Copy, this "project" still blows the net winnings (before 5% commission) from 74,249 bets on the 25 Lay bets at WFP less than 1.25, as 24 of these won.
It only needed one of these bigger lay bets (Winx) to get beat at $1.10 to correct that plus some, a lot hinging on one result.
16% their maximum overs seems light, but again not sure if dataset is real.
Although I haven't wasted any time analysing this, I can't get it out of my mind so maybe this post will bring some closure.
First, there is a major blunder in not realising that there is NO minimum Stake for SP, but minimum Liability.
Big, big difference. And what sort of hotshot is anyone who could be so ignorant!
Next, the exercise is completely wrong.
While every punter should understand the Kelly Criterion, I doubt there is any situation in real life where it can be applied directly.
Moreover adding Kelly Fractions within the same contest is mathematically wrong.
Consider a fair regular tetrahedron with implied True Odds of $4.
But you are offered
The Kelly Fractions are.
But staking 10% on both is wrong.
Because obviously this opportunity is equivalent to being offered $3 about 1 face of a fair coin.
Where the Kelly Fraction is 25%
But don't get me started.