dfsforge / NBA forecasts / Jason Preston assists

Jason Preston: assistsFITTED MODEL

G · LAC · NBA · baseline 2.8/game (2026, 4 games)
2.5
Median
0.2-5.6
80% range
1.1-4.0
50% range
6.7
Ceiling (p95)

Outcome distribution

0.22.55.6
Shaded: the 80% range. Dashed lines: p10 / median / p90.

Chance of reaching

assistsProbability
1.0+77%
2.0+59%
3.0+39%
4.0+25%
5.0+15%
Model receipt: on the held-out season, across 9094 G assists games, our 80% range covered 80.4% of actual outcomes (target 80%).
Baseline is the 2026 per-game average; the distribution shape is learned per position and per production level from historical game logs. When the weekly slate model is live the baseline switches to the weekly line.
More: Jason Preston player page · all NBA forecasts
Forecasts are probability distributions from dfsforge's variance model, fitted on historical game logs and validated on a held-out season - the coverage number printed above is that holdout result, not a training-set fit. Research output, not advice. Open the app · Accuracy receipts