Houston Rockets

As of: 2026-07-18 · Team: HOU · Coach: Ime Udoka · Up: 2026-27 Fantasy Basketball Encyclopedia

Verified roster ledger

The current NBA team page was reconciled on July 18 with the 2026-27 Basketball Reference contract table and the NBA's 2026 offseason transaction tracker. The evidence supports 13 standard contracts, one confirmed two-way, one probable continuing two-way, and four official-page holdovers whose Houston contracts are not established for 2026-27.

Josh Okogie and Dorian Finney-Smith are excluded, but not as free-agent departures — see the transactions ledger below. Finney-Smith was traded to Charlotte; Okogie signed elsewhere. An official roster page can lag contract reality, so the four holdovers receive dossiers but no baseline Houston minutes.

Transactions ledger

Two trades, both omitted from earlier drafts of this desk. They are the causal chain behind the guard logjam described throughout this file.

Net: Houston subtracted forward and wing minutes (Finney-Smith 16.8 MPG over 37 games, Okogie ~17) and added only guards (Smart, Bogdanović, Thornton). Four future seconds are gone, which constrains later depth-adding.

Trade market

Per Kelly Iko (Yahoo Sports, June 22 2026), rival teams in contact with Houston were told Amen Thompson is the "only near untouchable" player on the roster. The precise reporting matters: Iko does not say Houston is shopping Durant or Şengün, only that it has not shut down inquiries and is willing to hear scenarios rather than run the group back. Houston's front office concluded after the Finals that the team is "a ways away" from contention.

A three-team concept sending Durant to Detroit, Şengün to Boston, and Jaylen Brown to Houston was discussed by the three clubs (Brett Siegel); Houston never engaged, having invested in Durant a year earlier, and Brown was later traded to Philadelphia. So the concrete framework died, but the standing posture did not.

For fantasy, this is a team-change risk line on the desk's two highest-graded assets rather than a projection change. Neither dynasty grade is marked down here — Şengün and Durant carry their value to any destination — but a manager drafting either should price a non-trivial chance the team context in this file is stale by opening night.

Depth chart

The projected starting five follows the July 8 Houston Chronicle lineup analysis: VanVleet, Thompson, Durant, Smith, and Şengün. This is a projection, not a team-announced October lineup.

*Unsigned or unresolved for 2026-27 as of the cutoff.

Exact 240-minute baseline

This is a healthy, neutral active-night rotation. It sums to exactly 240 regulation minutes, excludes overtime, and does not pretend that the two-way and unsigned players appear simultaneously. The deep-reserve projections marked with an asterisk are conditional appearance lines rather than shares of the healthy rotation.

Read the MPG column as a healthy-night rate, not a season average. Multiplying the Proj. GP column by the Proj. MPG column yields 16,761 player-minutes against a season requirement of 240 × 82 = 19,680 — a gap of 2,919 minutes, about 35.6 per game, that the healthy-night table by construction assigns to nobody. That gap is the absence load: Durant's 16 projected missed games, Şengün's 12, VanVleet's 20, Adams' 27, and the rest. Those minutes are real and they flow downward, which means every reserve's season MPG runs meaningfully above his active-night number and the counting lines below are floors rather than medians for the bench. An independent availability-weighted model that does close to 19,680 lives in 2026-27 NBA minutes projections; this table is deliberately the other kind and is not a substitute for it.

Two rows in particular should not be read literally. Bogdanović at 7 and Capela at 4 are means of bimodal distributions, not roles. Neither is a seven- or four-minute player on a night he is in the rotation; both are closer to scratch-or-play. Bogdanović projects to roughly 14-18 minutes in the ~50 games he appears meaningfully, Capela to roughly 12-15 in ~45, with DNP-CDs otherwise — coaches play a center twelve-plus minutes or sit him, not four. Their D grades below are artifacts of averaging across the DNPs and should be read as "unrosterable in standard formats, live streaming target whenever the man ahead of him sits," not as judgments that the players are finished.

