Tyrese Maxey scored 25.9 points a game in 2023-24, up from 20.3 the season before, and his true shooting slipped from 60.5% to 57.3% over the same stretch (Basketball-Reference). What grew was his opportunity: almost four more minutes a night and a bigger share of Philadelphia's plays. That's where predicting player stats starts. Work out how much of tonight's game a player will get before you judge how well he'll play it.
Five inputs move a basketball player's line. Check them in this order.
| Input | What to check | Where to look |
|---|---|---|
| Minutes | His last three to five games against his season average | His game log on Basketball-Reference |
| Usage rate (USG%) | His share of the plays, and who's out tonight | NBA.com stats and the injury report |
| Matchup | The opponent's pace and defensive rating | Team ratings on Basketball-Reference |
| Rest | A second night of a back-to-back, and his status that day | The NBA's official injury report |
| Playoff rotation | Who the coach trusts late in close games | Box scores from the series so far |
Minutes come first
Every counting stat is a per-minute rate multiplied by minutes, so a change in minutes moves every line at once. Maxey shows how much: he scored 20.3 points in 33.6 minutes a game in 2022-23. Keep that scoring rate, stretch it over the 37.5 minutes he played in 2023-24, and you get about 22.7 points. Minutes alone explain roughly 2.4 of his 5.6-point jump.
The rest came from a bigger role. Philadelphia traded James Harden on November 1, 2023 (Basketball-Reference), and Joel Embiid played only 39 games that season (Basketball-Reference). Somebody had to take those shots.
Before you trust a minutes change, make sure it's real. One game tells you very little, because foul trouble or a blowout can swing a player's minutes a long way without touching his role. Treat a change as real once it has held for three straight games and you can point to a cause such as an injury or a trade. Chasing a big night with no reason behind it is recency bias, one of the habits that getting better at sports predictions means breaking. The same care applies when a player comes back from an injury: teams often hold him to limited minutes for his first few games, so his season average can overstate what he'll get.
Then look at tonight's game itself. When one team is a heavy favorite, the starters on both sides can spend the fourth quarter on the bench once the score gets out of hand, so a lopsided matchup puts his minutes at risk before tip-off. You can size that up from the odds in a few seconds with a favorite vs underdog read.
Usage rate (USG%) shows who gets the shots
The NBA.com stats glossary defines usage rate, written USG%, as "the percentage of team plays used by a player when they are on the floor." A play counts as his when he takes a shot, commits a turnover or goes to the line for free throws. Basketball-Reference's formula counts each free throw attempt as 0.44 of a play, since most trips to the line are two shots and some are one.
Five players share the floor, so the average player uses 20% of the plays while he's out there. Few get far past that: in 2025-26, 14 of the 164 players who logged 1,500 minutes had a usage rate of 30% or higher (Basketball-Reference). The highest single season on the all-time leaderboard is Russell Westbrook's 41.7% in 2016-17.
Usage moves when the roster does. Jaylen Brown's usage rate went from 28.9% in 2024-25 to 36.2% in 2025-26 (Basketball-Reference), a season in which Jayson Tatum played 16 games (Basketball-Reference). Brown's minutes barely changed, from 34.3 a game to 34.4. His shots went from 17.7 a game to 21.7, and his scoring from 22.2 to 28.7. The share shrinks again when the star comes back, so before you trust a fill-in scorer's big month, check whether the teammate he's filling in for is due back.
Shots follow usage more reliably than points do. In the 34 games Maxey played without Embiid in 2023-24, he took 21.8 shots a game, up from 18.8 when Embiid played, and his scoring barely moved: 26.3 points against 25.6 (from Maxey's and Embiid's game logs). So when a star sits, raise his teammates' shot counts first. Whether the extra shots turn into points is an efficiency question, and true shooting percentage is the number that answers it.
Pace and defense set the size of the night
Pace is the number of possessions a team plays per 48 minutes, per the NBA.com glossary. In 2025-26, Miami played at 103.4, the fastest in the league, and Boston at 94.8, the slowest (Basketball-Reference). Over 36 minutes, a player who uses 28% of his team's plays gets about 22 of them at Miami's speed and about 20 at Boston's. That's two fewer shots or trips to the line before the defense has done anything.
