Every prop line is a number someone set. But some are further from what the data shows than others. When Rithmm's models predict a player's steals total lands 0.5 to 0.75 below the sportsbook's posted line, the under has hit 81 times in 118 tries this NBA season. That's a 68.6% win rate and a 14.7% ROI on the season.
Tuesday's San Antonio vs Portland game activates that exact pattern on Victor Wembanyama.
San Antonio Spurs vs Portland Trail Blazers
April 29, 2026 | 1:30 AM ET
Rithmm's models project 1.0 steals for Wembanyama in this game. The sportsbook has the line set at 1.5. That 0.5-steal gap puts the prediction squarely in the delta window where the backing pattern has built its 81-37 record.
This pick is also tagged as a Parlay Piece inside the Rithmm app. The Parlay Piece tag means three criteria are met simultaneously: the model's prediction lands on the right side of the line with positive DTM, the prop has hit in at least 70% of the player's last 10 games, and the historical win rate clears 60%. All three apply here.
Wembanyama typically averages right around one steal per game. The line at 1.5 gives the sportsbook extra room, but Rithmm's model doesn't see him getting there Tuesday night. The pattern behind this pick has been profitable across 118 similar bets this season. It's not a Wembanyama-specific trend. It's a structural signal that shows up repeatedly when this delta window appears, across many players throughout the year.
That's the kind of context most bettors don't have before they place a prop. You see the line. You make a guess. Rithmm finds the window where the data is pointing in one direction and surfaces it automatically.
Wembanyama's steals prop is live in Rithmm right now. Open the NBA section, pull up Tuesday's San Antonio game, and you'll see the pick flagged alongside its pattern record and a plain-English explanation of what the model is detecting. No jargon, no black box. Just the data, laid out clearly.
