The Astros must make the staff earn the outs

The decision

Houston scored 48 more runs in 2026 and allowed 101 more. That is the offseason in two numbers. Yordan Alvarez supplied an extraordinary offensive season, but a pitching staff that walked 650 batters kept giving opponents additional chances. The Astros need a more dependable route from the first pitch to the final out before another expensive hitter becomes the priority.

The preferred plan signs Nick Martinez for two years and $26-30 million and retains free agent Steven Okert for two years and $16-20 million. With a separate $6-8 million depth reserve, annual additions are $27-33 million. Martinez receives 30 starts and 165 innings; Okert receives 60 relief innings. Those are S&S negotiating ceilings and planning workloads. Neither agreement exists.

Plan B replaces Martinez with Sonny Gray at one year and $20-24 million. Annual additions rise to $34-42 million, while the shorter starter contract lowers total new commitment. Plan C buys neither pitcher and redistributes every inning internally. It is a complete roster scenario, but it exposes more young pitchers and uncertain recoveries to meaningful workloads.

The corrected existing-roster envelope is $229.18-233.18 million before additions, benefits or tax. Plan A reaches $256.18-266.18 million. The report assigns 6,200 plate appearances, 162 starts at each position and 1,440 pitching innings. Houston can spend less than the preferred plan, but the savings have specific consequences for who must pitch and how often.

A division winner with a narrowing margin

The Astros won 88 games in 2024, 87 in 2025 and 81 in 2026. Their run margin deteriorated from plus-91 to plus-21 to minus-32. The 2026 offense recovered from the previous season, raising team OPS from .714 to .732. Run prevention moved in the other direction, with team ERA increasing from 3.86 to 4.45.

The postseason supplied little insulation from that trend. Chicago won the two Wild Card games in Houston: White Sox 6, Astros 3 on September 29 and White Sox 7, Astros 3 on September 30. Those are postseason results, kept separate from the regular-season tables and workload calculations. A two-game series cannot explain a whole roster, but it ended the season before Houston could resolve its deeper pitching problem.

The relevant comparison is not whether one more timely hit could have changed a playoff game. It is whether the retained pitchers can reduce the sustained pressure created by walks and short starts. In 2026, starters supplied 774 2/3 innings with a 4.81 ERA. Relievers carried 662 1/3 innings with a 4.08 ERA. The proposed staff shifts more work into the rotation without assuming every recovering starter returns at full strength.

Start with the actual roster

The dated official roster contains 39 names. November moves include the additions of Roddery Munoz and Cesar Salazar and the acquisition of Evan Reifert. The current team transaction account also resolves two incomplete API records: Jake Meyers elected free agency after being outrighted, and Ryan Weiss became a free agent after his option was declined. Neither is retained in the workload plan.

Spencer Arrighetti was traded to Toronto during the season. Colton Gordon also moved to Milwaukee. Their Houston and Sugar Land innings remain part of the historical performance evidence, but they are unavailable as internal 2027 depth. Jason Alexander is a free agent. Dustin Harris and Daniel Johnson likewise cannot be carried forward merely because they appear in Houston's 2026 batting totals.

Okert is a proposed retention purchase, not a pitcher already under team control. The same distinction applies to Daulton Varsho and the other departing veterans considered later. Current organizational affiliation, an old team stat line and an executable contract are different things. Any signing or minor-league selection must fit the actual roster when it occurs; the 39-name snapshot is not a promise that a place will remain open.

Alvarez is the foundation

Alvarez produced 42 home runs, 109 walks and a 1.033 OPS in 692 plate appearances. His .429 wOBA and .442 expected wOBA support the conclusion that this was elite offensive production, rather than a season built on fortunate placement of ordinary contact. The expected slugging figure was also above his actual result.

The planning challenge is that Houston cannot safely require another 692 plate appearances from its best hitter. Alvarez had only 199 in 2025. The preferred roster assigns 650 and protects much of his workload at designated hitter. It does not use his left-field availability as the solution to every other player's playing-time problem.

Isaac Paredes and Jeremy Pena give the lineup additional substance. Paredes combined 25 home runs with 71 walks and only 88 strikeouts. Pena hit 22 home runs with an .802 OPS in 483 plate appearances. Both provide a better starting point than a plan that depends on Walker and Altuve returning to their strongest previous seasons.

