A relative strength
Opponent free-throw attempt rate
22.8% · rank 3 of 366 · 38 games
Which defenders can protect the paint without surrendering free throws?
← Program librarySource edition Sep 12, 2026
The program dossier / men’s basketball
Duke
2025–26 performance, personnel and game context.
A starting point for 2026–27 preparation.
Independent preseason modelNo.1
+34.6 adjusted net
per 100 estimated possessions
35–32025–26 recorded finals
125.4Adjusted offense · 2026–27 baseline
90.8Adjusted defense · lower is better
66.8Model tempo · possessions / 40 min
Publisher lens / exact source team ID
A second read on Duke.
Checking the attributed publisher rating…
01 / Both ends of the floor
The possession profile.
35–338 recorded games · 38 paired efficiency samples · pace 66.8Mean opponent adj. net 11.8 · 38 rated opponents
Offense
Effective FG%
56.6%38 games · higher is betterRank 15 / 366 · 96th favorable percentileDefense
Opponent effective FG%
46.2%38 games · lower is betterRank 16 / 366 · 96th favorable percentileOffense
Turnover rate
15.9%38 games · lower is betterRank 139 / 366 · 62th favorable percentileDefense
Turnovers forced
17.3%38 games · higher is betterRank 147 / 366 · 60th favorable percentileOffense
Offensive rebound rate
38.4%38 games · higher is betterRank 7 / 366 · 98th favorable percentileDefense
Opponent offensive rebound rate
25.2%38 games · lower is betterRank 14 / 366 · 96th favorable percentileOffense
Free-throw attempt rate
38.2%38 games · higher is betterRank 87 / 366 · 76th favorable percentileDefense
Opponent free-throw attempt rate
22.8%38 games · lower is betterRank 3 / 366 · 99th favorable percentileShooting and ball-movement detail
| Metric | Observed rate | Games | Definition |
|---|---|---|---|
| Offensive efficiency | 122.2 | 38 | Points per 100 estimated possessions. |
| Defensive efficiency | 95.2 | 38 | Opponent points per 100 estimated possessions. |
| Two-point FG% | 60.2% | 38 | (FGM − 3PM) / (FGA − 3PA). |
| Opponent two-point FG% | 46.4% | 38 | Opponent (FGM − 3PM) / (FGA − 3PA). |
| Three-point FG% | 34.7% | 38 | 3PM / 3PA. |
| Opponent three-point FG% | 30.6% | 38 | Opponent 3PM / 3PA. |
| Three-point attempt share | 43.5% | 38 | 3PA / FGA; a style measure, not a quality grade. |
| Opponent three-point attempt share | 45.0% | 38 | Opponent 3PA / FGA; a style measure. |
| Free-throw accuracy | 72.9% | 38 | FTM / FTA. |
| Opponent free-throw accuracy | 71.2% | 38 | Opponent FTM / FTA; not attributed to defensive skill. |
| Assisted field-goal share | 58.8% | 38 | Assists / FGM; describes recorded ball movement. |
| Opponent assisted FG share | 52.3% | 38 | Opponent assists / FGM. |
Window changes describe the historical sample; they do not refit the preseason model. Full-season ranks require 10 games. Percentiles compare qualifying programs in the model’s rated field. Source coverage can differ by metric.
NCAA source context / 2025–26
What did the possessions look like?
2,534 source possession rows across 38 recorded team-games. These are descriptive NCAA source aggregates: they do not assign individual credit or enter the forecast.
02 / The film queue
Turn a number into a question.
A relative strength
Offensive rebound rate
38.4% · rank 7 of 366 · 38 games
Who crashes the glass, and how does that affect transition defense?
A closer look
Turnover rate
15.9% · rank 139 of 366 · 38 games
Where do turnovers occur under pressure, and who can relieve that pressure?
A closer look
Turnovers forced
17.3% · rank 147 of 366 · 38 games
Which defensive actions produce takeaways, and what shots do they concede?
These are stat-led film prompts, not verified explanations of a scheme. Lower-ranked factors can still be above average. A five-point-or-closer final occurred 9 times in the selected window; Duke won 6. That small-sample record is not a clutch-skill estimate.
