# A Welcome Regression

Entitlement's a dangerous feeling to harbour as a punter, as sporting events will always unfold as they wish, oblivious to how earlier outcomes might have treated your finances.

So, I'll not claim for a moment that I felt we were entitled to a series of favourable outcomes this weekend to redress the tragicomedy that was last weekend's wagering. Instead, I'll draw upon the nomenclature of the statistician and suggest that what we saw this weekend was an example of "regression to the mean", which, broadly, is the label attached to the phenomenon of witnessing a 'more normal' outcome directly after having witnessed a more extreme one.

It's a phenomenon with a proud history and one that we're all exposed to regularly.

For example, a shambolic experience at a restaurant is more likely to be followed by a more pleasurable one - if only you can bring yourself to return to the same restaurant. And, in a sporting context, it's how you explain the inevitable decline in performance of the previous year's Rookie of the Year.

The logic of regression to the mean is very straightforward. Each time you experience some repeatable event you experience just one of a number of possible "outcomes" for that event. If you could array the set of these possible outcomes you'd find that they tended to follow a vaguely bell-shaped distribution - that is, there'd be a few very bad (or very good outcomes), a few more less bad outcomes, and a far greater number clustered around "average". So, if you suffer a diabolical outcome this time, next time there are a far larger number of better outcomes for you to experience than there are worse outcomes, so you tend to experience one of the better ones.

Our win-every-bet-and-jump-12% performance this weekend was then potentially a particularly gratifying example of regression to the mean. Or, of course, it was just an example of extraordinarily good fortune to which we were completely unentitled.

Anyway, the summary is that the New Heritage Fund landed 3 from 3 to jump by over 13%; Prudence bagged 4 from 4 to jump by 4%; Chi-squared notched 3 from 3 to jump by 22%; and Line Redux landed 2 from 2 to jump by 23%.

Together, these results lifted the value of the Recommended Portfolio by about 12%, moving it back into profit by about 10% on the season. Other Portfolios rose by similar percentages on the weekend, leaving them up by between 13% and 25% for the entire season.

Here's the game-by-game picture for each Portfolio for the season so far:

MAFL Tipping was as accurate as MAFL Wagering, with BKB, Chi and ELO all tipping 4 from 4. ELO also tipped all 4 line betting results and is now at +11.09 units on level-stake line betting based on its tips, which is about a 6.2% ROI.

Week 2 of the Finals looks like this:

Comment