The older core needs realistic jobs

Altuve's OPS fell from .789 in 2024 to .771 in 2025 and .701 in 2026. Walker moved from .803 to .718 to .726. These are full MLB seasons, not selectively chosen stretches. Walker still hit 29 home runs in 2026, but the power did not produce an all-around offensive season comparable to the best hitters on this roster.

The response should be a defined workload rather than a demand that the lineup keep the same familiar shape. Altuve receives 500 plate appearances and 105 starts at second base. Walker receives 550 plate appearances, 120 starts at first base and 15 at DH. That leaves room for Paredes and the younger players while retaining the contracts in the payroll calculation.

Correa receives 350 plate appearances, with 50 starts at third base, 25 at shortstop and five at DH. This is a deliberately limited planning role following ankle surgery. It gives Houston a way to use a productive Correa without displacing Pena for a full season or pretending that four veteran infielders can all receive uninterrupted everyday work.

Contact quality sets the expectations

Expected statistics sharpen the argument against solving every problem through the veteran lineup. Altuve's .308 wOBA exceeded his .271 expected mark. Walker's .316 exceeded .305. Paredes also outperformed his expected figure, .355 to .323, although his plate discipline provides a separate source of value. These gaps are evidence to investigate, not automatic amounts to subtract from next year's production.

Cam Smith presents a different case. His .293 wOBA trailed a .316 expected mark, and his .390 slugging percentage trailed .420 expected slugging. He also supplied 21 home runs. Houston should preserve the opportunity for improvement while asking why similar headline OPS results in 2025 and 2026 have not yet turned better contact into broader production.

The smaller samples deserve even less certainty. Lucas Spence's .826 Houston OPS came in 63 plate appearances, while his expected wOBA was below the actual result. Zach Cole's brief MLB line points the opposite way. Neither sample settles an everyday job. Both belong in a competition with a specified amount of opportunity and a practical replacement route.

Handedness should change the lineup selectively

The table uses full-season MLB splits so traded players are not represented by only their Houston stretch. The useful question is whether a split can improve a particular assignment, not whether every left-handed or right-handed hitter should become a platoon player. The smaller samples should carry less influence than the overall record.

Alvarez remains a central hitter against either side. Houston should not reduce his opportunities merely to create a tidy platoon chart. The more useful decisions involve the final outfield places, backup catching and how frequently Altuve, Correa and Walker rest. A lineup can gain value by making those choices deliberately while keeping its best hitters in their primary jobs.

The same discipline applies to external candidates. Grisham's on-base profile, Mullins' range and Higashioka's catching experience solve different problems. None becomes the preferred acquisition simply because his handedness fits an empty box. The price, actual position, displaced innings and quality of the internal replacement must all agree.

Range is an independent problem

Houston has several reasons to protect the pitching staff with better range. Altuve recorded minus-nine Outs Above Average in 2026 after minus-five in 2025 and minus-nine in 2024. Pena moved from plus-eight in 2025 to minus-six in 2026. One reversal does not establish a permanent decline, but it makes an automatic defensive rebound an unsafe foundation for the staff plan.

Smith and Matthews each recorded plus-six OAA. Matthews' 167 opportunities were far fewer than Smith's 338, and his work covered multiple positions. Barrosa's plus-six came in 128 opportunities across his full MLB season. That is useful evidence for a defensive role, not proof that either player can sustain the same rate over an entire season in center field.

Walker remained positive at first base. Moving him freely around the diamond would sacrifice a defined contribution to solve an unrelated playing-time problem. Paredes and Correa should share the corners and shortstop relief within a coherent arrangement, while Matthews supplies genuine second-base flexibility. The defensive plan has to be workable before any run-prevention improvement is entered into the argument.

Center field is a development choice

The principal plan allocates 85 center-field starts to Matthews, 55 to Barrosa and 22 to Trammell. Matthews started 32 Houston games there in 2026 and ten at second base. Barrosa had only three Houston center-field innings, so the full-season range number must not be described as a large Astros center-field sample.

This is the largest everyday-position risk in the preferred roster. Matthews hit .587 OPS in 233 plate appearances, while Barrosa's full MLB season produced a .616 OPS in 174. Their gloves can help, but Houston would be accepting a weak offensive floor at an important position. The benefit is preserving money for the staff and receiving a meaningful evaluation of players already controlled.

The alternative is explicit. A 500-plate-appearance, 145-start center fielder displaces 300 plate appearances and 85 starts from Matthews, 150 and 55 from Barrosa, and 50 and five from Trammell. The additional purchase cannot be presented as free improvement while those internal players retain their original workloads.

