Matter of Stats

View Original

2013 : Round 6 - Wagers & Tips

This week Investors enter a new phase of the competition as the first of the Funds changes its wagering strategy. Based on empirical data, the Line Fund performs better in Rounds 6 to 11 than it does in the first 5 rounds of the season, so this week it doubles its standard wager to 2.5% of the Fund. Next week it'll be the Head-to-Head Fund's turn to ratchet up the risk and reward.

The Line Fund's responded to its longer leash by stretching it about as far as it will go: it's wagered in all but two of this week's games, shunning only the Roos giving Port 18.5 points start at Bellerive, and Adelaide receiving 22.5 points start playing the Hawks at Football Park.

Keeping its proverbial powder dry for when its turn comes to frolic off the leash, the Head-to-Head Fund has ventured just three wagers this week, all of them on teams priced at over $3 and none of them for more than 1% of the Fund. Its probability assessments in the other six contests were irrelevant as the home teams in all of these games were priced at less than $1.50 making them automatic 'no bet' games for this Fund.

No such restraint was required or expected of the Margin Fund, which is in the portion of the season where it bets most, a liberty that it's been hard to argue against given the Fund's performance so far. It's found six games that it fancies this week sufficient to wager on. Amongst its wagers are two, both courtesy of the partial advice it takes from Combo_NN2, priced some distance from our usual $7 or $7.50 fare. 

There's a wager on the Dons at $13 to win by 30 to 39 points - in a game where they're giving 80.5 points start, I'd point out -  and another on the Blues to win by 20 to 29 points, priced at $11. The Bookie_9-inspired wagers in these same games are on other than adjacent margin ranges, which has produced yet more odd-looking profit profiles for these games, as well as for some others, this week on the Ready Reckoner.

The combination of the Line Fund's 60% weighting in the Recommended Portfolio, its increased bet size, and its high level of activity is apparent in the range of these charts this week.

A surprise Tigers win is the best possible result of the weekend and would lift the Recommended Portfolio by 1.8c, though a Collingwood win by 31 to 49 points, a Sydney win by 45 to 59 points, or an Eagles win by 49 to 59 points would each be almost as good, ratcheting up the Recommended Portfolio by 1.7c.

The Richmond v Geelong game while representing the largest potential upside also represents the greatest potential downside for Investors, threatening a reduction in the price of the Recommended Portfolio of 1.7c should the Tigers lose by 19 points or more. Six other teams - Collingwood, Essendon, Gold Coast, Sydney, Carlton, and West Coast - could all knock 1.6c off the Recommended Portfolio price should they contrive to produce worst-possible outcomes.

In summary then, there's a lot more riding on each game this week, and consequently a lot more riding on the round.

TIPSTERS AND PREDICTORS

The Gold Coast v Fremantle game sees the greatest level of disagreement amongst the Head-to-Head Tipsters this week. They're split 8-5 in favour of the home team underdogs, the Gold Coast. In two other games they're split 9-4: in favour of the home team favourites in the shape of the Pies, and in favour of the away team underdogs in the shape of Port Adelaide.

As usual, the Head-to-Head Tipster Most Different is Home Sweet Home, which this week rates 41% on the disagreement index (the likelihood of it disagreeing with a randomly-selected Tipster's tip in a randomly-selected game).

Unlike the Head-to-Head Tipsters, the Margin Predictors are unanimously behind the favourites this week, disagreeing only in the margin by which they expect the favourites will win. Even this metric shows only minor levels of disagreement except in the Essendon v GWS game where the predicted margins range from 34 to 77 points, and in the Carlton v Melbourne game where they range from 26 to 69 points. Bookie_3's predictions are most different from the all-Predictor average this week, while ProPred_3's are the least different.

Meantime, dissention amongst the Head-to-Head Probability Predictors could best be described as moderate at most this week. The Adelaide v Hawthorn game registers the highest level of debate, with the Risk-Equalising variant of the TAB Bookmaker-derived Predictor rating the Crows as only 19% chances while WinPred assesses them at 42%. The span of the range of probability assessments is also in double digits for the Roos v Port, Tigers v Cats, Suns v Freo and Blues v Dees matchups. 

All told though, the mean absolute difference between individual Predictor's assessments and the average assessment for each game is under 4% points. The lowest mean absolute difference belongs to the Overround-Equalising variant of the TAB Bookmaker-derived Predictor (which surely deserves a shorter description, if only I could find one that didn't require me to explain it parenthetically immediately after using it each time) at 2.9% points, and the largest belongs to WinPred at 5.3% points.