Chicago White Sox at Houston Astros

2026-09-03, 2:10 PM ET, Daikin Park. Luis Castillo (R) against Hunter Brown (R). Every number derives from the public Statcast store and the engine's published boards. Charts are code-drawn. Probabilities, not picks. Statcast through 2026-09-02.

What decides tonight

Every number below is a probability this engine published before first pitch and grades that night, or a plain count. No scouting verdicts are ranked here — those live in the panels and carry their own sample sizes.

Hunter Brown (RHP)

1401 pitches, K% 27% vs LHB (194 BF) / 23% vs RHB (148 BF)

15 apps · median 23 BF · reaches the order 3.0× (max 3) · exits at 92 pitches (IQR 90–102, 15 starts) · fastball -0.4 mph late (13 deep starts). Fade is measured inside each start on nights that reached 85 pitches, so it is not flattered by the short outings.

League percentiles, starter pool

K%7025%
BB%913%
CSW%7329%
Whiff%5824%
Chase%1226%
xwOBA vs74.293
Hard-hit%5437%
GB%9555%
1st-pitch strike%1858%
published K forecast: E[K] 5.5
recent starts, K (n=5)
808/05
TOR
408/11
SF
508/16
SEA
408/22
ATH
1008/28
NYM

Arsenal, shape and results

Pitch # vL%vR% MPHIVBHB RPMWhiffCSW Chase 2K use 2K zone xwOBA RV/100
4-seam45439%23%96.2+17-8224619%26%20%31%37%.244-0.5
sinker41620%42%96.0+11-15214616%32%30%22%62%.305-1.2
knuckle-curve27622%17%83.4-16+6255535%29%31%29%28%.258+0.3
changeup13415%2%87.6+5-13147840%29%27%6%22%.431+2.0
slider1213%16%90.7+5+4226024%23%26%12%47%.329+1.0

RV/100 is run value per 100 pitches from the pitcher's side, so negative is good (a curveball at −1.3 costs the pitcher runs even if it misses bats — a real tension worth reading, not a contradiction). Whiff, chase and xwOBA are computed on this pitcher's own pitches this season; low-usage pitches carry small samples.

Locations, catcher view

4-seam
sinker
knuckle-curve

Movement

4-seam: +17 IVB, -8 HBsinker: +11 IVB, -15 HBknuckle-curve: -16 IVB, +6 HBchangeup: +5 IVB, -13 HBslider: +5 IVB, +4 HBhorizontal break (in), catcher viewinduced vertical break (in)

How to attack

  • Hunt the 4-seam early: 34% of his first pitches, only 19% whiffs, 44% in-zone on 0-0.
  • Lay off the changeup below the zone: 40% whiffs, only 37% of them are strikes.
  • Vs righties he leans sinker (42% of pitches to that side).

Luis Castillo (RHP)

2197 pitches, K% 17% vs LHB (319 BF) / 23% vs RHB (228 BF)

25 apps · median 22 BF · reaches the order 3.0× (max 3) · exits at 93 pitches (IQR 84–100, 22 starts) · fastball -0.2 mph late (15 deep starts). Fade is measured inside each start on nights that reached 85 pitches, so it is not flattered by the short outings.

League percentiles, starter pool

K%3019%
BB%757%
CSW%1425%
Whiff%4222%
Chase%7132%
xwOBA vs18.347
Hard-hit%2143%
GB%2537%
1st-pitch strike%9770%
published K forecast: E[K] 4.0
recent starts, K (n=5)
508/06
BOS
1008/12
CIN
108/17
CHC
308/22
NYM
108/28
MIN

Arsenal, shape and results

Pitch # vL%vR% MPHIVBHB RPMWhiffCSW Chase 2K use 2K zone xwOBA RV/100
4-seam92845%39%95.3+12-12227124%23%34%47%39%.349+0.9
slider54823%28%85.6+2+1223926%30%32%26%34%.321+1.2
sinker39711%28%95.5+4-1820799%24%24%12%59%.412-0.5
changeup32121%6%87.5-1-17196623%19%34%15%39%.293+1.5

RV/100 is run value per 100 pitches from the pitcher's side, so negative is good (a curveball at −1.3 costs the pitcher runs even if it misses bats — a real tension worth reading, not a contradiction). Whiff, chase and xwOBA are computed on this pitcher's own pitches this season; low-usage pitches carry small samples.

Locations and movement (4 charts)
4-seam
slider
sinker

Movement

4-seam: +12 IVB, -12 HBslider: +2 IVB, +1 HBsinker: +4 IVB, -18 HBchangeup: -1 IVB, -17 HBhorizontal break (in), catcher viewinduced vertical break (in)

Attack plan vs Houston Astros

  • Slider: top breaking whiffs are Trammell (50%), Peña (44%).
  • Changeup: top offspeed whiffs are Smith (44%), Peña (39%).
  • Careful with Paredes: 85% contact, 12% K rate vs RHP; pitch to contact at the edges.

