MAFL 2010 : Round 17 Results

Once the Roos had lost narrowly to the Dons on Saturday, Investors with the Recommended Portfolio needed the remaining four results to go their way to finish the weekend in the black. Carlton's second-half rejuvenation a little later on Saturday put paid to any such hopes so that, despite a perfect set of outcomes on Sunday, the Portfolio still dropped in value by a little under 1.4% across the weekend.

That leaves the Recommended Portfolio now down a tick over 10% on the season. I can't help but feel that it's going to take a well-backed long shot winner to drag this Portfolio into profitability for the season. We had such a chance this week with Port's upset victory over the Crows but unfortunately none of the head-to-head wagering Funds foresaw this result.

The Portfolio's of MIN#002 and MIN#017 fared a little worse than the Recommended Portfolio. MIN#002's Portfolio declined by about 4.3%, which has it now down almost 20% on the season, and MIN#017's Portfolio did what it's done now for 8 of the 12 rounds it's been active: it fell, this week by about 5.7% to leave it down 58% on the season.

Here's the latest Dashboard.

No tipster truly excelled on tipping this week, with the best return being only 5.5 from (effectively) 7.5 recorded by Easily Impressed II and Home Sweet Home. Curiously, these two tipsters now sit first and last on the MAFL Tipping table.

Second place on that table is shared by Easily Impressed I, Short Term Memory II and ELO, which are all now on 89 from 136 (65%). (Whilst the draw in the Saints v Hawks game was bad for Investors, at least it's ridded us of that pesky 0.5 in the tipping scores.) The bookies' representative, BKB, is now equal 6th on the MAFL Tipping ladder, four tips off the lead and looks to be in genuine danger of finishing the home-and-away season somewhere other than at the top.

BKB's Mean Absolute Prediction Error (MAPE) rose by almost one-half point this week, crunched as it was by the four games with victory margins in excess of 60 points. These blowouts increased the MAPE's of Chi, HAMP and LAMP by similar amounts, but had far less of an effect on ELO's, which rose by just 0.16 points.

This moved ELO's MAPE to 29.84 points per game and allowed it to grab 3rd place on this metric, nudging HAMP into 4th. ELO still has some distance to go though to bridge the gap to LAMP, which is 2nd on this metric on 29.51, and to BKB, which is 1st on 28.91.

Median APEs, by their nature are generally less negatively effected by blowout results, and this proved to be the case for most MAFL Margin Tipper's Median APEs this weekend. In fact, ELO's exceptionally fine margin tipping performance this week, which saw it record five games with sub-25 point absolute errors (and correctly predict six line results), allowed it to reduce its Median APE by half a point, lifting it into clear second and relegating HAMP into 3rd.

LAMP, which saw its Median APE increase by half a point this week, still lies in 1st on this metric. Its score is now 25.5 points but its lead over ELO has been halved to just one point.

MARS Rating ladder positions were again heavily traded this week, eleven teams finding themselves at the end of the round somewhere other than where they'd been at the start.

Carlton recorded the largest climb, jumping three places into 6th, while Melbourne was the only other team to register a multi-spot rise, climbing two places into 10th after belting the Swans. Essendon and Port Adelaide both enjoyed one rung climbs.

Seven other teams fell, six of them (Collingwood, Fremantle, Sydney, Adelaide, the Roos and West Coast) by just one place and the other - Brisbane - by two places. Even with this preponderance of falls over rises, nine teams still sport MARS Ratings above the 1,000 average.

HELP recorded 5 from 8 on line betting this week, its best single round performance since Round 13 and its equal-best performance of the season. It actually outperformed the Super Smart Model (SSM) on line tipping this weekend, as SSM managed just 3 correct line predictions despite registering a Mean APE marginally superior to BKB's.

The Bookie Probability Model's sole wager this weekend would have been on Port Adelaide, which turned out to be a lucrative selection at $3.75. That leaves the Bookie Probability Model with a perfect 3-for-3 wagering record since I started publishing its predictions in Round 16.

Tomorrow I'll work through the implications of this weekend's results for each team's finals aspirations and spoon fears.