The Start Week lets you decide from which week onward to factor games into the PPG calculation. For instance, if you pick Start Week 2, only games played in Week 2 or later count toward the Off Strength and Def Strength. This is useful if you want to ignore early-season results or focus on more recent performance trends.

By default, PPG compares each game to what the opponent had shown before that matchup. Turning on retroactive averages means every game is compared against the opponent's full-season scoring and allowance once the year is complete.

Use this when you want to know how strong a performance looks after the season wraps up. Leave it off to see the in-the-moment context that was available entering the game.

Accuracy reflects how often the model’s predicted winner (based on the strength gap) matched the actual winner. It’s simply correct picks ÷ total graded games.

This view only counts games with final scores recorded after the selected start week.

These are the same controls as the rankings page:

• Start Week filters out early-season games before the chosen week.
• Retroactive averages decide whether opponent baselines use only prior weeks (off) or the whole season (on).

Games are grouped by the predicted strength gap. Wider gaps indicate higher confidence.

Each band shows how often the favored team actually won and the average implied probability in that band. It’s a quick way to see if big projected edges translate to higher hit rates.

Calibration compares what the model thought would happen to what actually happened across ranges of predicted win probability.

Ideally the average predicted chance matches the true win rate in each bucket. Divergence signals that the probability curve might need tuning.

This table highlights a subset of recent games so you can spot where the model aligned with reality and where it missed.

Rows in yellow denote incorrect calls. Use the predicted margin and probability columns to gauge whether the miss was a low-confidence upset or a surprising result.