Key factors

Context onlyFacts that explain the pick. The model doesn't read them.

Worked out from the stored data, never estimated.

League context

Context onlyTable position isn't in the model.

Where the teams sit in the Thai League 2, before kick-off.

Phrae United: recent fixtures

Partly in the modelThe model uses each team's last 6 league matches. Cup and other matches here are context.

Matches finished before this one, newest first, with the score from their side.

Kanchanaburi: recent fixtures

Partly in the modelThe model uses each team's last 6 league matches. Cup and other matches here are context.

Matches finished before this one, newest first, with the score from their side.

Head to head

Partly in the modelThe model uses league meetings from this season and last, moving expected goals by at most 4%. Older and cup meetings are context.

How these two have done against each other. Recent meetings count for more than old ones.

Form, home and away

In the modelThe model rates each team's home and away records separately.

The last 5 and 10 matches, then Phrae United at home and Kanchanaburi away only: a home fixture is read from home matches where there are enough.

Team strength, attack and defence

Context onlyShown for context. The model's own ratings come from goals scored and conceded this season and last.

Schedule and rest

Context onlyRest and fixture congestion aren't in the model.

Squad availability

In the modelMissing players move the model's expected goals, chiefly a missing top scorer.

Players out or doubtful, as reported the night before.

Counter-signals

Context onlyFacts pointing the other way. The model doesn't read them.

Evidence pointing the other way.