LeedsvsCrystal Palace

Leeds are favoured at 55% to beat Crystal Palace, a margin that reflects their superior league position and recent momentum rather than a dominant underlying advantage. The model's confidence rests on Leeds' four-game unbeaten run and their defensive solidity in recent weeks, but Crystal Palace's injury troubles—particularly the absence of key attacking and defensive personnel—may matter more than the raw probability suggests. This is a match where Leeds' home record and Palace's away struggles could prove decisive, though both sides arrive with thin seasonal samples.

Where the model stands

The model assigns Leeds a 55% win probability, with draws at 23% and Crystal Palace at 22%. This represents a modest but meaningful edge for the hosts. The Elo ratings support this view: Leeds sit at 1515 compared to Palace's 1444, an 71-point gap that translates into the model's slight-to-moderate favouritism. The model leans toward Over 2.5 goals at 58%, suggesting an expectation of attacking play despite both teams' recent defensive frailties. The handicap market—Crystal Palace +0.5 at 56%—indicates the model sees Leeds winning by a single goal as the most likely decisive outcome.

Form and the table

Leeds occupy sixth place with 8 points from four matches (2W-2D-0L), whilst Crystal Palace languish in 17th with just 3 points (1W-0D-3L). The gap is stark. Leeds have not lost in their last five games, winning one and drawing three; they scored 1.8 goals per game over that span whilst conceding only 0.8. Crystal Palace, by contrast, have lost their last match and won only once in five, averaging 1.5 goals for and a concerning 2.8 against. At home, Leeds have recorded 1 win and 1 draw from 2 matches (2.00 points per game). Away, Palace have managed 1 win and 1 loss from 2 (1.50 ppg), a record that underperforms their overall season trajectory.

The injury list for both sides is substantial. Leeds are without Mateo Hošević, J. Rodon, Ilija Gruev, Wilfried Nzonté and G. Gudmundsson. Crystal Palace's absences are more acute: A. Disasi (suspended following red card), D. Henderson, Jean-Philippe Mateta, C. Riad, Eddie Nketiah and I. Sarr. Palace's attacking depth is compromised, with Mateta unavailable and Nketiah sidelined. T. Mitchell (2 goals) remains their leading scorer this season, though the sample is extremely thin at four matches played.

Model vs market

The model and market-implied probabilities align closely on the match winner: both place Leeds at 55%. The draw probability shows a marginal difference (model 23%, market 24%), whilst the market slightly underestimates Crystal Palace at 20% versus the model's 22%. These are negligible divergences, suggesting consensus on the underlying match dynamics. The total-goals market shows the model leaning Over 2.5 at 58%, a lean that reflects recent defensive vulnerabilities on both sides rather than attacking prowess.

What the AI panel thinks

The AI panel shows broad agreement on a Leeds win, with probabilities ranging from 55% to 75%. The Scorebase Model (55%), GPT-5.6 (58%), Grok (58%), Gemini (58%) and Kimi K3 (62%) cluster around the model consensus, whilst Claude (68%) and Qwen 2.5 (75%) express stronger conviction in Leeds. On the handicap, most models favour Leeds -0.5 (ranging from 55% to 65%), with only the Scorebase Model backing Palace +0.5 at 56%. The total-goals market splits: five models lean Over 2.5, whilst Claude and Grok prefer Under 2.5.

Verdict

Leeds' unbeaten run, superior table position and home advantage form a coherent case for favouritism. Crystal Palace's injury burden—especially in attack—compounds their away-day struggles. The model's 55% for Leeds is justified but not emphatic; this is a fixture where a draw or a narrow Palace defeat remain plausible outcomes. The thin seasonal sample (four matches) warrants caution on any strong conviction.

Prediction: Leeds 55% (model pick); Over 2.5 goals 58% (lean).