Between September 16 and September 19, 2026, the dixon_coles_v1 model logged 42 settled picks across 28 soccer matches, finishing with a 19-21-2 record and a net loss of -3.89 units (0.475 win rate). Spread plays managed a positive return at 9-7-1 for +1.42 units across 17 picks, but totals dragged the overall ledger down, going 10-14-1 for -5.31 units across 25 picks. Four days of match data represent a narrow sample size, and these short-term figures neither prove nor disprove any systematic edge.

Spreads Yield Modest Returns

The spread market produced a net gain driven by a few well-priced hits where model probability stood apart from bookmaker pricing. The largest single return came in Serie A, where the model assigned a 0.3389 probability to Lazio -1.5 against Venezia at a decimal price of 3.850. Lazio secured a 2-0 away victory to cover the line, returning +2.85 units.

In the Premier League, Brentford covered as a 0.5 home underdog against Chelsea in a 3-0 win, matching a model probability of 0.6992 at 1.550 odds; Igor Thiago and Jaidon Anthony each provided goals in the victory. Atlético Madrid covered -1.0 in a 4-0 win over Osasuna (0.4965 model probability, 1.813 price), receiving goals from Jonathan David (who added an assist), Álex Baena, Lee Kang-in, and Robin Le Normand. Real Betis also delivered on home spread duties, beating Getafe 1-0 on a goal from Abde Ezzalzouli to cover -0.5 (0.5332 probability, 1.690 price).

Not all spread positions landed. Espanyol failed to cover -0.5 at a 0.5031 probability and 1.826 price in a 1-3 home loss to Elche, where Fernando Niño scored two goals and Germán Valera added a third. Lens 0.5 also missed in Ligue 1 (0.6053 model probability, 1.806 price), losing 2-1 to Monaco as Paris Brunner scored for Monaco and Matthieu Udol scored for Lens.

High-Probability Misses Undercut Totals

The totals market proved far more punishing, burdened by high-confidence misses that offset underdog covers. The model's strongest projection in the totals market sat on Newcastle vs. Hull City under 2.5 goals, assigning a 0.9113 probability at a 2.430 price. Newcastle won 2-1, yielding three total goals and handing the ledger a 1.00 unit loss (though Hull City 1.5 spread covered at a 0.9957 probability and 1.599 price).

Another volatility spot occurred in Serie A, where Monza's 2-1 win over Sassuolo cleared the 2.5 line, sinking an under pick evaluated at a 0.3733 model probability and 1.952 price. Gustavo Varela scored both goals for Monza in that fixture. Conversely, the model hit a high-priced total in Germany, taking under 2.5 goals in VfB Stuttgart vs. Borussia Dortmund at a 0.3450 model probability and 3.050 price; Dortmund held on for a 0-1 away win to yield +2.05 units.

Within the window's tracking dataset, two settled picks lacked individual final scorelines and were graded within aggregate performance metrics without explicit match scores.

Market Breakdown

The complete output across the four-day span breaks down as follows:

Market Picks Record Win Rate Net Units
Spread 17 9-7-1 0.529 +1.42
Total 25 10-14-1 0.400 -5.31
Overall 42 19-21-2 0.475 -3.89

Evaluating model health over a four-day frame offers context on line movement and pricing discrepancies, but a 42-pick sample remains far too small to draw broad conclusions about long-term edge.