2015 - Round 15 : Blank Lines

In a week where seven home team are giving start and the two others aren't receiving much, the Line Fund has opted to keep its wallet in its pocket, leaving the Head-to-Head Fund's four forays as the round's only activity for Investors.

None of the four wagers threatens to knock much more than 1c off the already substantially diminished Overall Portfolio, and none offers to add much more than that amount either.

The Roos wager, narrowly, carries the round's greatest upside and downside, while the Dogs bet carries the smallest upside and the Lions bet the smallest downside.

Five of the week's games are entirely wager-free and, in total, only 3.4% of the Overall Portfolio is at risk. Favourable results in the four games that do carry bets would add 3.2c to the Portfolio - a very small first step in a far more significant climb back to profitability. (But I'd take it.)


This week sees historically high levels of disagreement amongst the Head-to-Head Tipsters, the overall Index of 29% the second-highest for a single round this season.

That Index has been driven up by the fact that five of the week's contests involve a minority group of seven Tipsters or more. These renegade groups are largest in the two games in which the TAB Bookmaker has installed equal-favourites, those games seeing 12 Tipsters forming the holdout group as well as a draw-tipping Bookie_3.

Keeping the Index constrained though are the two unanimous picks (the Tigers and the Eagles) and the game in which only Home Sweet Home stands between the Swans and similar unanimity.

Home Sweet Home is not, however, the Predictor Most Extreme this week, that title belonging instead to Easily Impressed II whose Disagreement Index of 49% is amongst the highest recorded by any Tipster this season. Three other Tipsters also have higher Indexes than Home Sweet Home: Easily Impressed I registering 43%, and Short Term Memory II and Shadow each coming in at 40%.

Even the Tipsters with the lowest levels of disagreement have produced Indexes of 21%. What's good about these generally high levels of disagreement is that they offer the promise of some movement at the top of the MoS Leaderboard, which has been very rare of late. More specifically, amongst the four equal-leaders, only Combo_7 and Bookie_9 have made identical selections.


The Margin Predictors are also exploring historical extremes in the current round, though in their case these extremes are lows. The-all Predictor Mean Absolute Deviation average of 4.9 points per game per Predictor is, in fact, the lowest recorded for a single round this season.

Combo_NN2 has the week's highest MAD of 11.3 points per game, with C_Marg (7.5) and ENS_Linear (6.7) registering the two next-highest results. ENS_Linear's high MAD is largely a result of its prediction of a 58 point win by the Swans, that prediction the first this season that has differed substantially from ENS_Greedy's.

The week's low MAD belongs to Bookie_9, which lurks ominously in 2nd right now behind C_Marg on the MoS Leaderboard.

Taking a game-by-game view we see that the Eagles v Crows game is the only one for which the MAD has reached double figures, the 11.3 points per Predictor in that game almost 4 points higher than the next-highest MAD of 7.5 points per Predictor for the Lions v Swans game.

The Dogs v Suns game has provided the astonishingly low MAD of just 2.2 points per Predictor, and the Giants v Saints game has provided the equally astonishing prediction range of just 11 points.

Low MADs don't mean certain winners, however, and we see five contests with average predicted margins of less than two goals, and only one game with an average predicted margin greater then four goals.


Disagreement amongst the Head-to-Head Probability Predictors is also at relatively low levels this week, the all-Predictor average of 4.6% points per game per Predictor the fourth-lowest of the year.

C_Prob is the only Predictor substantially above that average, its 6.5% per game result the highest of all Predictors for the second week in a row and for the third time in four weeks. Bookie-LPSO has the round's lowest MAD of 3.8% per game, that result only fractionally below Bookie-OE's MAD, which differs only in the 4th decimal place.

As for the Margin Predictors, it's the Eagles v Crows game generating the highest levels of disagreement, this the only game in which the all-Predictor MAD has reached double figures. The smallest MAD, of just 2.3% per Predictor, has been attained in the Lions v Swans game for which the prediction range is also a very narrow 9% points.

Only two home teams have been assessed by the Line Fund algorithm as better than even-money chances this week, neither of them by an amount sufficiently above the 50% mark to warrant a wager. The week's best chances belong to the Pies (63%), the Suns (56%), and the Saints (56%).


Even though the magnitudes of the various disagreement measures may be at their extremes this week, you can see from the table below that the general patterns of who's divergent and who's not is broadly consistent with history.