2017 - Round 16 : Still in the Fog

/

Normally, at this point in the season, some semblance of an order emerges and you feel as though most games in most rounds will adhere to a familiar script. Sure there'll be the odd fluffed line and lighting mishap, but the play will be performed largely as written. This year, it's all improv, all the time.

Still, the MoS Head-to-Head Tipsters are mostly in agreement, though there are three games in which there are three or four dissenting Tipsters in the minority.

In the Collingwood v Essendon game, six Tipsters have lined up behind the underdog Pies, whilst in the West Coast v Port Adelaide matchup, three have tipped the underdog Power. And, in the Saints v Tigers game, their are four dissenting voices tipping the underdog Saints.

Thanks largely to Home Sweet Home (who is now, I will remind you, only two tips behind BKB), the All-Tipster Disagreement Index for the round is 27%, which is the highest it's been since Round 13 and, before that, Round 3.

The Margin Predictors' average mean absolute deviation (MAD) of 3.7 points per Predictor per game is quite low - the lowest since Round 12, and also slightly below the all-season average of 4.2 points per Predictor per game.

Driving up that average are C_Marg (6.4) and MoSHBODS_Marg (6.2) and, to a lesser extent, Bookie_3 (4.5) and MoSSBODS_Marg (4.1). That's pretty much in keeping with their season-long MAD proclivities.

From a game-by-game perspective, it's the Roos v Dockers game that has the highest MAD (5.7 points), ahead of the Eagles v Power (4.3), and Saints v Tigers (4.1) games.  But, in the last two of those matchups, the range of forecasts spans no more than about 15 points, which is about the case for every game in this round.

For the Head-to-Head Probability Predictors it's the Roos v Dockers (7.3% points), and Saints v Tigers (6.9%) games that have sparked the broadest disagreement, the range of probability estimates in both of those games spanning about a 16% point range.

MoSHBODS_Prob and C_Prob have the round's most divergent probability estimates and both have MADs of just over 5% points per game.

WAGERS

This week it's the Head-to-Head Fund that is significantly more active than the Line Fund, its six wagers representing a little over 15% of the original Fund (the highest since Round 11) compared to the Line Fund's four wagers representing just under 5% of that Fund.

The week's upside belongs mostly to the Kangaroos (+1.8%), the Brisbane Lions (+1.7%), and St Kilda (+1.4%), whilst the downside belongs mostly to the Kangaroos (-2.3%) and St Kilda (-1.1%).

In total, about 5% of the Overall Portfolio is at risk, and a best-case set of results is capable of adding just over 6% to that Portfolio.

I thought it might be interesting this week to review the performances of the three MoS Funds so far this season, in particular looking at how they've fared for different levels of estimated overlay.

On the left are the results for the Head-to-Head Fund, which shows that it has made most of its return in games where the estimated overlay was over 1.22. There have been 18 such games, which have resulted in bets averaging 3.2% of the Fund and providing a return equal to 38% of the original Fund. (Note that there have been 5 bets where the overlay was under 1.05. These are bets that were placed in error due to an Excel formula glitch. Fortunately, we've roughly broken even on these.)

The only range in which a net loss has been recorded is for overlays in the 1.09 to 1.14 range where just 4 of 17 bets have been successful and the loss has been 17% of original Funds.

In the middle are the results for the Line Fund. It too has made most of its return on wagers where the estimated overlay has been largest - here overlays of more than 15 points, which has been the case in 17 games, with 12 of them being eventual winners worth, in total, 25% of original Funds.

Note that the overlays shown here are relative to adjusted handicaps in games where the line market was not at even money. To date, in deciding whether or not the 8 point minimum threshold has been reached, I've measured relative to unadjusted handicaps, which is why there have been a handful of wagers made at overlays of slightly under 8 points as shown here. These wagers have roughly broken even.

Again we find just one range of overlays for which a loss has been recorded, here 10 to 12 point overlays where just 6 of 13 wagers have been winners and the overall loss has been 3% of original Funds.

Lastly, on the right we have the results for the Overs/Unders Fund. Unlike the other two Funds, this one has made most of its money on smaller overlays (6 to 8 points) though almost as much on slightly larger overlays (10 to 14 points).

The only range for which it has a less than 50% win record is 8 to 10 points where it's landed just 7 of 15 wagers.

MoSSBODS AND MoSHBODS

Here are the MoS twins' views on team and total scoring for the week.

Commentary to follow.