Four games, first tip at 7:30 PM ET, 86 players in the pool. The matchups: Las Vegas hosts Toronto, Atlanta visits Portland, Indiana takes on Connecticut, and Washington meets Los Angeles. That last one shows up twice in the raw data, which tells you the doubleheader math is real — there are enough minutes in this slate to build from.
The shape of tonight is this: injuries have reshuffled the minute distributions in at least three games, which means the model's injury-edge signal is running hot. On a four-game slate, that's not background noise — it's the primary construction lens. The Vegas-Toronto game in particular is dripping with redistributed minutes across multiple players, and the Atlanta-Portland game isn't far behind. Indiana and Washington-LA round out the slate with their own angles, but the first two games are where most of tonight's projected upside is concentrated.
The model is working from exponentially weighted per-minute rates (half-life of 10 games), blended with opponent defense and pace context, then run through 5,000 Monte Carlo draws per player. Floor and ceiling figures are each player's own measured volatility — p10 and p90 of that distribution. Worth keeping in mind as you read through: this is a probabilistic range, not a promise.