A backtest pick from the 2022-23 to 2025-26 seasons: the model's total was worked out from that season's earlier games only, exactly as it would have been before the game, and compared with the market's line at the time.
| Model | Market | Actual | |
|---|---|---|---|
| Total goals | 59.0 | 56.5 | 54 |
| Model total | 59.0 |
| Market line | 56.5 |
| Model − line | +2.5 |
This league plays the over, when the model is 2+ goals above the line. Pick: Over 56.5 at 1.89. Loss: 54 goals.
Each team gets an attack and a defence rating from this season's games before this one (N-Lubbecke 14, Lubeck-Schwartau 14). 1.000 is league average: an attack of 1.100 scores 10% more than its opponents usually concede, and a defence of 0.900 lets in 10% fewer than its opponents usually score. The ratings allow for who each team has played, and each one is pulled towards 1.000 as if the team had also played 6 average games, so a few games do not swing it too far. Figures are rounded, so the last digit may differ.
| Step | Calculation | Result |
|---|---|---|
| League average goals for the home side | from 127 games played earlier in the season | 29.95 |
| League average goals for the away side | the same games | 28.78 |
| N-Lubbecke attack × Lubeck-Schwartau defence | 0.995 × 1.029 | 1.024 |
| N-Lubbecke goals | 29.95 × 0.995 × 1.029 | 30.65 |
| Lubeck-Schwartau attack × N-Lubbecke defence | 0.976 × 0.974 | 0.951 |
| Lubeck-Schwartau goals | 28.78 × 0.976 × 0.974 | 27.36 |
| Both teams together | 30.65 + 27.36 | 58.01 |
| League correction | how far this league's real totals ran from the raw figure, 2022-23 to 2024-25 | +0.96 |
| Model total | 58.01 + 0.96 | 59.0 |