Catching requires more than a batting line

Yainer Diaz hit .772 OPS with 14 home runs in 361 plate appearances. His offensive contribution should not obscure the receiving evidence: minus-4.18 framing runs and minus-5.17 total catching runs in the full-season Statcast records. The categories are related and should not be added as separate penalties.

The main plan gives Diaz 110 catching starts and 500 total plate appearances. Salazar takes 32 starts, Price 20. Salazar's positive framing result came in a limited sample; Price's was smaller still. Their recent Sugar Land performance is more informative about batting depth than their combined 35 MLB plate appearances, but it does not guarantee major-league receiving or offense.

Higashioka is a priced contingency at one year and $3-4 million, with 180 plate appearances and 45 catching starts. He replaces 100 plate appearances and 25 starts from Salazar and 80 and 20 from Price. His own 2026 framing figure was negative, so this would be an experience and depth purchase, not a demonstrated receiving upgrade. The money is additional to the depth reserve.

The rotation must reduce free bases

The staff's 650 walks represented 4.07 per nine innings over 1,437 innings. Starters issued 378 walks in 774 2/3 innings; relievers issued 272 in 662 1/3. More strikes alone will not solve every matchup, but this much traffic makes it difficult to preserve starters' pitch counts, control innings and keep the bullpen in favorable roles.

Brown remains the best retained starting foundation, with a 3.19 ERA and 3.32 xERA in 107 1/3 innings. Lambert supplied 155 2/3 innings with a 3.82 ERA and 3.70 xERA. The plan asks for 155 and 150 innings respectively, rather than treating Brown's shorter season as proof of 200-inning availability or Lambert's strong year as an unlimited workload.

Javier, Imai and Burrows require distinct scrutiny. Imai walked 51 batters in 76 2/3 innings. Burrows allowed 21 home runs in 94 2/3. Javier finished with a 5.14 ERA and 4.76 xERA. Their salaries and potential are reasons to work on the problems, not evidence that the problems have already been corrected.

Build a bullpen with usable middle innings

Hader's 1.01 ERA and 1.93 xERA in 44 2/3 innings supplied a dominant result over a limited workload. The plan uses 55 innings and retains a medical review of his biceps history. It does not require him to absorb an entire departed reliever's workload on top of his own.

Blubaugh is important because he can cover more than three outs. His 96 innings included 42 walks, however, and the 3.66 ERA was below his 4.27 xERA. King and Sousa also need careful interpretation: both had expected ERAs below their actual results, but that does not make every difficult inning a product of bad luck. The next staff needs improvement in execution and role selection.

Reifert adds strikeout potential, having struck out 81 in 58 Durham innings. He also walked 34. A 2.48 Triple-A ERA is not sufficient evidence to assign him immediate high-leverage MLB work. His 40-inning principal role is an opportunity with a command condition. Santa and VanWey likewise receive useful work without their very small MLB samples being projected as established full-season performance.

Health belongs in the workload calculation

The dated team update lists Correa after ankle tendon repair, Loperfido after hip surgery and Walter after Tommy John surgery. Burrows had elbow neuritis and shoulder inflammation; the current account says he was available by the postseason. These are dated recovery facts, not an examination or a guarantee of future availability.

Loperfido and Walter receive no principal workload. Blanco also remains outside the main pitching allocation. Their salaries or roster positions are not removed to make the plan appear cheaper. A successful return would create a genuine choice about who loses playing time, rather than an extra layer of production added to an already complete season.

Wesneski receives 130 innings after 61 MLB innings in 2026. That is an aggressive progression requiring current medical and workload judgment. Brown, Javier, Imai, Burrows and Hader also need current review. The combined failure case removes Javier and Imai after the no-purchase plan, then reallocates all 40 starts and 210 innings. Individual hitter contingencies remain separate; several concurrent absences would require a fresh lineup rather than mechanically adding every branch together.

The farm can supply choices, not a complete rescue

The upper-minors evidence is mixed. Spence had a .748 OPS at Sugar Land and .804 at Corpus Christi before his brief Houston success. Price produced .758 at Triple-A, Salazar .778, and Ferreras .837. Those are usable development signals with different ages, positions and sample sizes, not interchangeable projections of MLB offense.

The pitching depth has similar tradeoffs. Ullola struck out 90 in 66 1/3 Sugar Land innings but walked 42. Hendrickson offered 113 2/3 innings and 27 walks, with a 4.75 ERA. Mayer's 2.61 Corpus Christi ERA and 100 strikeouts in 69 innings support an emergency development path, while his additional Triple-A work must also be considered before promotion.

