Player development / Hitting Lab

Is it a swing-decision problem or a bat problem?

Pittsburgh hit .701 OPS against left-handed pitching (17th) and .749 against righties (6th). Against lefties that was about −12 runs (≈ −1.2 wins) against an average lineup: −6 from swing decisions, −7 from contact, and +1 from the lineup’s handedness (30% of the pitches lefties threw went to left-handed hitters, against 34% for the league). Keystone splits each hitter into swing decisions and contact because each needs a different fix. The Lineup Lab checks how much of the gap the 2027 platoons close.

For hitting coordinator, MLB hitting coach, manager Model card 2026 regular season · Statcast pitch & bat tracking

Swing decisions vs LHP −6.5 runs decision value −0.17 per 100 pitches, about 0.49 runs each · 22nd of 30 · 7,774 pitches

Contact vs LHP −6.5 runs −0.17 per 100 swings · 19th of 30

Handedness mix vs LHP +1.1 runs vs. an average lineup · 30% of pitches to left-handed hitters (league 34%) · 11th of 30

Swing decisions vs RHP +8.2 runs decision value +0.10 per 100 pitches · 13th of 30 · contact −4.0 runs (18th)

Where the 2027 lineup loses runs to lefties

The Winter Plan’s hitters (decisions feed), each one’s regressed skill against left-handers over the pitches he should see from them in 2027. Against same-side peers: what coaching can change.

The plan’s chain (if Lowe declines the qualifying offer), at the plate, handedness included: Nick Gonzales takes second from Brandon Lowe (+0.5 runs against left-handers, −21.6 against right-handers over 535 PA), and Isaac Paredes plays third instead of the in-house option, Jared Triolo (+3.2 and +12.9 over 562 PA). Together: +3.7 runs against left-handers and −8.6 against right-handers. Offense only; the Winter Plan adds defense, and the Lineup Lab sets the order.

Decisions × contact · vs LHP

Regressed estimates (2025–26 pooled, the split against one hand shrunk on its own) with ±1 SD. Skill against same-side peers. Circle size = plate appearances behind the estimate (min 100). Dashed: a tentative tag.

Jake Mangum

CF · bats S · vs LHP · 2027: Bench · 476 PA

Limit exposure · tentative

Below the line on both halves against both hands, even after regressing: a defense or bench bat, not a platoon.

57% chance both halves are really below the line (−0.55): under 70%, so treat the tag as a lead, not a plan.

Decisions est. ± SD · 2026 raw: −1.45 ± 0.30 · −1.92 value/100 pitches
Contact est. ± SD · 2026 raw: −0.67 ± 0.72 · −0.88 /100 swings
With the matchup: −1.09 runs/100 pitches, handedness included
PA 2026 · pooled: 134 · 229
wOBA / xwOBA: .313 / .289
Chase: 46.4%
Hittable pitches taken: 25.1%
Zone contact: 86.7%
Bat speed: 68.5 mph
Squared-up / swing: 20.8%
EV90 · Barrel: 101.8 · 4.2%

Full profile

Pittsburgh hitters · vs LHP

Sorted by 2026 plate appearances. Decisions and contact are regressed estimates ± SD (2026 raw in the detail panel); a diagnosis needs 100 pooled PA. 2027 roles from the Winter Plan.

Hitter	2027	PA	xwOBA	Decisions	Contact	Chase	Takes	Z-Contact	EV90	Diagnosis
Bryan Reynolds LF · S	LF · 602 PA	240	.345	+0.96 ±0.3	+0.69 ±0.7	27%	23%	83%	106.7	Core
Brandon Lowe 2B · L	Leaving if he declines the QO	211	.266	−0.33 ±0.3	+0.60 ±0.7	41%	29%	78%	104.7	Core
Nick Gonzales 3B · R	2B · 535 PA	195	.299	+0.29 ±0.3	−0.56 ±0.7	31%	18%	83%	100.8	Contact fix ?
Jared Triolo SS · R	Bench · 415 PA	143	.330	+0.44 ±0.3	−0.94 ±0.8	26%	33%	89%	102.5	Contact fix ?
Ryan O'Hearn RF · L	1B · 474 PA	140	.280	−0.26 ±0.3	+0.11 ±0.7	33%	41%	86%	102.1	Core ?
Oneil Cruz CF · L	CF · 458 PA	138	.315	−0.76 ±0.3	+0.70 ±0.8	30%	43%	81%	111.2	Approach fix
Jake Mangum CF · S	Bench · 476 PA	134	.289	−1.45 ±0.3	−0.67 ±0.7	46%	25%	87%	101.8	Limit exposure ?
Esmerlyn Valdez RF · R	RF · 344 PA	111	.370	+1.36 ±0.4	−0.06 ±0.9	21%	29%	66%	108.6	Core
Henry Davis C · R	Release by Nov 17	92	.235	+0.43 ±0.3	−1.26 ±0.8	38%	14%	85%	104.5	Contact fix
Konnor Griffin SS · R	SS · 369 PA	91	.359	−0.87 ±0.4	+0.57 ±0.9	37%	33%	82%	107.8	Small sample
Spencer Horwitz 1B · L	DH · 461 PA	90	.294	+0.58 ±0.3	−0.70 ±0.8	24%	35%	88%	101.2	Contact fix ?
Rafael Flores Jr. C · R	C · 266 PA	77	.342	+0.08 ±0.4	+0.97 ±1.0	27%	36%	78%	105.5	Small sample
Nick Yorke 3B · R	Call-up · 10 PA	61	.270	−0.18 ±0.4	−0.79 ±1.0	33%	29%	78%	103.8	Small sample
Jacob Gonzalez SS · L	Traded to HOU	33	.295	−0.37 ±0.4	+0.03 ±1.0	30%	19%	80%	106.1	Small sample
Endy Rodríguez C · S	Bench · 289 PA	28	.298	+0.41 ±0.4	−0.57 ±1.0	31%	38%	79%	102.9	Small sample
Ronny Simon RF · S	Bench · 78 PA	24	.382	−0.32 ±0.5	+0.01 ±1.1	48%	20%	97%	104.8	Small sample
Tyler Callihan LF · L	Minors	7	.365	+0.45 ±0.5	−1.53 ±1.1	22%	18%	80%	92.0	Small sample

