Just a quick post to answer a question raised in the comments section of an earlier blog on the team-by-team performances of the MoS Tipsters and Predictors. There the requester asked me to undertake another team-by-team analysis, here to summarise each team's performance on the Line market this season.
The results appear in the table below.
Adelaide, it turns out, have the best record on the TAB Line market this season, covering the spread in 7 of the 10 games in which it has given start, and defending the start it's been given in 5 of the 8 games where it has started as the underdog. Overall the Crows have won on the Line market two-thirds of the time, which is well above the chance 50% level.
The margins in their games, taken from their point of view, have been 14.4 points per game in their favour. So, for example, in games where they've given 30.5 point start, say, we'd expect them to win by almost 45 points.
In absolute terms, the games in which Adelaide has been involved have finished with margins about 23 points different from that implied by the handicap.
(Note that, as the footnote to the table explains, I've made adjustments to the handicap in those markets where the line was no longer being offered at a $1.90 price. This adjustment serves to make the line equivalent to what it would need to be to be priced at $1.90.)
Scanning the entire table reveals a few other things of note:
- Teams receiving start have fared slightly better than those giving it (51% v 49%)
- Melbourne have been successful when receiving start 80% of the time, but successful only 38% of the time when giving it
- The Roos have been successful when giving start 62% of the time, but successful only 20% of the time when receiving it
- Richmond have fared worst on the Line market generally, winning only 22% of the time
- Only 6 teams have a worse-than-50% record, while 10 have a better-than-50% record
- The TAB bookmaker has done best in estimating margins in Western Bulldogs games (mean absolute error 18.4 points per game) and worst in estimating margins in Collingwood games (mean absolute error 40.2 points per game)
One word of caution though before I finish: the variability in team-by-team performances we see here is almost certainly due to nothing other than chance. Bookmakers don't last if they persistently under- or over-estimate team abilities.