03 / How the season moved
Read beyond the average.
● Offense● Opponent offenseRaw points / 100 possessions · game order
Mar 29, 2026 · neutral vs UConn · L 72–73 · offense 114.7 / defense 116.3
Higher offense and lower opponent offense are better. Opponent strength and venue are not adjusted here. Missing paired boxes create gaps.
Open the game-by-game scouting log · 38 games
| Game / venue | Result | Current opponent rank | Off. eff. | Def. eff. | Pace | eFG% | TO% | ORB% |
|---|---|---|---|---|---|---|---|---|
| UConnMar 29, 2026 · neutral | L 72–73 | 10 | 114.7 | 116.3 | 62.8 | 58.3 | 20.7 | 40.0 |
| St John'sMar 27, 2026 · neutral | W 80–75 | 14 | 124.8 | 117.0 | 64.1 | 58.0 | 14.0 | 41.4 |
| TCUMar 21, 2026 · neutral | W 81–58 | 48 | 116.8 | 83.6 | 69.3 | 62.2 | 24.5 | 39.1 |
| SienaMar 19, 2026 · neutral | W 71–65 | 221 | 114.7 | 105.0 | 61.9 | 44.9 | 12.9 | 44.4 |
| VirginiaMar 15, 2026 · neutral | W 74–70 | 35 | 116.4 | 110.1 | 63.6 | 44.6 | 14.2 | 48.8 |
| ClemsonMar 14, 2026 · neutral | W 73–61 | 26 | 115.5 | 96.5 | 63.2 | 55.0 | 17.4 | 32.1 |
| Florida StMar 12, 2026 · neutral | W 80–79 | 81 | 119.0 | 117.5 | 67.2 | 51.6 | 22.3 | 53.7 |
| North CarolinaMar 7, 2026 · home | W 76–61 | 22 | 116.9 | 93.8 | 65.0 | 49.3 | 12.3 | 42.9 |
| NC StateMar 3, 2026 · road | W 93–64 | 45 | 144.3 | 99.3 | 64.5 | 63.6 | 10.9 | 45.2 |
| VirginiaFeb 28, 2026 · home | W 77–51 | 35 | 126.4 | 83.7 | 60.9 | 60.4 | 13.1 | 25.9 |
| Notre DameFeb 25, 2026 · road | W 100–56 | 114 | 139.8 | 78.3 | 71.5 | 55.9 | 12.6 | 38.9 |
| MichiganFeb 21, 2026 · neutral | W 68–63 | 6 | 110.0 | 101.9 | 61.8 | 50.9 | 17.8 | 39.4 |
| SyracuseFeb 17, 2026 · home | W 101–64 | 102 | 157.4 | 99.8 | 64.2 | 72.1 | 6.2 | 40.7 |
| ClemsonFeb 14, 2026 · home | W 67–54 | 26 | 108.2 | 87.2 | 61.9 | 51.9 | 16.1 | 28.1 |
| PittFeb 11, 2026 · road | W 70–54 | 71 | 114.2 | 88.1 | 61.3 | 63.7 | 21.2 | 33.3 |
| North CarolinaFeb 7, 2026 · road | L 68–71 | 22 | 106.7 | 111.4 | 63.8 | 50.8 | 12.5 | 31.3 |
| Boston CollegeFeb 4, 2026 · home | W 67–49 | 152 | 109.1 | 79.8 | 61.4 | 53.0 | 14.7 | 22.6 |
| Virginia TechJan 31, 2026 · road | W 72–58 | 79 | 125.0 | 100.7 | 57.6 | 56.8 | 13.9 | 44.4 |
| LouisvilleJan 27, 2026 · home | W 83–52 | 32 | 128.0 | 80.2 | 64.9 | 57.4 | 18.5 | 46.9 |
| Wake ForestJan 24, 2026 · home | W 90–69 | 60 | 138.9 | 106.5 | 64.8 | 55.6 | 13.9 | 50.0 |
| StanfordJan 17, 2026 · road | W 80–50 | 77 | 126.9 | 79.3 | 63.0 | 60.0 | 15.9 | 40.7 |
| CaliforniaJan 15, 2026 · road | W 71–56 | 104 | 109.9 | 86.7 | 64.6 | 50.8 | 15.5 | 40.0 |
| SMUJan 10, 2026 · home | W 82–75 | 51 | 117.9 | 107.8 | 69.6 | 59.8 | 18.7 | 26.7 |
| LouisvilleJan 7, 2026 · road | W 84–73 | 32 | 112.8 | 98.0 | 74.5 | 57.0 | 16.1 | 22.9 |
