On Sunday the Pies and the Hawks recaptured a significant portion of what the teams before them had collectively surrendered earlier in the weekend, but their combined efforts could only mitigate and not extinguish the overall loss. By Sunday night that loss amounted to a little over 4c, leaving the Recommended Portfolio up by 8.7c on the season.

From a Portfolio perspective, most of that loss was attributable to the Line Fund, whose 2 from 6 performance served to strip 5.5c from the value of the Fund and half that amount from the value of the Portfolio. The Head-to-Head Fund, despite landing the Pies' result on Sunday, still wound up shedding just over 1c on the weekend having been let down by the wayward 10.19 Roos on Saturday, and the Margin Fund, despite correctly predicting the margin range in the Swans v Dons game, was predictably wrong on 11 other occasions and so shed 5c.

In summary, a small but not devastating loss for the Portfolio, and some early confirmation that the timing of the Line Fund bet-size reduction was about right.

TIPS AND PREDICTIONS

Only six favourites collected the competition points this weekend, but this didn't prevent Silhouette, Short Term Memory II and Shadow from landing eight from nine correct predictions, and Combo_NN_2, Consult The Ladder and Easily Impressed II from landing seven.

The all-Tipster average for the week was six, weighed down in particular by scores of only four from Follow The Streak, Short Term Memory I and Easily Impressed I.

The Tipsters at the top of the MoS Leaderboard all scored six from nine, however, which leaves Bookie_9 still two tips clear of Combo_7 and Combo_NN_1, and three tips clear of Bookie_3.

Amongst the Margin Predictors it was a rare week of triumph for C_Marg, which this week recorded the round's best Mean Absolute Prediction Error (MAPE) of just 21.9 points per game. That result was enough to catapult it into second-last position, relegating Combo_NN_2 into last, but with the gap to Combo_7 at the head of the Leaderboard still 327 points, I suspect that C_Marg has left its run for the title a little late.

Four other Margin Predictors recorded sub-25 MAPEs for the round - Win_3 (24.6), H2HU3 (24.7), H2HU10 (24.7) and H2HA3 (24.9) - and the all-Predictor average came in at 26.4 points per game.

Taken across the entire season, the all-Predictor average MAPE is now 29.8 points per game and 11 Predictors have profitable Line betting performances. Seven of them have selected the correct team in at least 58% of the contests this year, while three of them have profitable season-long SuperMargin performances, and eight have profitable season-long SuperMargin performances when wagering only when they predict home wins or draws.

C_Prob lies second amongst the Head-to-Head Probability Predictors despite recording the second-worst probability score for the round, superior only to the performance of ProPred which, on the strength of that performance, now sits last on the ladder. WinPred recorded the round's best performance but still sits in a very comfortable fifth. Bookie-OE remains in first place. C_Prob, I remind you - though, to be honest, mainly to remind myself - uses no bookmaker information in its probability assessments, so for it to be so close to the lead this far into the season is about as close to astonishing as I can find things these days.

The Line Fund algorithm had its fourth-straight round of negative probability scoring, which left it with a negative score for the season as a whole.