Then there's the defense. Oklahoma City allowed 107.7 points per 100 possessions in 2025-26, the best mark in the league, and Washington allowed 122.7, the worst. Boston paired the slowest pace with the fourth-best defense, while Washington played faster than the league average of 99.4. A scorer who draws Boston starts the night with fewer chances against a better defense than one who draws Washington. For one number that sums up a team at both ends, use net rating.
Team numbers set the baseline. After that, look at who'll guard him. A team with a top defender at his position can take a real bite out of his night even at an average speed, so check the likely matchup before you settle on a number.
Rest decides whether he plays at all
Teams average 14.2 back-to-backs in the 2026-27 schedule, and none plays more than 16 (NBA.com). On the second night of one, the biggest swing is a missed game. The league's player participation policy even allows pre-approved back-to-back restrictions for older players and for players with heavy career workloads or serious injury histories (NBA.com). The pattern shows up in game logs. From his season debut on November 18, 2025, LeBron James played 7 of the Lakers' 11 second nights and 53 of their other 57 games (Basketball-Reference).
Teams must list any player whose status is in doubt by 5 p.m. local time the day before most games, and by 1 p.m. local time on game day for the second game of a back-to-back (NBA injury report rules). A status you read the night before a second game can change by lunch, so check it again before you commit.
Playoff rotations shrink to the players a coach trusts
The playoffs change two inputs at once. Games slow down: league pace fell from 99.4 in the 2025-26 regular season to 95.7 in the 2026 playoffs, and points per 100 possessions fell from 115.8 to 112.0 (Basketball-Reference's season and playoff pages). Scoring dropped from 115.6 points per team game to 107.9. Free throw attempts went the other way, from 23.5 to 24.8, so the whistle didn't go quiet in the 2026 playoffs.
Coaches also shorten the bench. The champion Knicks had seven players average 16 or more minutes in the playoffs, and nobody else topped 14 (Basketball-Reference). Basketball-Reference's regular-season and playoff per-game pages show what that did to five players.
| Player | Minutes, 2025-26 season | Minutes, 2026 playoffs | Points, season / playoffs |
|---|---|---|---|
| Cade Cunningham, Pistons | 33.9 | 40.9 | 23.9 / 28.1 |
| Victor Wembanyama, Spurs | 29.2 | 34.1 | 25.0 / 23.8 |
| Shai Gilgeous-Alexander, Thunder | 33.2 | 36.3 | 31.1 / 27.6 |
| Harrison Barnes, Spurs | 25.8 | 9.2 | 9.9 / 2.3 |
| Aaron Wiggins, Thunder | 21.8 | 5.8 | 9.4 / 1.5 |
Cunningham's jump is almost all minutes. His regular-season scoring rate over 40.9 minutes works out to 28.8 points, and he averaged 28.1. Wembanyama and Gilgeous-Alexander played more and still scored less, in slower games against playoff defenses. For bench players, the minutes can drop to almost nothing.
Elimination games push it further. In Game 7 of the Western Conference finals, Gilgeous-Alexander played 42:58 and Wembanyama played 41:54, while Wiggins didn't play and Barnes played 3:25 (box score). In the first round, Maxey played 45:15 in Game 7 at Boston after averaging 38.0 minutes in the regular season, and he scored 30 (box score). Before a playoff game, look at who closed the last one and who the coach sent in first when a starter picked up early fouls, because those are the players he trusts.
What an early read is worth in GAGE
Say a wing has averaged 22 points in 30 minutes this season. His team's lead guard went down a week ago, and he's played 36 or more minutes in every game since. The odds still lean on his season average and give Points 25+ a 40% chance, so a basketball-style pick in GAGE would show YES +6.0 and NO +4.0. Pick YES, and you'll score +6.0 if he reaches 25 or lose 4.0 if he doesn't.
The odds keep moving until the game starts. Say they reach 55% as everyone else catches up to his new role. One of your friends who takes YES then gets +4.5 if it lands and −5.5 if it misses, while your pick keeps its 40% lock. If 55% is the true chance, your pick is worth +1.5 points on average and theirs is worth zero. The full rules are on how scoring and lock-in work, and GAGE is on the App Store.