EPL winner 2017/18 - Statistical Analysis

comment by DeltaHedge (U8662)
posted 1 minute ago
comment by Darkphoenix (U11503)
posted 56 seconds ago
Being a Maths teacher and spending most of my time staring at data, making it interpret the story I need it to be.. I am intrigued, how right are they usually?
----------------------------------------------------------------------
I agree, any statistical model is just prediction. For e.g., the same website said Trump won't even get nominated for Republicans, we all know how that prediction turned out to be
----------------------------------------------------------------------
To be fair, if you ask Republican politicians now, I'm sure they won't nominate Trump anymore

My hobby is AI optimisation algorithms, borne from finance sector. Even with data set of of 90,000 matches its very difficult to create any statically significant optimisation or perdition.

Comment Deleted by Site Moderator

Rubbish.

comment by Posh Mufc Great Hafi Not Arrogant Just Better (U6578)
posted 1 minute ago
Rubbish.
----------------------------------------------------------------------

All I needed to hear.

Hafi saying rubbish to this is like when Lance Henriksen says this in Aliens..

"Doesn't surprise me. A2 models always were a bit twitchy"

I'm surprised at Hafi's reaction; take a look at this part of the methodology:

'Shot-based expected goals are an estimate of how many goals a team “should” have scored given the shots they took in that match'

He's just annoyed they've stolen his big idea.

We are 3 games in and I won't be suprised if that same table last season had us finish top or be 2nd after 3 games but we finished 6th.

https://fivethirtyeight.com/features/whos-likely-to-win-the-five-big-european-soccer-leagues/
Premier League
CHANCE OF …
TEAM MAKING CHAMP. LG. WINNING PREMIER LG.
Chelsea 96% 62%
Liverpool 79 15
Arsenal 71 9
Tottenham 70 9
Man. City 54 4
Man. United 29 <1

Teams with a greater than 5 percent chance of qualifying for Champions League
....
This was on January 20th for last season. Liverpool finished 4th instead of 2nd, Arsenal finished 5th instead of 3rd, Spurs finished 2nd instead of 4th and Man City finished 3rd instead of 5th and Man Utd finished 6th like they predicted at January so hope this prediction after 3 games comes true of 1st

This is a running prediction and is updated after each game as more data is available. But everyone probably is sure of who will PL half way through the season barring catastrophic failure by that team.