2012 Round 10 Results : Giving A Little Back

This wasn't a good round for Investors.

Had Friday night's game finished about 10 minutes sooner though, it would have been very different, since at that point the Saints were leading and the lead was in the magical 3-9 point range needed to land the Line, the SuperMargin and the Head-to-Head wagers. That result would have turned the 6c loss we eventually suffered into an 8c gain. Even if the Saints had clung on to win by any margin we'd still have made a profit on the contest, albeit a smaller one.

But, we didn't, and only two other results in the round went our way, both of them line victories by the Home team, leaving us down about 6.7c on the round and now up only 4.4c on the season.

Some of our losses were due to the relatively poor performances of favourites this week, which collectively won only 5 of the 9 games. Once again the all-Tipster average mirrored the bookmaker's performance, making for the second-lowest average witnessed so far this season.

Best performed amongst the Head-to-Head Tipsters were Home Sweet Home, Shadow and Silhouette, each scoring 6, while worst performed was Easily Impressed I, which scored only 3. Eleven tipsters still head the Leaderboard, and they're now on 69 from 90 (77%). Surely at some point a leader will emerge to guide us all to tipping glory.

Margin prediction proved diabolically difficult this week, as no Margin Predictor could muster an average below 40 points per game. Best was Combo_7's score of 41.04, and worst H2H_Adjusted_3's score of 46.42. The all-Predictor average of 44.02 points per game was easily the highest average for a completed round this season.

Bookie_9 still heads the Leaderboard, now on 28.72 points per game ahead of Combo_7 on 29.01 and then Bookie_3 on 29.48. Five more Predictors still have sub-30 MAPEs across the season.

The line betting performance of the Margin Predictors fell away too, the all-Predictor average slumping to 4 from 9, the second-lowest average for the season. Combo_NN2 selected 6 from 9 however, dragging it above 50% for the season and reverting us to the position where every Margin Predictor has a season-long average performance above 50%.

Consistent with the general performance-destroying nature of the round, only five Predictors continue to show a profit on SuperMargin wagering had their margin predictions been used for SuperMargin wagering in every game. Just two more Predictors would have spun a profit if only their predictions of Home team victories or draws had been believed and acted upon.

And, to complete the picture of devastation, all the Head-to-Head Probability Predictors and the Line Fund algorithm recorded negative probability scores for the round. Amongst the Head-to-Head Probability Predictors, the TAB bookmaker's results were the least poor, which was sufficient to rocket him to the top of the Leaderboard ahead of WinPred and ProPred that are now each effectively tied for 2nd. They're not called "upset" results for nothing.

Despite the relatively high number of upset victories this week, most games followed the season's trend of the eventual victors leading at the first change. Of the 87 games in which there's been a clear leader at quarter-time this season, 68 have been led by the eventual winner. That was true in 7 of the 8 games that met this description this week (the Cats v Suns game was tied at quarter-time).

The table below provides information about how the leads at quarter-time have translated into wins or losses so far this season.

The first row tells us that there have been 3 games so far this season that have been drawn at quarter-time. I've split these 50-50 between the "Win" and "Loss" columns, denoting the fact that one half of the teams that have been in this position at quarter-time have won, and one half have lost.

The second row tells us that there have been 24 teams that have led by 1 to 5 points at quarter-time and that 13 of them have gone on to win, which represents a winning percentage of 54%. The figure of 78% tells us that teams that have led by 1 to 5 points or more have gone on to win 78% of games this season. Columns on the far right give the same information for all of the games in Season 2011.

One striking feature of this table is that it shows a 98% success rate for teams with leads of just 2 goals or more at quarter-time. This compares with a success rate of only 75% for teams in similar positions last season. 

Here is the data for the positions at the end of the 2nd and 3rd Quarters.

Here too we see the importance of leading this season. Teams with only a one goal lead at half-time have won over 90% of the time, and teams with any lead at all at three-quarter time have gone on to win almost 90% of contests.