White Sox bats vs Hunter Brown

Season quality vs tonight's starter hand, percentile-colored vs league: red elite, paper average, blue poor. P columns are the engine's published board numbers for this game. Batting order as posted. 30d contact is the last-30-day contact rate minus the season rate; an arrow is drawn only when the move clears two standard errors of its own noise, and n is the recent swing count.

#BatterPAxwOBAK%Chase%Contact%HH%P(1+ hit)P(1+ HR)30d contact
1Sam Antonacci (L)381.36014%21%87%36%63%5%n174
2Munetaka Murakami (L)316.37433%23%61%56%58%15%n215
3Miguel Vargas (R)424.37118%21%82%41%63%11%n204
4Andrew Benintendi (L)371.31922%33%75%46%59%9%n147
5Braden Montgomery (S)210.34817%26%76%42%61%9%n220
6Colson Montgomery (L)386.29637%33%64%45%60%11%n233
7Chase Meidroth (R)401.28723%28%82%33%67%9%n185
8Randal Grichuk (R)73.28821%30%71%43%··n95
9Jake Rogers (R)119.25729%24%78%33%53%9%n78

Houston Astros bats, where our staff attacks

Whiff by pitch group vs our starter's hand. Blue cells are the attack zones. P columns are the published board numbers. Batting order as posted. 30d contact is the last-30-day contact rate minus the season rate; an arrow is drawn only when the move clears two standard errors of its own noise, and n is the recent swing count.

#BatterPAFB whiffBRK whiffOFF whiffxwOBAChase%Contact%1st-pitch swingP(1+ hit)P(1+ HR)30d contact
1Jeremy Peña (R)30413%44%39%.32638%72%47% n30565%11%▼ -6% n213
2Yordan Alvarez (L)44613%24%16%.48227%84%30% n44668%14%n188
3Isaac Paredes (R)4349%23%22%.32026%85%16% n43468%8%n212
4Nick Allen (R)10920%34%·.23726%77%31% n109··n60
5Daulton Varsho (L)35218%22%25%.29436%80%35% n35263%11%n204
6Christian Walker (R)42620%38%29%.29632%73%35% n42663%9%n184
7Taylor Trammell (L)20030%50%38%.31528%64%34% n20054%9%n122
8Yainer Diaz (R)20216%21%25%.32743%82%44% n20261%10%n133
9Cam Smith (R)36417%43%44%.32129%73%35% n365··n175

Houston Astros hitters, one card each

One card per hitter: the published probabilities this engine grades tonight, his whiff shape by pitch group, and his season against this starter's hand with the sample stated. Times-through-order and leverage splits are deliberately absent — at a median 45 plate appearances the third-time-through number carries a standard error roughly three times the effect it would be claiming.

1

Jeremy Peña (R)

P(1+ hit) 65% · P(1+ HR) 11%

FB13%
BRK44%
OFF39%

vs RHP 0.326 xwOBA, K 21% (304 PA)

2

Yordan Alvarez (L)

P(1+ hit) 68% · P(1+ HR) 14%

FB13%
BRK24%
OFF16%

vs RHP 0.482 xwOBA, K 18% (446 PA)

3

Isaac Paredes (R)

P(1+ hit) 68% · P(1+ HR) 8%

FB9%
BRK23%
OFF22%

vs RHP 0.320 xwOBA, K 12% (434 PA)

4

Nick Allen (R)

FB20%
BRK34%
OFF·

vs RHP 0.237 xwOBA, K 23% (109 PA)

5

Daulton Varsho (L)

P(1+ hit) 63% · P(1+ HR) 11%

FB18%
BRK22%
OFF25%

vs RHP 0.294 xwOBA, K 20% (352 PA)

6

Christian Walker (R)

P(1+ hit) 63% · P(1+ HR) 9%

FB20%
BRK38%
OFF29%

vs RHP 0.296 xwOBA, K 25% (426 PA)

7

Taylor Trammell (L)

P(1+ hit) 54% · P(1+ HR) 9%

FB30%
BRK50%
OFF38%

vs RHP 0.315 xwOBA, K 32% (200 PA)

8

Yainer Diaz (R)

P(1+ hit) 61% · P(1+ HR) 10%

FB16%
BRK21%
OFF25%

vs RHP 0.327 xwOBA, K 13% (202 PA)

9

Cam Smith (R)

FB17%
BRK43%
OFF44%

vs RHP 0.321 xwOBA, K 25% (364 PA)

Matchups: CWS hitters vs Hunter Brown

Where this starter's actual usage meets each hitter's measured holes and damage, on the same map. An edge is only reported where the pitcher genuinely goes — a hole against a pitch he does not throw is not a matchup — and only where the rate is measurably worse than league for a typical hitter in that same spot. Every line carries its sample size. Batter-versus-pitcher history is shown as trivia and is not used in any of it: the samples that actually occur are far too small to support a conclusion, and most public matchup content is built on them anyway.

