Welcome to Part 3 of Brett’s Brownlow Breakdown. In the lead-up to the Brownlow 2020, Brett – the man behind our new Total Brownlow package – will share his key thoughts on betting on the Brownlow medal.
One thing I’ve noticed with all Brownlow Medal previews and predictions is that they all focus on ‘projected votes’. That seems to be what everybody wants to see.
I think they’re actually somewhat misleading.
Yes, they’re a guide, but ultimately what’s important is a price. You need to create that price with your model.
Eyeballing vote projections won’t help you figure out a correct price. They can help you to establish a very simple trading policy, but that’s about it.
For example, in a head-to-head market… your projected vote count might have Player A two votes ahead of Player B.
But what does that mean price-wise? What price should each player be?
So a vote projection is not very useful for establishing whether a bet represents value or not.
Our approach is based on probabilities, just like any other pricing model.
We generate a probability of a player receiving either one, two three or zero votes in every game – they’re the four possibilities.
We do that for all 44 players, in all games.
From there, simulations add all these player performance probabilities together. A player who played 22 games has 22 multinomial distributions for projected vote outcomes. They’re connected together and then it’s all iterated 5,000 times.
That’s the process we go through to come up with what we need: a price.
So our advice here is: put aside the vote projections and predictors for a moment. Knowing exactly how many votes a player will get isn’t the big question.
We’re betting here, so what’s important is the price, and whether that price offers value.
In all the work we’ve done so far on this year’s count, there’s something we’ve noticed:
The ‘Winner without Neale‘ markets are absolutely horrible.
Why? We’re seeing that there’s a glut of players behind the favourite, to the point where we estimate there’s a 93% chance of ties.
So with the dead-heat rules, any price you see should really be half the odds – that’s what you’ll be paid out in the highly-likely event that there’s a tie.