? = tentative: under a 70% chance the true value is past the line (for Limit exposure, that both halves are). The line (−0.55) is about the bottom fifth of regressed values against left-handers (−0.60 for decisions, −0.60 for contact among 240 hitters).

Who hits left-handed pitching · outside the org

100+ PA vs LHP in 2026, ranked by estimated runs per 100 pitches against lefties with the matchup added back (a left-handed hitter facing a lefty starts about 0.41 runs per 100 pitches behind a right-hander; decision value counts at 0.49 runs a unit). Players their clubs are unlikely to trade are left out (22 in this range).

Hitter	Team	PA vs L	vs LHP	Skill	xwOBA	2027 WAR	Age	Available?	2027 cost
Isaac Paredes 3B · R · the plan’s acquisition	HOU	158	+0.29 81st	+0.27 100th	.328	3.3 plan	28	At a price	Club option $13.3M
Seiya Suzuki RF · R	CHC	165	+1.25 4th	+1.23 7th	.389	2.6	32	Free agent	4 yr × $31.3M est.
Gleyber Torres 2B · R	DET	124	+1.17 6th	+1.15 8th	.384	1.1	30	Free agent	2 yr × $7.5M est.
Heriberto Hernández LF · R	MIA	163	+0.95 12th	+0.93 15th	.400	1.6	27	At a price	Pre-arb $780K
Jo Adell RF · R	CLE	190	+0.83 19th	+0.80 26th	.399	0.7	28	At a price	Arb $5.2M
Yandy Díaz DH · R	TB	197	+0.79 23rd	+0.77 32nd	.351	1.1	35	At a price	Vesting option $13.0M
George Springer DH · R	TOR	159	+0.75 25th	+0.73 35th	.286	0.7	37	Free agent	1 yr × $6.3M est.
Paul Goldschmidt 1B · R	NYY	181	+0.74 27th	+0.71 37th	.371	0.0	39	Free agent	1 yr × $2.3M est.
Taylor Ward DH · R	SEA	192	+0.69 29th	+0.67 44th	.351	1.2	33	Free agent	2 yr × $9.2M est.
Jahmai Jones DH · R	BOS	156	+0.69 30th	+0.67 45th	.353	0.3	29	At a price	Arb $1.1M
Andrew Vaughn 1B · R	MIL	132	+0.69 31st	+0.66 46th	.349	0.6	29	At a price	Arb $7.7M
Jake Bauers 1B · L	MIL	160	+0.65 35th	+1.03 11th	.362	1.7	31	Free agent	2 yr × $17.7M est.
Ty France 1B · R	SD	138	+0.60 37th	+0.58 55th	.337	1.0	32	Free agent	2 yr × $10.8M est.

“vs LHP” adds back the league’s value for his batting side against lefties; “Skill” leaves it out, which is right for coaching but not for choosing who faces a lefty. The Lineup Lab asks the same question against a named 2027 incumbent and in runs per season, and the FA Board and Opportunity Finder price the fit.

Does it hold up?

2024–25 estimates against what the same hitters did in 2026

Swing decisions, 247 hitters with 300+ PA: r = 0.77 for the estimate, 0.73 for the raw 2025 value; chase rate alone −0.73. Error 0.54 vs. 0.63 decision value per 100 pitches.
Contact: r = 0.63 vs. 0.57 raw; zone contact rate alone −0.06. Error 1.02 vs. 1.30; slope 0.81 (estimates still a little too spread out).
Against left-handers, 231 hitters with 100+ PA: Error with the split 0.791 vs. 0.816 using his overall estimate alone (decisions); 1.396 vs. 1.411 (contact).

Read it this way: the estimates beat last season’s raw numbers, and decision value carries about the same signal as chase rate alone. The split against one hand cuts the error by 3% (decisions) and 1% (contact) over the hitter’s overall estimate, so a diagnosis against lefties is mostly his overall diagnosis. The swing models are cross-fit by season, so the 2024–25 values were scored by models trained on the other seasons, 2026 included. The models don’t know who is batting, so that leak is league-level only.