| Florida StJan 3, 2026 · road | W 91–87 | 81 | 130.5 | 124.8 | 69.7 | 64.2 | 21.5 | 40.0 |
| Georgia TechDec 31, 2025 · home | W 85–79 | 137 | 118.2 | 109.9 | 71.9 | 47.7 | 16.7 | 47.7 |
| Texas TechDec 21, 2025 · neutral | L 81–82 | 16 | 115.0 | 116.4 | 70.4 | 61.5 | 15.6 | 20.7 |
| LipscombDec 16, 2025 · home | W 97–73 | 144 | 117.6 | 88.5 | 82.5 | 58.0 | 26.7 | 52.6 |
| Michigan StDec 6, 2025 · road | W 66–60 | 13 | 102.7 | 93.3 | 64.3 | 46.4 | 12.4 | 25.0 |
| FloridaDec 3, 2025 · home | W 67–66 | 4 | 102.8 | 101.3 | 65.2 | 50.9 | 18.4 | 31.4 |
| ArkansasNov 28, 2025 · neutral | W 80–71 | 23 | 114.1 | 101.3 | 70.1 | 55.9 | 17.1 | 38.7 |
| HowardNov 23, 2025 · home | W 93–56 | 223 | 131.5 | 79.2 | 70.7 | 66.9 | 15.6 | 33.3 |
| NiagaraNov 22, 2025 · home | W 100–42 | 331 | 154.4 | 64.9 | 64.8 | 73.6 | 13.9 | 45.8 |
| KansasNov 19, 2025 · neutral | W 78–66 | 17 | 118.2 | 100.0 | 66.0 | 52.6 | 12.1 | 37.1 |
| Indiana StNov 15, 2025 · home | W 100–62 | 140 | 130.6 | 81.0 | 76.6 | 60.6 | 15.7 | 48.5 |
| ArmyNov 12, 2025 · road | W 114–59 | 340 | 148.0 | 76.6 | 77.1 | 69.2 | 13.0 | 43.3 |
| W CarolinaNov 8, 2025 · home | W 95–54 | 245 | 124.3 | 70.6 | 76.4 | 54.7 | 13.1 | 32.5 |
| TexasNov 5, 2025 · neutral | W 75–60 | 34 | 112.6 | 90.1 | 66.6 | 51.0 | 15.0 | 32.4 |
04 / The people behind the possessions
Personnel and workload.
2025–26 recorded affiliations, independent of the split selector. This is not a confirmed 2026–27 roster. Inspect roster observations →
| Player | GP | MIN/G | PTS/G | REB/G | AST/G | TS% | Est. usage | Usage games | AST/TO | 3PA |
|---|---|---|---|---|---|---|---|---|---|---|
| Cameron BoozerF | 38 | 33.5 | 22.5 | 10.2 | 4.1 | 65.3 | 30.0% | 38 | 1.6438 games | 13838 games |
| Isaiah EvansG | 38 | 28.3 | 15.0 | 3.2 | 1.3 | 59.0 | 24.9% | 38 | 1.1238 games | 28038 games |
| Dame SarrG | 38 | 23.5 | 6.4 | 3.8 | 1.1 | 51.4 | 15.8% | 38 | 1.0538 games | 13338 games |
| Cayden BoozerG | 38 | 22.8 | 7.7 | 2.3 | 3.0 | 59.2 | 17.3% | 38 | 2.2238 games | 6638 games |
| Caleb FosterG | 33 | 24.9 | 8.3 | 3.5 | 2.8 | 54.1 | 17.9% | 33 | 2.5633 games | 9833 games |
| Maliq BrownF | 38 | 20.2 | 4.9 | 5.2 | 1.6 | 63.4 | 12.7% | 38 | 1.3838 games | 3038 games |
| Nikolas KhameniaG | 38 | 19.7 | 5.7 | 3.3 | 1.0 | 56.7 | 15.0% | 38 | 1.3038 games | 10338 games |
| Patrick Ngongba IIC | 32 | 21.9 | 10.1 | 5.8 | 2.0 | 64.5 | 21.1% | 32 | 1.5232 games | 3132 games |
| Darren HarrisG | 36 | 9.6 | 3.3 | 0.8 | 0.4 | 54.3 | 16.3% | 36 | 2.6036 games | 6936 games |
Usage is a minutes-based estimate over the stated sample, not a direct measure of on-court possessions. Low-minute estimates are noisy. AST/TO is unavailable when recorded turnovers are zero; missing fields are never filled with zero.