1. Sam Antonacci

Edge: batterWhen the ball goes up and away — and he puts 8% of his pitches there — the batter does real damage: 0.426 expected value per ball in play, against a 0.272 league average in that spot. (19 balls in play.)

Tonight's published, gradeable numbers for this batter: 63% to get a hit and 5% to homer. Those are what the record grades — the notes above only explain why.

0 for 3 lifetime. At 3 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

2. Munetaka Murakami

Edge: pitcherHe throws the four-seam off the plate on 21% of his pitches to this side. The batter swings and misses 46% of the time there, against 31% for a typical hitter — +15 points. (31 swings.)

Edge: batterWhen the ball goes down and away — and he puts 6% of his pitches there — the batter does real damage: 0.406 expected value per ball in play, against a 0.261 league average in that spot. (20 balls in play.)

Tonight's published, gradeable numbers for this batter: 58% to get a hit and 15% to homer. Those are what the record grades — the notes above only explain why.

1 for 3 lifetime. At 3 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

3. Miguel Vargas

Edge: batterWhen the ball goes middle and in — and he puts 5% of his pitches there — the batter does real damage: 0.481 expected value per ball in play, against a 0.332 league average in that spot. (31 balls in play.)

Tonight's published, gradeable numbers for this batter: 63% to get a hit and 11% to homer. Those are what the record grades — the notes above only explain why.

0 for 3 lifetime. At 3 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

4. Andrew Benintendi

Edge: pitcherHe throws the four-seam up-middle on 3% of his pitches to this side. The batter swings and misses 23% of the time there, against 17% for a typical hitter — +5 points. (31 swings.)

Edge: batterWhen the ball goes up-middle — and he puts 6% of his pitches there — the batter does real damage: 0.400 expected value per ball in play, against a 0.331 league average in that spot. (22 balls in play.)

Tonight's published, gradeable numbers for this batter: 59% to get a hit and 9% to homer. Those are what the record grades — the notes above only explain why.

0 for 3 lifetime. At 3 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

5. Braden Montgomery

Edge: batterWhen the ball goes down-middle — and he puts 6% of his pitches there — the batter does real damage: 0.395 expected value per ball in play, against a 0.357 league average in that spot. (22 balls in play.)

Tonight's published, gradeable numbers for this batter: 61% to get a hit and 9% to homer. Those are what the record grades — the notes above only explain why.

0 for 2 lifetime. At 2 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

6. Colson Montgomery

Edge: pitcherHe throws the four-seam off the plate on 21% of his pitches to this side. The batter swings and misses 48% of the time there, against 31% for a typical hitter — +17 points. (87 swings.)

Edge: batterWhen the ball goes down-middle — and he puts 6% of his pitches there — the batter does real damage: 0.510 expected value per ball in play, against a 0.357 league average in that spot. (18 balls in play.)

Tonight's published, gradeable numbers for this batter: 60% to get a hit and 11% to homer. Those are what the record grades — the notes above only explain why.

0 for 3 lifetime. At 3 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

7. Chase Meidroth

Edge: batterWhen the ball goes middle and in — and he puts 5% of his pitches there — the batter does real damage: 0.409 expected value per ball in play, against a 0.332 league average in that spot. (23 balls in play.)

Tonight's published, gradeable numbers for this batter: 67% to get a hit and 9% to homer. Those are what the record grades — the notes above only explain why.

0 for 2 lifetime. At 2 plate appearances this is trivia, not evidence — it is shown because people ask for it, and it is not used anywhere in the analysis above.

8. Randal Grichuk

Not enough of a track record in the spots this pitcher uses to say anything useful. No verdict rather than a guess.

No published probability for this batter tonight, so the notes above are observations, not predictions.

9. Jake Rogers

Not enough of a track record in the spots this pitcher uses to say anything useful. No verdict rather than a guess.

Tonight's published, gradeable numbers for this batter: 53% to get a hit and 9% to homer. Those are what the record grades — the notes above only explain why.

Houston Astros: the running game and the defense

Sprint speed says who can run; attempt rate says who will. Rate is steals of second per time on first, shrunk toward the league’s 5.3% at 22 opportunities, with the raw count beside it. Verdict needs 15+ times on first. Sprint speed and Outs Above Average are Baseball Savant leaderboard metrics — measured from tracking data we do not have and did not compute.

Behind the plate: Jake Rogers — runners went 39-for-51 against him (24% caught). That is mostly the pitcher: measured split-half reliability for a catcher’s own caught-stealing rate is +0.59, and the spread between catchers is smaller than the sampling noise. Read it as what happened, not as a rating.

