Final Four: Pace vs. Efficiency Math
A forecast for a basketball game needs two scores. A projection of 69.12 points for Team A cannot be compared with a full-game total, which includes both teams. This worked example separates possessions, scoring rates, and the market being evaluated.
Every team, rating, pace, and price below is hypothetical. The calculation illustrates a model you can audit; it is not a current Final Four prediction or a validated betting system.
Keep the Units Consistent
An offensive rating measures points scored per 100 possessions; a defensive rating measures points allowed per 100 possessions. Under this convention, lower defensive ratings are better. A rating of 90 means fewer points allowed than 110 on the same basis.
Tempo measures possessions per team over a specified game length. The two teams have nearly equal numbers of possessions, so adding their individual paces double-counts the opportunities used in a per-team score calculation. See Ken Pomeroy's ratings glossary for these definitions and the distinction between raw and adjusted ratings. The simple matchup formula below is our illustrative assumption, not a reproduction of that rating system.
Declare Both Teams and the Baseline
Assume a neutral court, a 40-minute regulation game, no overtime, and ratings measured on the same scale and period:
| Input | Team A | Team B |
|---|---|---|
| Offensive rating, points per 100 possessions | 120 | 110 |
| Defensive rating, points allowed per 100 possessions | 100 | 90 |
| Historical tempo, possessions per team per 40 minutes | 68 | 68 |
Assume the league-average offensive rating is 100. For this exercise, estimate each matchup scoring rate as its offensive rating multiplied by the opponent's defensive rating divided by the league average. This multiplicative model requires validation before use on real games.
Calculate a Complete Score
Use 68 possessions per team as the baseline forecast. That is an explicit assumption, not a conclusion that either team controls the pace.
- Team A rate: 120 × 90 / 100 = 108 points per 100 possessions, or 1.08 points per possession.
- Team B rate: 110 × 100 / 100 = 110 points per 100 possessions, or 1.10 points per possession.
- Team A score: 1.08 × 68 = 73.44 points.
- Team B score: 1.10 × 68 = 74.80 points.
- Combined score: 73.44 + 74.80 = 148.24 points. Projected margin: Team B by 1.36 points.
Fractional scores describe model means, not possible final scoreboards. Keep unrounded inputs until the final display.
Change Pace Without Silently Changing Efficiency
Holding both scoring rates fixed gives a simple sensitivity check:
| Possessions per team | Team A points | Team B points | Combined points |
|---|---|---|---|
| 64 | 69.12 | 70.40 | 139.52 |
| 68 | 73.44 | 74.80 | 148.24 |
| 72 | 77.76 | 79.20 | 156.96 |
Each additional possession per team adds 1.08 + 1.10 = 2.18 combined points under these assumptions. The 64-possession scenario is a sensitivity case; it does not establish that a slower opponent will force that tempo. A fuller model can allow efficiency to vary with pace, lineups, and game state, but needs evidence for those adjustments.
Compare the Matching Market
Against a hypothetical regulation-only full-game total of 145.5, the baseline mean of 148.24 is 2.74 points higher. The 64-possession scenario is 5.98 points lower. Compare 69.12 only with a Team A team-total market under the same period rules, never with the combined total.
A mean above a line does not give the probability of finishing above it. At an illustrative -110 price, a binary win/loss selection needs a win probability above 110 / 210 = 52.38% to have positive expected net return. Estimating that probability needs a distribution of scores and validated uncertainty, not just the difference between a mean and a line. An overtime-inclusive market also needs an overtime-inclusive forecast.
For an OwnTheLines exercise, keep an external worksheet of both ratings, the assumed possession count, forecast time, available line and price, and settlement period before recording a supported selection. Review the settled result separately from the assumptions. The platform does not provide this pace model or guarantee team-total availability. No market overreaction or pricing advantage has been demonstrated here.
Continue with March Madness Math for tournament probabilities and NBA Point Spreads vs. Totals for the distinction between score means and market probabilities.
Pace and Totals Questions
Q: Is a higher defensive rating better?
A: No. When the rating measures points allowed per 100 possessions, lower is better. Compare ratings with the same definitions and adjustment method.
Q: Should I add the teams' paces?
A: No. Estimate one shared possessions-per-team figure for the game. In this hypothetical example the baseline is 68, with 64 and 72 used only as sensitivity cases.
Q: What does the 69.12-point projection represent?
A: It is Team A's score at 64 possessions. Team B projects to 70.40, giving a combined regulation total of 139.52 points.
Q: Does a projected total establish an advantage?
A: No. You also need the score distribution, settlement scope, offered price, and evidence that the model estimates probabilities reliably.