Search MAFLOnline
Subscribe to MAFL Online


Contact Me

I can be contacted via


Latest Information


Latest Posts
« 2011 Round 1 Results: Everything Went Better Than Expected | Main | And So It Begins Anew »

2011 Round 1: Anything But Quiet

The Head-to-Head and Line Funds, bereft of any relevant form data, have peered through the murk that shrouds Round 1 wagering activity and spotted a swag of teams to love. Well, to feel vaguely affectionate about for this week anyway.

Investors have head-to-head wagers in six contests and line wagers in seven. (Actually, we'd have had wagers in all eight line markets if the TAB accepted 50c wagers, such was the Line Fund's lukewarm enthusiasm for a tilt at Sydney +8.5 playing Melbourne).

Here's the detail (click on this image and any of those that follow for a larger version):

The first row of the table is for Thursday's Carlton versus Richmond clash. Carlton, though the AFL-designated Home team for the contest, has played at the MCG only 8 times in the past 12 months, while Richmond's played there a dozen times, which is why there's a minus 4 Venue Experience figure for this game. We've a 12.7% head-to-head bet on Carlton at $1.35 and a 5% line bet on them at $1.90 giving 19.5 start.

Our 8% Collingwood wager is another head-to-head wager that's significant in terms of size, but at $1.06 it's not a bet that's brimming with upside. For that you'd need look at the 5.6% wager on the Lions at $2.25 or the 3.1% wager on the Dons at $2.35.

Looking next at our line bets, the two biggies are the one we've already considered on Carlton, and another that's almost as large, 4.7% on Collingwood giving 47.5 points start. In all, we're receiving start for five of our line bets and giving it in only two.

Incidentally, the reintroduction of starts smaller than 6.5 points - with Adelaide receiving 4.5 points start and being priced at $1.90 - did come as a surprise to me. In recent seasons past the usual behaviour has been for the TAB bookmaker to defer posting a line market for close games until nearer game time, then post it at 6.5 points but with a price something other than $1.90. It'll be interesting to see if this approach, which I much prefer, persists into subsequent rounds.

Here's the Investor TL;DR Table for this week:

So we know which teams we'd like to win, but which teams do we think will win? (If you think these two sets should contain the same teams, please read some of my early newsletters explaining why it makes perfect sense to wager on teams that you think are less likely to win than to lose.)

(For details on how the heuristic tipsters have been altered to cope with the introduction of the bye, see this blog.)

The heuristic-based tipsters from Shadow to Follow the Streak are obliged to concur with Consult the Ladder in Round 1, so there's not a lot to be made from the uniformity of their opinions this week. ProPred and WinPred though, whose tips are shown in the rightmost columns, do have the putative benefit of freewill, and their tips disagree with BKB in a number of games. ProPred fancies the underdog Cats over the Saints, the Lions over Freo, Sydney over Melbourne, and the Roos over the Eagles; WinPred also likes the Cats and the Lions, and also tips the Crows to topple the Hawks.

Across all the head-to-head tipsters there's unanimity only for the Carlton v Richmond, and the Collingwood v Port Adelaide clashes.

These are also the only games on which the Margin Tipsters all agree. On average, they predict the Blues to win by about 5 goals and the Pies to win by over 50 points. Here's the detail:

(For details on the development of the margin tipsters, see this blog and this one. Since these two blogs were written I've decided to add another neural network-based margin tipster, CN2, which uses as inputs only the Prices and MARS Ratings of each team and whether or not the game is an Interstate clash.)

Looking across all 13 tipsters, Carlton is predicted to win by anywhere between 14 and 41 points, which is quite a range and which explains why the standard deviation of the predicted margin is a relatively high 8.6 points. Eleven of the thirteen tipsters predict that the Blues will cover the spread.

The Geelong v St Kilda game causes much angst amongst these tipsters, with nine of them predicting margins of around 1 point. The largest predicted margin in absolute terms is a St Kilda win by 8 points, which is the prediction of CN1, one of our two neural networks. With St Kilda giving 6.5 points start in this game, this makes CN1 the only tipster predicting that the Saints will cover.

For the Collingwood v Port Adelaide game, tipster debate revolves around the margin of Collingwood's victory but not around its likelihood. W7, the model based solely on WinPred's probability estimates, is at the low end in foretelling a 33 point Pies victory. Somewhat oddly, W3, the other model based solely on WinPred's probability estimates, is at the high end, predicting a 71 point victory for the Pies. Across all tipsters the variability is such that this game sports the largest standard deviation in predicted margins at almost 11 points per tipster. Nine of the tipsters see the Pies covering the 47.5 point spread.

Most tipsters see the Adelaide v Hawthorn clash as a close-run thing, with one side or the other prevailing by less than a goal. B9 had the most difficulty deciding before eventually coming down on the side of the Crows though if it were a betting algorithm I'd wager it'd be on the draw. 

The two outliers for this clash are W3, which has the Crows winning by 12, and CN1, which has the Hawks winning by 31. These two predictions contribute substantially to the overall variability in the predicted margins for this game, which is the second largest of all eight. Only two tipsters foresee the Hawks covering the 4.5 point spread.

Brisbane v Fremantle is another game that's predicted to be a close one by most tipsters, with all but CN1 predicting a single-digit margin. W3 and HA7 have the largest predicted victory margin for the Lions at 8 points, while CN1 is again the outlier for the Away team, predicting Freo to win by 26 points. Only CN1 and B3 see Fremantle winning by enough to cover the 7.5 point spread.

Yet another game generally expected to be close is Essendon v the Dogs for which 10 of the 13 tipsters predicts a single-digit margin of victory. If Essendon win, it'll be particularly narrowly, as their greatest tipped victory margin is just 3 points. The tipster most confident about the Dogs' chances is CN2, which has them winning by 20 points. Only three tipsters predict the Dogs will cover the 10.5 points spread.

All 13 tipsters are predicting single-digit victories in the Melbourne v Sydney matchup, and 10 of them have the margin at a goal or less. B3 and B9 have the largest Melbourne margins at 9 points, making them the only two tipsters to predict that Melbourne will cover the 8.5 point spread, while CN2 has the largest predicted Sydney victory margin at just 1 point.

To finish the round we've one further game for which the majority of tipsters are predicting narrow results. W3 is predicting the largest Eagles victory at 9 points, which makes it the only tipster predicting that the Eagles will cover the 8.5 point spread, while CN1 is yet again an outlier, this time in predicting the Roos to win by 11.

Overall, this is a round in which the margin tipsters favour the teams receiving start, with only Carlton and Collingwood expected by the majority to cover their respective spreads.

Lastly, as another way to look at the confidence with which various of our tipsters predict one team or other to win, here's a diagram showing the range of probabilities that the TAB, ProPred, WinPred and the (unadjusted) H2H probability predictor attach to the Home team's chances in each game, and the range of probabilities that the Line Fund algorithm attaches to a Home team victory on line betting for each game.

In all but the last two games of the round, ProPred, WinPred and H2H generally all think that the TAB has underestimated the Home team's chances, which is partly the reason we have head-to-head wagers on the first six games of the round.

The Line Fund is most confident about the Home team's chances in the line market in the 1st and 3rd games, and most confident about the Away team's chances in the final game of the round. This is why our largest line bets are on these three contests.

Reader Comments

There are no comments for this journal entry. To create a new comment, use the form below.

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>