05 / Recruiting evidence
What changed around Duke?
This panel keeps dated school publications, source-listed roster observations and prior college production in separate evidence lanes. A source record does not establish eligibility, availability or a future role.
11Source-listed players
9Same-program prior records
0Different-program records
59%Prior minutes represented
Role workload watch
Which source-listed role is represented?
Prior minutes are grouped by the source-reported position on the current listing. “Unrepresented” means prior minutes not attached to a same-program listing in this release; it does not establish a departure, transfer, eligibility or expected role.
| Role | Listed | Prior minutes | Returning minutes | Incoming prior minutes | Returning share | Unrepresented |
|---|---|---|---|---|---|---|
| Guard | 9 | 3,758 | 3,758 | — | 100% | — |
| Forward | 0 | — | — | — | — | — |
| Center | 2 | 724 | 724 | — | 100% | — |
0 listed players have no recognized source position. The workload denominator is the matched prior-production sample (4,482 minutes).
Dated school publications
3 announced additions
| Player | Evidence | Prior program | Prior MPG | Prior PPG | Latest publication |
|---|---|---|---|---|---|
| Bryson HowardPrep addition | Addition announced ↗Duke announced Bryson Howard as a 2026–27 addition. | Heritage High School | — | — | Nov 20, 20251 dated record |
| Cameron WilliamsPrep addition | Addition announced ↗Duke announced Cameron Williams as a 2026–27 addition. | St. Mary’s High School | — | — | Nov 20, 20251 dated record |
| Maxime MeyerPrep addition | Addition announced ↗Duke announced Maxime Meyer as a 2026–27 addition. | IMG Academy | — | — | Nov 20, 20251 dated record |
Roster observation
Who is listed in the source release?
11 source-listed players. Prior rates below are attached by exact source ID where available; missing values remain missing.
| Player | Observation | Prior MPG | Prior PPG | Prior TS% |
|---|---|---|---|---|
| Dame Sarr →G | Same programDuke Blue Devils | 23.5 | 6.4 | 51.4% |
| Cayden Boozer →G | Same programDuke Blue Devils | 22.8 | 7.7 | 59.2% |
| Caleb Foster →G | Same programDuke Blue Devils | 24.9 | 8.3 | 54.1% |
| Nikolas Khamenia →G | Same programDuke Blue Devils | 19.7 | 5.7 | 56.7% |
| Patrick Ngongba II →C | Same programDuke Blue Devils | 21.9 | 10.1 | 64.5% |
| Darren Harris →G | Same programDuke Blue Devils | 9.6 | 3.3 | 54.3% |
| Cameron Sheffield →G | Same programDuke Blue Devils | 2.9 | 0.5 | 33.3% |
| Jack Scott →G | Same programDuke Blue Devils | 2.9 | 0.6 | 66.7% |
| Ifeanyi Ufochukwu →C | Same programDuke Blue Devils | 4.4 | 0.8 | 44.7% |
| Brock Davis →G | No prior appearance observedNo prior program listed | — | — | — |
| Sebastian Wilkins →G | No prior appearance observedNo prior program listed | — | — | — |
Recruiting review edition Sep 8, 2026 · coverage is partial (14 programs reviewed). See the full methodology and coverage map before drawing a roster conclusion.
Look ahead / 2026–27
The next assignments.