#Batter Sprint Steal rate Threat OAA
1Jeremy Peña (R)29.2 ft/s4.8% (16/338)holds-3 SS
2Yordan Alvarez (L)24.9 ft/s1.1% (3/346)holds·
3Isaac Paredes (R)25.0 ft/s0.5% (1/390)holds+0 3B
4Nick Allen (R)26.9 ft/s7.0% (14/193)occasional+2 SS
5Daulton Varsho (L)27.3 ft/s6.3% (13/202)holds+4 CF
6Christian Walker (R)25.9 ft/s0.8% (2/364)holds+2 1B
7Taylor Trammell (L)27.0 ft/s7.0% (8/109)occasional·
8Yainer Diaz (R)25.8 ft/s1.1% (2/269)holds·
9Cam Smith (R)29.2 ft/s6.2% (18/285)holds+5 RF

Chicago White Sox: the running game and the defense

Sprint speed says who can run; attempt rate says who will. Rate is steals of second per time on first, shrunk toward the league’s 5.3% at 22 opportunities, with the raw count beside it. Verdict needs 15+ times on first. Sprint speed and Outs Above Average are Baseball Savant leaderboard metrics — measured from tracking data we do not have and did not compute.

Behind the plate: Yainer Diaz — runners went 108-for-133 against him (19% caught). That is mostly the pitcher: measured split-half reliability for a catcher’s own caught-stealing rate is +0.59, and the spread between catchers is smaller than the sampling noise. Read it as what happened, not as a rating.

#Batter Sprint Steal rate Threat OAA
1Sam Antonacci (L)28.2 ft/s8.2% (17/200)occasional-5 LF
2Munetaka Murakami (L)27.1 ft/s2.0% (3/182)holds+0 1B
3Miguel Vargas (R)28.3 ft/s3.9% (15/389)holds-2 3B
4Andrew Benintendi (L)26.6 ft/s0.8% (1/268)holds·
5Braden Montgomery (S)28.5 ft/s1.7% (1/103)holds·
6Colson Montgomery (L)26.6 ft/s0.9% (1/227)holds+6 SS
7Chase Meidroth (R)27.2 ft/s3.1% (13/428)holds+4 2B
8Randal Grichuk (R)27.3 ft/s1.7% (1/110)holds·
9Jake Rogers (R)26.3 ft/s2.0% (1/86)holds·

Bullpens

Sorted by 30-day workload. Day cells show pitches on the three most recent league game dates; dark means heavy recent use. Season rates percentile-colored vs all pitchers with 150+ pitches.

White Sox bullpen 8 arms
Reliever08/3109/0109/02 games since 30-day K%Whiff%xwOBA Top pitches
Trevor Richards (R)38··2 heavy305p/10g22%21%.3084-seam 91 (66%) · changeup 80 (21%)
Erick Fedde (R)···9 heavy260p/6g16%18%.334sweeper 82 (37%) · sinker 94 (33%)
Grant Taylor (R)·14·1238p/12g33%26%.2504-seam 98 (40%) · curveball 84 (24%)
Sean Newcomb (L)··41today heavy218p/8g23%21%.3104-seam 95 (38%) · slurve 83 (30%)
Hagen Smith (L)···5209p/8g35%35%.2804-seam 94 (50%) · slider 83 (30%)
Jordan Hicks (R)·25·1175p/9g27%25%.282sinker 100 (39%) · 4-seam 99 (29%)
Huascar Brazobán (R)···8166p/10g21%24%.295sinker 96 (54%) · changeup 91 (34%)
Bryan Hudson (L)·17·1143p/11g22%22%.2804-seam 92 (72%) · sweeper 79 (28%)
Houston Astros bullpen 8 arms
Reliever08/3109/0109/02 games since 30-day K%Whiff%xwOBA Top pitches
AJ Blubaugh (R)··15today229p/11g25%26%.3074-seam 96 (51%) · sweeper 85 (27%)
Enyel De Los Santos (R)27··2227p/11g22%25%.2904-seam 95 (40%) · changeup 85 (30%)
Bryan Abreu (R)16··2225p/12g30%32%.3194-seam 96 (46%) · slider 87 (39%)
Josh Hader (L)··14today191p/13g36%33%.213sinker 95 (63%) · slider 84 (37%)
Bennett Sousa (L)23··2191p/12g22%30%.317slider 86 (72%) · 4-seam 94 (19%)
Steven Okert (L)18··2189p/13g26%29%.276slider 80 (64%) · 4-seam 90 (29%)
Bryan King (L)···12 heavy109p/5g18%16%.3234-seam 92 (47%) · sinker 92 (31%)
Kai-Wei Teng (R)·35·1 heavy87p/4g22%25%.339sweeper 84 (39%) · sinker 93 (29%)