Janek and Schiavone remain development cases rather than automatic answers to the MLB catching problem. Neyens, Hughes and Radel occupy longer-horizon prospect slots. The preferred plan does not bring them forward to satisfy a numerical shortage in 2027. Promoting a player should follow demonstrated readiness and an actual job, not the need to make a roster spreadsheet balance.

Correct the contractual starting point

The payroll source's 2027 guaranteed-contract headline omits the exercised Imai and Paredes options. Adding $18 million and $13.35 million raises the eight-contract allocation to $183.417 million. The Paredes top-ten MVP mutual-option provision remains a release-date check, with award voting still pending.

These amounts are annual allocations. Correa's figure includes a prorated bonus, and the separate credits include the corresponding offset plus $10 million from Minnesota. Altuve and Alvarez also have bonus allocations. Treating the displayed allocation as the exact amount of checks written during the calendar year would mix different accounting bases.

Walker and Javier reach the end of their guaranteed terms after 2027. Imai's later option is separate from his now-retained 2027 season. Those future decisions affect flexibility, but they do not reduce today's requirement to fund the roster. A possible trade also creates no usable savings until there is an actual counterparty, salary allocation and acceptable return.

Arbitration and the full roster cost

Nine retained arbitration estimates total $38.6 million. Paredes' exercised option replaces his earlier arbitration estimate, and Meyers' departure removes him from the retained arbitration group. The estimates are not agreements. Blanco's salary remains included despite the absence of a principal workload.

The bridge adds an $18-22 million pool for pre-arbitration players, minor-league depth and callups, plus the identifiable Weiss buyout, then applies the Correa credits. It produces $229.18-233.18 million before acquisitions. Benefits, tax, cash timing and any future collective-bargaining changes require separate treatment. The assumed roster pool is not a claim that the 2027 minimum salary is already known.

The prior-year player-payroll estimate of approximately $239.90 million gives scale, not permission. Plan A would exceed it. Houston therefore faces a real decision: authorize a higher proposed allocation, remove salary through a justified transaction, or accept the additional internal workload. This report credits none of those possibilities before it occurs.

Compare the jobs before repeating the names

Martinez is preferred because 26 walks in 180 innings directly address Houston's traffic problem. His 109 strikeouts and 3.98 xERA also explain the limit on the bid. The 3.10 ERA is not the only figure in the evaluation. He replaces the additional starts otherwise assigned to Burrows, Ullola, Pecko and Hendrickson, plus small increases for four retained starters.

The same $26-30 million two-year ceiling appears in other S&S dossiers. Here it purchases a control-oriented 165-inning role. Washington's broader rotation construction and Detroit's alternative-starter job are different contexts, and neither proposal reserves Martinez for Houston. Griffin at $28-32 million is a distinct alternative with more strikeouts and a higher walk total, rather than a second signing added to Plan A.

Okert's $16-20 million two-year ceiling matches the recent Detroit and Washington planning basis. Houston's job is retention of 60 relief innings after other departures. The earlier Kansas City inquiry used $8-10 million for one year. Similar annual spending does not make the commitments equivalent: the second year is a real additional risk. Clarke is cheaper but comes with a 4.42 xERA, which argues against pricing him as interchangeable late-inning production.

Three funded plans and optional upgrades

Plan A's $27-33 million annual additions combine Martinez at $13-15 million per year, Okert at $8-10 million and the $6-8 million reserve. Total new commitment, including that one-year reserve, is $48-58 million. Plan B uses Gray at $20-24 million for one year, producing $34-42 million annual additions and $42-52 million total commitment. Plan C uses only the reserve.

The center-field inquiry is separately funded. Varsho or Mullins at one year and $8-10 million raises Plan A's annual additions to $35-43 million and its total allocation to $264.18-276.18 million. Grisham at two years and $18-22 million instead raises annual additions to $36-44 million and the total allocation to $265.18-277.18 million. None of these purchases is hidden inside the depth reserve.

Mullins brings the strongest recent OAA result of those three, but his .264 expected wOBA is a major offensive concern. Grisham's .337 expected wOBA offers a different argument, alongside minus-six OAA. Varsho's range helps, while his Houston and full-season batting lines justify a strict price limit. A qualifying offer, changed medical evidence or a more expensive market must be priced before an agreement, not appended afterward to an already strained budget.

