🎯

Fixture360 only lists fixtures from a curated set of leagues, chosen specifically to keep our prediction model accurate.

📰

Shock result as Chelsea defy the odds to beat Fulham 1-2.

📝 Match Report

How It Unfolded

Chelsea beat Fulham 1-2 at Craven Cottage, Regular Season - 33, in the Premier League. With the result on the board, the sections below measure it against everything the model and the numbers projected in advance.

The Model vs The Result

The Poisson model went into this projecting Fulham 1.45 xG and Chelsea 1.09 xG, a combined 2.54. The scoreboard read 1-2 for 3 actual goals. Chelsea outscored their 1.09 projection by 0.9. Those figures were built on strength ratings of Fulham attack 1.01 / defence 0.94 against Chelsea attack 0.81 / defence 0.97, drawn from 70/70 games (CurrentSeason).

On the result, the model split it Fulham 45% | Draw 27% | Chelsea 28%, with Fulham to win its most likely call at 45%. The actual Chelsea win had been the model's second-ranked read at 28%, so the result diverged from the top call without being a shock.

Goals & Markets

The model rated Over 2.5 goals at 47%. The game delivered 3, so it went over — the opposite of the projected direction. Over 1.5 had been 72% and landed. On both teams to score, the model sat at 51% and the match saw both sides score — another call in the model's favour. Historically the two sides average 63% Over 2.5 across their same-competition games (Fulham 60%, Chelsea 66%), a base rate that agreed with today's over. Their combined BTTS history runs at 61%, which matched the both-scored outcome.

Trading Patterns vs Reality

Fulham's trading profile (70 games, 35 home) set some clear expectations: BTTS lands in 59% of their matches — today it did.

Chelsea's trading profile (70 games, 35 away) set some clear expectations: BTTS lands in 64% of their matches — today it did.

Form vs Result

The form lines were close — Fulham 1.36 PPG, Chelsea 1.67 PPG — offering no strong steer, and the Chelsea win broke the near-deadlock.

Verdict

Grading the result against the stored data, 1 of 3 pre-match signals came in: result (miss), Over 2.5 (miss), BTTS (hit). The data was on the wrong side of this result more often than not — a reminder that a single match sits well inside the model's variance.

💡 Key Insights