Wednesday, 4 May 2011

Cornering The Market

Long-time readers will know that the Elo based rating spreadsheet has steadily evolved since it's inception, and currently uses shots, shots-on-target and corners as factors in rating a team's performance in addition to goals, which taken alone can produce a distorted view of a match.

For example, on 20 December 2010, Manchester City played Everton.

Shots - Manchester City 25, Everton 4.
Shots-on-Target - Manchester City 15 Everton 3.
Corners - Manchester City 11 Everton 0.

Result: Manchester City 1 Everton 2.

The result here is not an accurate reflection of the match, but what would be? The Reep ratio (one goal:nine shots) has stood the test of time, but for weighting the value of corners I used my own research. It is conventional wisdom that corners lead to goals, but as Moneyball and Soccernomics both showed, conventional wisdom isn't always right.

A fascinating post from Soccer By The Numbers concludes that the value of corners is overrated. "Turns out that more corners don't equal more goals".

I use corners as a factor in rating weighting (very poetic) not as a tool for predicting goals, but it did occur to me that perhaps the in-play markets overreact to the award of a corner. If the finding that the "yield from corners ranges from 0 to .07" in the EPL is correct, and the blog is meticulously written, so I have every confidence in its findings, when the price in a market drops by three ticks for a corner, there should be value in opposing this move. Corners are not the goal scoring opportunities they are generally considered to be, but still strike terror into the hearts of traders hoping the status quo will be maintained, and gladden the hearts of traders betting on a change.

No comments: