives back seven runs a year to Misiorowski on rate. The Phillies are 71-58 and 83% to make it, in an East that Atlanta is running away with.
In July we tempered him on a +0.63 FIP gap. That gap is now +0.04. His 2.20 ERA and 2.24 FIP are, for practical purposes, the same number, the tightest match of any National League starter on this board. He has the second-best ERA among NL qualified starters, a 25.5% strikeout-minus-walk rate, and a projected 14-11 with a 2.56 ERA and 206 strikeouts. The reason he is only at 6.6% is arithmetic rather than doubt: 168 projected innings against Misiorowski's 182 and Sánchez's 197. He is 12-9 on a 75-54 Atlanta team that is 99% to reach October, which is the sort of record that costs a pitcher votes for reasons that have nothing to do with him.
The most valuable arm on this page with no path to the trophy. Sixteen starts for Detroit and three for Los Angeles since the deadline, a 2.85 ERA with a 2.74 FIP behind it, a 25.8% strikeout-minus-walk rate, a 0.96 WHIP: a genuinely elite line. The problems are the two the July piece already named, only worse. He has missed time, so the projection stops at 154 innings. And having changed leagues in August, his season now competes in the National, where Misiorowski and Yamamoto are 17 and 12 runs ahead of him. He projects to a 2.85 full-season ERA, tenth-best in baseball among arms with 120 projected innings, at 0%. Several pitchers in this article will almost certainly finish with a worse season and more Cy Young votes.
One more arm the board can't use.Shohei Ohtani has an 8-2 record and a 1.79 ERA over 85⅔ innings, the second-lowest ERA in the National League among anyone with 80 innings, and he does not appear on the Cy Young board at all: our eligibility cut is 100 projected innings, and he has not pitched since July 3. Those innings are not wasted, though. They are the entire margin in his MVP case, worth +22.2 runs saved, which is why he leads the National League MVP board at 99.6% while his bat alone would leave him in a three-run race. The companion piece works through what it means to lead an award on a résumé you stopped adding to seven weeks ago.
The July NL cautionary tales, six weeks on.Eduardo Rodriguez, our loudest regression siren at a +2.07 FIP gap, has seen his ERA rise from 2.10 to 2.61 exactly as advertised, though a 13-4 record on a 68-61 Arizona team has kept him alive at 2.0%. Chase Burns is 14-2 with a 2.51 ERA where it was 2.50, his FIP gap halved from +1.02 to +0.56, and the model still gives him 0.4%. That is the same disagreement as Gray's, in the same shape, and Cincinnati's 61-68 record makes the vote case moot either way. Zack Wheeler, whose July snub we half-defended at 4%, is 10-4 with a 2.92 ERA and has fallen to 0% behind three teammates and a stacked field. And Paul Skenes, whom we called the field's most unlucky arm, is now its unluckiest by a wider margin, at a 3.95 ERA on a 3.01 FIP for a −0.94 gap, on a 63-67 Pittsburgh team. That is a season's worth of nothing going right.
Where they stand, and where the model lands them
Every arm named above: banked 2026 line through August 21, against the median full-season projection across 500 sims
Actual, through Aug 21
Projected full season
Pitcher
W-L
ERA
IP
K
WHIP
W-L
ERA
IP
K
WHIP
Cam SchlittlerNYY
11-6
2.16
158
193
0.96
13-8
2.53
196
234
1.02
Parker MessickCLE
9-8
2.54
149
152
1.03
11-10
2.78
181
185
1.08
Dylan CeaseTOR
7-5
2.42
138
201
1.06
9-7
2.83
175
251
1.13
Troy MeltonDET
7-2
1.60
90
76
1.00
9-4
2.48
127
108
1.09
Sonny GrayBOS
16-3
2.79
139
123
1.16
18-5
3.02
175
159
1.22
Drew RasmussenTB
13-5
3.01
134
138
0.95
15-7
2.96
170
173
0.99
Nick MartinezTB
12-4
3.05
142
83
1.14
14-7
3.12
180
109
1.18
Shane McClanahanTB
9-6
3.25
102
94
1.12
10-7
3.60
129
119
1.17
Logan GilbertSEA
10-7
3.35
150
157
1.01
12-9
3.20
183
189
1.05
Jacob deGromTEX
9-8
3.77
127
157
1.13
11-9
3.47
157
190
1.14
Max FriedNYY
4-4
2.81
86
79
1.00
6-6
2.74
118
110
1.06
Paul BlackburnNYY
4-2
1.77
71
58
1.06
6-3
2.32
103
85
1.12
Joe Ryan 15-day ILMIN
6-8
3.65
126
138
1.14
no further innings projected
Jacob MisiorowskiMIL
13-5
1.68
145
216
0.76
15-7
2.27
182
267
0.85
Yoshinobu YamamotoLAD
12-7
2.61
152
145
0.90
15-9
2.62
193
187
0.96
Cristopher SánchezPHI
16-4
2.51
162
187
1.22
19-6
2.73
197
225
1.23
Chris SaleATL
12-9
2.20
135
166
1.03
14-11
2.56
168
206
1.08
Tarik SkubalLAD
7-7
2.85
114
135
0.96
10-8
2.85
154
188
0.99
Eduardo RodriguezARI
13-4
2.61
155
110
1.21
15-6
2.91
193
143
1.24
Chase BurnsCIN
14-2
2.51
136
159
1.11
15-4
2.84
166
194
1.17
Zack WheelerPHI
10-4
2.92
120
144
1.01
12-6
2.94
152
181
1.06
Jesús LuzardoPHI
11-5
3.35
156
194
1.15
14-7
3.37
196
244
1.18
Paul SkenesPIT
9-11
3.95
139
171
1.14
11-13
3.99
170
206
1.17
Shohei Ohtani not since Jul 3LAD
8-2
1.79
86
95
0.95
no further innings projected
Team colored red for AL, blue for NL. Projected ERA is converted from projected runs allowed at this season's league-wide earned-run share, so read it as a close estimate rather than an exact figure. Joe Ryan and Shohei Ohtani are projected no further innings, so their season totals are already final. BBMI season simulator, August 22 run.