A full season of plate appearances

The 6,200-plate-appearance allocation is a season-long plan. It includes callups and substitutions rather than claiming that eighteen hitters fit on the Opening Day roster. It also avoids counting a full season for every veteran and prospect who could plausibly contribute. The amount of work assigned is the first constraint on any production argument.

The infield balance is deliberate. Paredes gets 100 starts at third base and 35 at first; Correa gets 50 and 25 at third and short; Altuve gets 105 at second, with Matthews taking 40. Alvarez receives 115 DH starts and 30 in left. Walker keeps first base as his primary job. Each position totals 162 starts, and no individual receives more than 162.

There is a cost to this arrangement: Correa's expensive contract receives a reduced role, and the outfield asks young hitters to improve. A trade could eventually change the allocation, but an unspecified salary dump is not a roster plan. Until an acceptable transaction exists, the analysis should show the actual crowding and the choices it imposes.

A full season of pitching

The preferred rotation supplies 870 innings across 162 starts, an average of 5.37 innings per start. Relievers receive the remaining 570 innings. This is a planning total of 1,440, not a prediction of precisely how many innings Houston will play; home losses, extra innings and shortened games affect actual totals.

Ullola's 20 starting innings and 25 relief innings are combined into a 45-inning MLB assignment. Pecko's 20 and 30 combine to 50. Blubaugh's 85 relief innings acknowledge an established multi-inning role without silently adding starts. These distinctions matter because the same prospect cannot be counted as a full rotation replacement and a full bullpen solution simultaneously.

The principal workload already asks for progression from several pitchers. Its value comes from reducing the size of those asks through one dependable external starter, not eliminating all development risk. If Martinez's price rises beyond the ceiling, Houston should compare the Gray branch and internal plan with current evidence instead of pretending that any available veteran supplies the same innings at the same quality.

What happens when the plan fails

The internal plan redistributes all of Martinez's 30 starts and 165 innings and all 60 of Okert's relief innings. It preserves the same total season and payroll obligations. The combined no-purchase, no-Javier, no-Imai case then replaces a further 40 starts and 210 innings, using Brown, Wesneski, Burrows, Ullola, Pecko, Hendrickson and Mayer.

The combined test balances numerically, but its competitive implications are severe. Burrows, Ullola and Pecko would take much larger roles, and Mayer would need a major-league opportunity before one has been assumed in the principal plan. The scenario demonstrates the dependence on depth. It does not demonstrate that the substitutes would match the removed pitchers' performance.

The Correa, Diaz and Alvarez branches also name the removed plate appearances and starts. They are individual tests. Several simultaneous hitter failures would alter both the workload limits and the defensive arrangement, especially at catcher and shortstop. At that point, Houston would need a new combined roster and a newly priced acquisition decision rather than a stack of incompatible optimistic assumptions.

The scale of improvement

Houston went 39-42 at home and 42-39 away. It scored 344 runs at home and 390 on the road, while allowing 375 and 391. These are observed splits, not a park-adjusted estimate of true talent. The road offense's stronger result is another reason to avoid explaining the entire season through a general claim about the home ballpark.

The sensitivity table uses the simple formula 162 times runs scored squared divided by the sum of runs scored squared and runs allowed squared. It shows the scale of possible changes, not a forecast or a price-to-wins conversion. Preventing 40 runs would materially alter the team's underlying margin. Losing a large amount of Alvarez's offense would move the same calculation sharply in the other direction.

Track walks, starter length, range and young-player usage alongside the record. April wins alone will not show whether those weaknesses have improved.

The recommendation and dated roster

Randy Flores was introduced as general manager on November 2, with the managerial appointment still pending in the dated source. The roster recommendation does not depend on inventing a future manager's preferences. It depends on the existing contracts, the young players' actual opportunities and a staff that needs fewer walks and more dependable starts.

Approve the Martinez and Okert negotiating ranges only if ownership accepts the corrected total allocation. Keep the Gray branch available as a different annual-cost and commitment decision. Preserve an explicit center-field inquiry, but do not quietly spend that money before deciding which internal development opportunity it replaces. Keep Loperfido, Walter and Blanco outside the principal workload until their individual cases support more.

The roster and market evidence is dated November 3, 2026. The November 3 review reconciles named-player status, observed options, unsigned arbitration estimates, medical reporting and competing candidate roles. The current recommendation is to buy control and usable innings, make the veteran infield fit honestly, and give the young outfield a defined test. Houston's next contender needs a more reliable pitching foundation beneath its best hitter.