I sometimes feel that I'd benefit from putting a little more time between the end of the weekend's last game and composing the related Results blog - but then, who'd want to read about the Round on the following Tuesday or Wednesday?

So, without any benefit at all of hindsight, let me instead provide you with my assessment of the weekend's wagering, which I'd summarise as, on balance, encouraging. Objectively, I acknowledge, it finished as a net loss to the Recommended Portfolio of just under 1c, the magnitude of that loss significantly reduced by Port Adelaide's Sunday afternoon endeavours, but with two of the SuperMargin wagers finishing within a kick of glory - in the Suns v Giants game, in fact, within a point - and with the Line Fund snagging two from three, I find myself buoyed rather than despairing. Call me glass half full for once ...

In any case and in summary, the Head-to-Head Fund landed 1 from 3 wagers during the weekend, gaining 0.6c in so doing, the Line Fund, as mentioned, called correctly in 2 of 3 bets, gaining 2c as a result, while the Margin Fund was the round's only net loser, failing in each of its eight attempts, but daring greatly in the attempt. All told, the Recommended Portfolio shed 0.8c to finish the week down by 3.7c.

TIPS AND PREDICTIONS

The best of MatterOfStats' Head-to-Head Tipsters recorded six correct tips for the round, this result being achieved by the three WinPred-based and the three ProPred-based Tipsters alone. The all-Tipster average was just under five correct tips and the low score was three, a score attributable to Easily Impressed II, Shadow and Short-Term Memory II.

Bookie_9 now leads all Tipsters outright on the Leaderboard with a 36 from 54 (67%) record, just a lone tip ahead of the ever-threatening BKB as well as Bookie_3, Combo_7 and Combo_NN_1.

Combo_7 is also placed highly amongst the Margin Predictors. In fact, it leads them all, with a season-long Mean Absolute Prediction Error (MAPE) of 32.5 points per game. The two RSMP Predictors fill the next two places, both less than about 3 goals behind Combo_7.

Combo_NN_2 recorded the Round's best average MAPE at 18.9 points per game, a mere 0.3 points per game ahead of C_Marg which, despite its stellar performance, remains resolutely and depressingly in last place. (Incidentally, I discovered that C_Marg had been being excluded from nearest-to and furthest-from the pin calculations, an error which you'll see I've corrected this week. It's interesting to note that the Predictors currently in 4th and in last positions are both most notable for the frequency with which they've been the nearest and the furthest Predictors from the actual result. C_Marg's problem is that, when it's bad it's very, very bad. That said, it has the best line betting record of any of the Margin Predictors, having picked the winner 61% of the time.)

Less than six goals in predictive accuracy now separate the first 13 Margin Predictors, and seven Predictors have line betting accuracy rates sufficient to generate profits, five of them from the 10 best Predictors in terms of MAPE, and two of them from the four worst Predictors on that same metric.

Bookie - OE leads the Head-to-Head Probability Predictors, with the two other most directly TAB Bookmaker-derived Probability Predictors rounding out the top three. C_Prob continues to loiter close to the podium, occasionally stepping onto the dais for a game or two before relinquishing the honour. The Line Fund algorithm, having returned its fourth positive probability score for an entire round this season, now also shows a positive probability score for the season as a whole.

SUPERMARGIN PERFORMANCES

Nine of the Margin Predictors selected the correct SuperMargin bucket in two of this week's games, but Combo_NN_1, which was one of the nine, remains as the only Predictor with a profitable SuperMargin record for the season as a whole.

Combo_NN_2 was, perhaps, the round's most unfortunate Predictor, selecting the correct bucket in just one game, but missing by a single bucket in four other games.