BallRanks
Premium
Menu
Blog

Player outlook

Reed Sheppard can lose the point guard job and keep his fantasy value

Houston's crowded backcourt will take the ball out of Sheppard's hands, but steals, blocks, and spot-up threes never needed it. The real threat is the minutes, and the right punt build turns him into a bargain.

October 8, 2026· Ball Ranks
Reed Sheppard
G · HOU
Reed Sheppard
Our rank
#62

Reed Sheppard's three best categories are worth more on the Ball Ranks board than his entire ranking. Steals, threes, and blocks together outscore the full nine-category line that puts him 62nd per game. Field-goal percentage takes back the difference.

No other player on the board is projected for at least 1.4 steals, 0.8 blocks, and two threes per game. Sheppard clears all three in 26.6 minutes. The rest of Model Zero's line is 13.3 points, 3.0 rebounds, 3.4 assists, 43.0% shooting, 82.5% at the line, and 1.5 turnovers across 72 games.

Houston's offseason is about to take the ball out of his hands. For fantasy purposes, that matters less than it sounds.

None of his best categories needs the ball

Fred VanVleet missed all of last season with a torn right ACL. Amen Thompson handled most of the lead-guard work, and Sheppard took over as the starting point guard at times. VanVleet is now fully cleared, Houston added Marcus Smart and Bogdan Bogdanović, and Ime Udoka has said Sheppard is likely to play more off the ball when he shares the floor with Smart.

Sort his value by what needs the ball. Assists and points are minor contributors to his rank; together they are worth less than his steals alone. The categories carrying him are the ones an off-ball guard keeps: steals, blocks, and catch-and-shoot threes.

The rate is what makes him unusual. His projected 1.5 steals and 0.8 blocks become 3.1 combined per 36 minutes. Among guards in our top 150, only Ausar Thompson and Dyson Daniels project higher. Neither comes close to his 2.8 threes.

Only three players rank among the board's top 20 in both steals and threes: Luka Dončić, Tyrese Maxey, and Sheppard. Dončić and Maxey rank fourth and fifth overall. Sheppard sits at 62nd because of the rest of his line, not because the categories he wins are any less scarce. A role change that leaves those two categories intact leaves the hard part of his value intact.

His steals rose and fell with his minutes

Sheppard played all 82 games last season, but his role moved month to month, and the defensive stats moved with it. From October through December, he averaged 25.8 minutes, 1.6 steals, and 2.6 threes. In January he dropped to 22.6 minutes, and the line sagged everywhere: 1.0 steal, 2.0 threes, 2.1 assists, and 38.5% shooting.

Then the role came back larger. From February through April, he averaged 28.3 minutes, 1.6 steals, 0.8 blocks, 3.3 threes, and 3.9 assists. March was the full-time version: 30.9 minutes, 1.9 steals, 1.0 block, 3.1 threes, and 4.6 assists, though with 2.6 turnovers.

Reed Sheppard's Steals per game trend

Steals per game by month, 25-26.

Full data
Reed Sheppard Steals per game by month, 25-26
MonthSteals per game
Oct1.3 (4 games)
Nov1.9 (13 games)
Dec1.4 (13 games)
Jan1 (17 games)
Feb1.4 (12 games)
Mar1.9 (16 games)
Apr1.3 (7 games)

Model Zero's 26.6 minutes sits between the January floor and the March ceiling, at almost exactly last season's 26.2 average. The projection is not betting on the starting job. It is betting that the steals-per-minute rate survives a bench role.

Punt field-goal percentage turns him into a fourth-round pick

Sheppard's 43.0% shooting on 11.8 attempts is the largest negative in his line, followed by 3.0 rebounds. Remove either one and the rest of his profile plays up fast.

Take field-goal percentage out of the scoring and his remaining eight categories rank 38th on the board. Take rebounds out instead and he ranks 36th. Drop both, the natural shape of a guard-heavy build, and he ranks 24th. Punt points helps less than it might seem, lifting him only to 49th, because his scoring was never the drag. Punt assists and punt FT% barely move him at all.

His 82.5% free throws and 1.5 turnovers are what make the punt work. Once field-goal percentage is gone, nothing else in his line is a serious drag, so the threes and steals arrive almost free.

He also climbs in season-long formats. Ball Ranks' totals board has him 44th, up 18 spots from his per-game rank, on 72 projected games.

A 10-minute playoff game is the warning

The case against Sheppard was on display in April. He took 20 shots in 36 minutes of Houston's first-round opener against the Lakers and made six, then played 10 minutes in Game 2. Across six playoff games, he shot 30.7% from the field and averaged 1.0 rebound, though he still produced 2.2 steals in 32.1 minutes. Houston lost the series 4–2.

VanVleet is the clearest threat to the minutes. He projects for 26.4 minutes across 61 games, and Smart projects for 25.2. The expected plan has Sheppard coming off the bench, while Thompson projects for 36.7 minutes.

If Udoka trims Sheppard to his January workload, the steals and threes shrink with it, and his field-goal percentage becomes a bigger share of what is left. Holding 26 minutes on a deeper roster is the assumption most likely to fail.

There is an early cushion. VanVleet will not play the second night of back-to-backs to open the season, which opens guard minutes on those nights.

Draft the defense, then build around the shooting

In a standard 12-team nine-cat league, Sheppard is an early sixth-round pick. Punt FG% and punt rebounds builds should take him in the fourth, and a manager punting both can take him at the end of the second round without reaching. Roto managers can move him up to the fourth on the totals rank alone.

Builds protecting field-goal percentage should leave him at his sixth-round price. For everyone else, he is the cheapest way on the board to buy elite steals and threes in the same pick.

See Reed Sheppard's full player profile, or compare Houston's crowded backcourt on the 2026–27 fantasy basketball board.

Sources and data notes