OK, so lets assume we can't play out the season....
(Yay.... Shoddy, CWC and SCC lovez me now!)
I have read several articles about using PPG to determine the relative positions.
Most of the stuff I have seen written is seriously flawed for a few reasons.
The prime example is using a formula of PPG home X home games remaining + PPG away X away games remaining
Where this model falls woefully short is that it doesn't account for strength of schedule.
In simple terms you would expect to get more points from Bournemouth at home than Liverpool at home.
The answer is actually fairly simple.
You would use the % distributions from a betting model that line makers/ bookies use to set betting markets.
You would do this for every single game.
So how do we know this model is accurate.
Its simple, we have a large sample of games and if it was way off the bookies would be broke.
(HINT: Its not and they are not broke!)
Another major flaw with the PPG model is it destroys the advantage of goal difference.
It would be extremely unlikely that two teams would finish on the same exact number of points.
Lets say West Ham finish on 35.11 points and Bournemouth finish on 35.12 points.
Even though they basically have the same points and West Ham have a 3 better GD, they get f**ked.
Another flaw is points won are worth more than games in hand. Brighton's 2 points are worth far more than Aston Villa's game in hand.
I wont bore you with the details but some interesting outcomes as the teams are pretty close.
Norwich are clearly f**ked and will finished bottom in every calculation.
Brighton will almost certainly stay up in every scenario.
Their two points are massive.
The rest are pretty dam close with GD being eliminated.
AV are pretty much gone as the game in hand actually counts against them.
I will make one prediction that I guarantee will be 100% correct.
There will be Blubbing........Lots of it.....
We may need Noah's Fucking Ark........
Posted By: usacanary, May 15, 04:17:16
Written & Designed By Ben Graves 1999-2024