Correction 2026-07-18: Jabari Smith Jr. raised from 29 to 32 active minutes, with the three minutes taken from VanVleet (27→26), Sheppard (26→25), and Smart (12→11); his points grade moves C+ to B-. The earlier line cut the starting power forward by 6.1 MPG off a 35.1 actual while Houston's forward room got thinner, not deeper — Finney-Smith traded to Charlotte and Okogie to Utah, with the only additions being three guards (Smart, Bogdanović, Thornton). No reporting supports a Smith minutes cut; coverage has him holding the starting job on a five-year extension (SI). The compression belongs in the guard rotation that actually got crowded. Durant's 31.0 is deliberately left alone: it is not an unsourced cut but a reported one — Rafael Stone said at Summer League, "I definitely want to take some of the load off Kevin. Last year I do think he played too many minutes" (HoopsHype, July 10 2026). Total remains exactly 240.

The nine-category lines are this desk's medians, not vendor projections. Points grades use the reference formula PTS + 1.2×REB + 1.5×AST + 3×STL + 3×BLK − TO; platform settings can materially reorder players. The linked 2026-27 NBA minutes projections is an independent availability-weighted model and is deliberately more aggressive on Thompson/Şengün and less confident in Smith's role. That disagreement is preserved rather than hidden.

Fantasy hierarchy

• Nine-category core: Şengün is the best balanced early-round target if a manager can absorb roughly 70% FT and elevated turnovers. Thompson is a category-warping top-40 bet with elite FG%, rebounds, and steals but almost no threes. Durant still has top-25 per-game skill, although 38-year-old totals risk argues against paying only for rate. Sheppard is the swing pick: 26 minutes can already return strong threes and steals; a VanVleet setback creates top-40 upside.
• Middle and late rounds: Smith offers scalable threes, rebounds, blocks, and low turnovers but may again lack usage. Eason can be a stocks/rebounds specialist without starting. VanVleet's assists and steals survive the conservative workload, yet the knee recovery makes both efficiency and totals fragile.
• Points leagues: Şengün and Thompson lead, then Durant. Sheppard gains less from steals than in nine-cat under some platform formulas. Smith, Eason, and VanVleet are format-sensitive; the remaining veterans are streamers unless the rotation breaks open.
• Dynasty: Thompson is foundational and is the one player Houston has told rival teams is near untouchable. Şengün is a comparable on-court asset but not a comparable Houston asset — the June 22 reporting has Houston willing to hear inquiries on him, so his dynasty case should be made on the player rather than on the fit described in this file. Sheppard is the high-upside second tier; Smith and Eason are useful but role-capped. Durant is win-now only, while Thornton and Copeland are deep developmental stashes.

Public ranks disagree in useful ways. Dynatyze's July 3 board placed Şengün 15th, Thompson 16th, Sheppard 58th, Durant 62nd, Smith 95th, and Eason 104th. FantasyPros had Thompson 22nd and Sheppard around 62nd in its July expert view. A Reddit way-too-early nine-cat board on April 14 put Durant 18th, Şengün 25th, and Thompson 31st, while SI placed Thompson only 43rd on June 4. The gap is mostly format, availability, and shooting-punt assumptions.

Role battles and swing events

The depth problem is real. Smart's July 10 signing added a former starter who averaged 28.5 minutes for the Lakers, while Bogdanović's July 8 one-year deal added another veteran ballhandler. At center, Şengün, Adams, and Capela cannot all repeat their prior active-game workloads. This baseline treats Smart, Bogdanović, and Capela as the first players squeezed, but Udoka could instead reduce Sheppard, Eason, or Smith.