Today against the finish line
ERA through August 21 (dark dot) against projected full-season ERA (colored dot)
1.502.002.503.003.504.10
American League
Melton
1.60 to 2.48
Schlittler
2.16 to 2.53
Cease
2.42 to 2.83
Messick
2.54 to 2.78
Gray
2.79 to 3.02
Rasmussen
3.01 to 2.96
National League
Misiorowski
1.68 to 2.27
Sale
2.20 to 2.56
Sánchez
2.51 to 2.73
Burns
2.51 to 2.84
E. Rodriguez
2.61 to 2.91
Yamamoto
2.61 to 2.62
Skubal
2.85 to 2.85
Skenes
3.95 to 3.99
With two-thirds of the sample already banked, an August ERA is hard to improve, and the model expects most of this board to give a little back down the stretch. There are two exceptions. Rasmussen, in green, is the only arm here projected to lower his ERA. Skubal, in grey, is projected to finish exactly where he sits. Median projection across 500 sims.
The bottom line
Both favorites are stronger than they were, and both got there by not regressing. That is the whole finding. When we projected Schlittler at a 2.74 ERA and Misiorowski at 2.57 in July, we were building in the give-back that August ERAs usually deliver. Instead Schlittler moved 11 points and Misiorowski six, and the projections came down to meet them rather than the other way around.
Where the model still disagrees with the room is the same place it disagreed in July, and it has not been vindicated. Sonny Gray is 16-3 and Chase Burns is 14-2, and our board has them at 3.2% and 0.4% because their peripherals say a good pitcher is being carried by his defense and his offense. Six weeks was not enough to settle that, and the last five weeks probably won't be either. The difference between "the ERA will crack" and "the ERA will hold and the votes will follow the record" is the honest open question on this page.
Watch the innings from here. Seven arms are projected past 190 and three of them are on this page: Sánchez at 197, Schlittler at 196, Yamamoto at 193. Misiorowski, at 182, carries the lightest workload of any leader, and in a runs-saved award that is the one place his lead is exposed.
Projected strikeout leaders
Median full-season strikeouts across 500 sims
Misiorowski
267
Cease
251
Luzardo
244
G. Williams
239
Schlittler
234
Sánchez
225
Detmers
214
Sale
206
Red bars AL, blue bars NL. Misiorowski's 267 would lead baseball, and 80% of his sims land between 259 and 276. Paul Skenes also projects to 206.
Methodology: ERA and stat lines recomputed from official box-score data through August 21. Cy Young probabilities and projections are BBMI season-simulator outputs from the August 22 run: 500 replays of the 481 remaining games, with each projected line the median across runs and ranges the 10th–90th percentile. Cy Young ranks on runs saved, meaning a pitcher's projected run prevention measured against the league average of 4.54 runs per nine, scaled by projected innings. Projected ERA is converted from projected runs allowed at this season's league-wide earned-run share, where 91.8% of runs are earned, because the simulator does not model errors at the individual level; ERA figures should be read as close estimates rather than exact. A pitcher's league is the league of the club he is on now, which is why Tarik Skubal's season appears in the National League. Pitchers who are done for the year keep their banked line and are projected no further innings. "FIP gap" is FIP minus ERA, on a league-derived FIP constant of 3.085; a large positive gap flags an ERA likely to rise. The odds are the share of 500 simulated finishes in which a pitcher leads his league in runs saved. They are not a forecast of the writers' vote. Team records and playoff probabilities are from the August 22 BBMI playoff simulator, 10,000 runs. For the hitting companion, see the 2026 MVP races. Not betting advice.