2014 - Round 14 : Lined Up

This week's portfolio of wagers seems much better aligned with the Funds' recent form, so I feel a lot more comfortable with what we have in Round 14 than with what we had last week. Which, oddly, makes me uncomfortable ...

The centre-piece is six Line bets, equal to the most we've had in any single round so far this season and representing 30% of that Fund and so 15% of the entire Portfolio. Four of these wagers are on underdogs receiving between about 2 and 5 goals start, and two are on favourites, one slight and one raging.

These half-dozen Line bets are served with a side order comprising a trio of Head-to-Head wagers - all on underdogs of varying relative abilities and none sized larger than 2.7% - and in a foam of SuperMargin wagers (because there's just about nothing as insubstantial as a SuperMargin wager). As I write this, the TAB Bookmaker has yet to post a SuperMargin market for the Fremantle v Brisbane game, so for the purposes of posting this blog tonight I've assumed we'll secure $8 prices for both of the bets we're proposing. I'll provide an update once we've secured a price, assuming the TAB ultimately offers one. (EDIT: They have and I was optimistic. We got $7 for both buckets.)

Together, the 15 wagers we have represent 17.9% of the Portfolio, making this the fourth consecutive round in which we've set the record for the largest proportion of Portfolio Funds at risk.

Seven games have the potential to materially change the value of the Recommended Porfolio this week, the exceptions being the Essendon v Adelaide game, on which we've no wagers at all, and the Hawthorn v Collingwood game, on which we've only a pair of largely inconsequential SuperMargin bets that can provide no more than a 1c gain or loss.

(EDIT: See foot of blog for updated Ready Reckoner.)

St Kilda carries Investors' greatest hopes, promising a 3.7c gain should they spring the upset on the Eagles. Other significant gains are possible via upsets by the Suns over the Cats (3.6c), or the Giants over the Blues (2.9c). Fremantle and Port Adelaide could deliver gains of similar magnitude, but would require finer motor skills to deliver them - Fremantle would lift the Portfolio by 3.5c if it were to win by between 56 and 59 points, while Port Adelaide could lift it by 3.3c should it win by between 50 and 59 points. That seems, as I think I've said before, like landing a Jumbo on a cricket pitch.

Six teams carry downside risks of 2.5c or more, chief amongst them the Gold Coast who would destroy 3.3c of value if they were to lose by 14 points or more. Port Adelaide could knock 2.9c off the Portfolio price by winning by 46 points or less, the Lions could do the same damage by winning by 39 or less, GWS could shave off 2.8c by losing by 17 points or more, the Tigers could jettison 2.5c by losing by 30 or more, while the Dees could wipe off the same amount by thoughtlessly losing by 22 points or more. 

Upside and downside is, as you can see, very widely spread this week but lumpy nonetheless. A dream round would add just over 22c to the value of the Recommended Portfolio, while a nightmare round would slough off just under 18c.


We've returned to more traditional, by which I mean minimal, levels of disagreement amongst the Head-to-Head Tipsters this week, with only the Essendon v Adelaide game causing any significant divisions. In that game the Tipsters are split 14-14, though the weight of opinion amongst the better-performed Tipsters, including the TAB, is very slightly with the Crows.

In other games there are eight Tipsters predicting an upset win by the Giants over the Blues, six foreseeing an upset Pies win over the Hawks, four believing that the Dees can go back-to-back for the first time this season, and three touting the prospects of a Dogs upset over Port. In three other games Home Sweet Home is martyred once again by its slavish devotion to the team playing nearer where it sleeps.

Not surprisingly then, Home Sweet Home is the weekend's Tipster Most Different with an extraordinarily high Disagreement Index of 61%. STM I, FTS and EI I are all, jointly, next-most disagreeable, all with Index values of 38%.

The Margin Predictors are unanimous in all but two games this week, sitting either side of the fence in only the Hawthorn v Collingwood and Essendon v Adelaide clashes. That said the average predicted margin in nearer zero in the Suns v Cats and the Giants v Blues games than it is in the Hawks v Pies.

The average margin in the Hawthorn v Collingwood game is elevated by the large margin predictions of, in particular, Combo_NN_1 and Combo_NN_2, who both have the Hawks winning by 5 goals or more.

Bookie_3, which is the extreme Predictor in four contests, has the round's largest Mean Absolute Deviation (MAD). Combo_NN_2 has the next-largest MAD, though it's the extreme Predictor in only two games. ProPred_7, Combo_NN_1 and C_Marg have the third-, fourth- and fifth-highest MADs and are each the extreme Predictor twice also.

The Head-to-Head Probability Predictors are, like the Margin Predictors, unanimous in all but the Hawks v Pies and Dons v Crows contests. For these Predictors too the range of opinions in the Hawks v Pies game is vast, here spanning probability estimates of the Hawks' chances ranging from 36% to 73%. That's comfortably the greatest range of any of the weeks' contests. In fact, for all but two of the games the range is less than 20% points.

WinPred, who climbed into third place on the MatterOfStats Leaderboard last weekend on the back of some brave and differentiated opinions, has provided another set of relatively divergent views this week. It's the extreme Predictor in five games and has the highest MAD of all Predictors. C_Prob is second on this metric though is the extreme Predictor in only three contests. Bookie-RE is only third on the metric despite being the extreme Predictor in four games.

The Line Fund algorithm has rated four teams' chances at 60% or more on Line betting this week: Collingwood (82%), Richmond (69%), Melbourne (65%), and Fremantle (60%).

ChiPS continues to be most influenced by the teams' relative Ratings, selecting as winners the higher-Rated team in all nine contests this week.

The margin of victory that it assigns to Fremantle is the second-highest margin it has predicted all season, behind only its prediction that Hawthorn would beat GWS by 87.4 points in Round 11.

ChiPS recommends four wagers, one each on Richmond, Gold Coast, GWS and St Kilda. Make of that what you will.


Swapping the $7 SuperMargin prices for Fremantle with the $8 I assumed above gives us this profile for that game.