Health and availability

• VanVleet had right ACL reconstruction and repairs to both menisci on September 25, 2025. He said in July that he expected to be ready for the 2026-27 opener, but that target sits near the long end of a normal 9-12 month return and does not guarantee unrestricted minutes. This is the roster's largest fantasy branch.
• Adams had left-ankle surgery in January and missed the final 42 games; the team expected a full recovery. His 55-game projection includes both recovery and rotation risk.
• Bogdanović missed 59 games with a hamstring problem before Houston signed him. His rate projection is less concerning than whether he can both stay available and beat the guard logjam.
• Durant played 78 regular-season games but missed Houston's playoff opener with a bruised right knee. The injury was not reported as chronic; age and workload, not that bruise alone, drive the 66-game projection.
• Eason played 60 games. Houston then re-signed him for five years, strong evidence that the franchise still values his role even though it does not settle starter versus reserve minutes.

2025-26 team baseline and 2026-27 outcome range

Houston finished 52-30 and lost its first-round series 4-2. The official end-of-season game notes show the mechanism: 115.2 points per game, league-leading rebounding, and 110.0 opponent points per game, but only 25.4 assists. VanVleet's return should help organization; Smart and Bogdanović add decision-making depth; the spacing and overlapping-center problems remain.

• Quantitative prior: Andrew's July 17 RAPM model in 2026-27 NBA team outcomes projects 55.1 wins, a 47-62 P10-P90 band, and a 99.4% playoff probability. It also flags many replacement/rookie rows and high rating uncertainty.
• Desk median: 52-30, top four in the West, with the defense and offensive rebounding carrying an uneven half-court attack.
• Ceiling: 57-60 wins and a conference-finals path if Durant remains elite, VanVleet and Adams return without meaningful restrictions, and Sheppard/Thompson improve the playoff offense.
• Floor: 43-47 wins and a play-in/first-round exit if the rehab returns stall, Durant misses a long stretch, or non-shooting lineups again let opponents overload Şengün and Thompson.

Evidence and media scan

• Official: NBA roster, end-of-season game notes, contracts, and offseason deals, accessed July 17-18.
• Role reporting: Houston Chronicle lineup combinations, July 8; the projected opening five and the roster's increased versatility are reporting/analysis, not a coach guarantee.
• Fantasy: Dynatyze, FantasyPros Thompson, FantasyPros Sheppard, Athlon's Sheppard breakout case, and CBS's VanVleet-return warning, checked July 18.
• Video/podcast: The April 7 Dunker Spot segment argued that Thompson and Sheppard's development would determine Houston's postseason ceiling and specifically highlighted Sheppard's spacing around Durant/Şengün actions.
• Community: Fantasy Reddit supplied early draft-price checks and daily-line examples; Rockets Reddit supplied lineup and Summer League hypotheses. These are sentiment/mechanism inputs only. A July 15 thread correctly cautioned that Summer League production has little direct regular-season correlation.
• Social: The user-supplied Alex Barutha July 17 boom-or-bust post did not name a Rocket. Searches of Ethos Fantasy NBA, Dynatyze, and RotoWireNBA did not yield an independently recoverable, dated Houston projection beyond the linked web items by the cutoff; absence from a search result is not a ranking signal.

Open questions

• Is VanVleet fully cleared for five-on-five work at camp, and what is his opening-month cap?
• Does Udoka prioritize Sheppard's shooting, Smart's defense, or Bogdanović's veteran creation as the ninth-man need changes?
• Can Adams and Capela both remain in the nightly rotation behind Şengün, or is one purely injury insurance?
• Are Newton and Crawford under current 2026-27 two-way/qualifying agreements, and where do Holiday, Tate, and Green sign?
• Does Houston consolidate depth before opening night, changing the minute tree and the 55-win model input?
• Do Durant or Şengün actually move, or was the June posture only leverage? Every projection in this file is conditional on both opening the season in Houston.
• Does Thompson sign his rookie-scale extension before the October deadline, and at what percentage of the cap — Bobby Marks' near-max or Shankar's 20-23%?
• Houston sits at 13 standard contracts against a 14-man minimum. Does a veteran center signing (Nick Richards and Jonas Valanciunas have both been floated) zero out the Capela line?
• If Crawford accepts his two-way qualifying offer, which of Newton or Copeland loses a spot?