Checking the live schedule edition…
Army vs Duke
53.4Army
89.5Duke
Duke win estimate98.4%
Model signalStrong lean
Projected home margin36.1
Projected total142.9
80% home-margin range20.2 to 52.1
Estimated possessions66.3
Why the model tiltsfour-factor edge
Shot qualityAway 0.2 pts
Duke attack 57.1% · Army defense 53.6%Army attack 49.0% · Duke defense 45.4%
Ball securityAway 0.1 pts
Duke attack 15.7% · Army defense 15.7%Army attack 17.5% · Duke defense 17.6%
Second chancesHome 4.0 pts
Duke attack 36.2% · Army defense 29.6%Army attack 27.8% · Duke defense 25.1%
Free-throw pressureAway 2.6 pts
Duke attack 37.0% · Army defense 36.2%Army attack 29.4% · Duke defense 26.0%
Cameron Indoor Stadium
Preseason baseline · roster changes are not model features.
No verified pregame market line imported. Check the forecast record →
Coastal vs Duke
54.3Coastal
86.1Duke
Duke win estimate97.5%
Model signalStrong lean
Projected home margin31.8
Projected total140.3
80% home-margin range15.9 to 47.7
Estimated possessions66.1
Why the model tiltsfour-factor edge
Shot qualityHome 3.9 pts
Duke attack 57.1% · Coastal defense 49.6%Coastal attack 49.0% · Duke defense 45.4%
Ball securityAway 3.2 pts
Duke attack 15.7% · Coastal defense 12.9%Coastal attack 17.1% · Duke defense 17.6%
Second chancesHome 5.3 pts
Duke attack 36.2% · Coastal defense 28.3%Coastal attack 27.8% · Duke defense 25.1%
Free-throw pressureAway 3.5 pts
Duke attack 37.0% · Coastal defense 34.0%Coastal attack 32.5% · Duke defense 26.0%
Cameron Indoor Stadium
Preseason baseline · roster changes are not model features.
No verified pregame market line imported. Check the forecast record →
Michigan St vs Duke
67.2Michigan St
73.8Duke
Duke win estimate70.7%
Model signalLean
Projected home margin6.6
Projected total141.0
80% home-margin range-9.4 to 22.5
Estimated possessions66.0
Why the model tiltsfour-factor edge
Shot qualityHome 2.1 pts
Duke attack 57.1% · Michigan St defense 46.6%Michigan St attack 53.8% · Duke defense 45.4%
Ball securityHome 0.1 pts
Duke attack 15.7% · Michigan St defense 16.8%Michigan St attack 16.6% · Duke defense 17.6%
Second chancesHome 2.7 pts
Duke attack 36.2% · Michigan St defense 23.7%Michigan St attack 34.9% · Duke defense 25.1%
Free-throw pressureAway 5.3 pts
Duke attack 37.0% · Michigan St defense 30.9%Michigan St attack 37.4% · Duke defense 26.0%
United Center
Preseason baseline · roster changes are not model features.
No verified pregame market line imported. Check the forecast record →
Read the evidence correctly.
The top-line adjusted ratings use the full published preseason model. Split statistics below them are unadjusted observations from 2025–26. Opponent ranks are the current model’s ranks, not what was known before each historical game. The “top 50” split is descriptive and does not enter model evaluation.
Rates pool numerators and denominators from games with all fields required for that metric. Missing values remain unavailable; every metric shows its game count. Ranks require at least 10 games and compare programs in the model’s rated field. Ties share rank; percentiles use their average position. High percentiles mean favorable values only for metrics with a defined direction.
Player workloads refer to recorded 2025–26 appearances. Estimated usage divides FGA + 0.475 × FTA + turnovers by minutes-prorated team opportunities in those same games. It assumes team opportunity rates were constant while a player was on court; it is not a lineup or play-by-play measurement. Overtime is included in available minutes. Minutes and usage samples can differ.
Source statistics: SportsDataverse bulk releases, CC BY 4.0. Calculations and model: Silvermine. Definitions draw on publicly documented Four Factors and basketball rate statistics; these are not another publisher’s proprietary ratings. Model notebook →