45% Chance Top 2 and 85% in Play Offs

All advertisments are hidden for logged in members, why not log in/register?

Kenilworth

Up the Punters ❤️⚪️⚫️⚔️
Joined
Aug 6, 2009
Messages
15,346
Reaction score
18,943
Location
Not Kenilworth
Chap who does stats and posts on Twitter has a predictive model of the range teams will finish in the EFL divs based on expected Goals (xG)

Interestingly it has the Blades at 45% predicted probablility of finishing top 2 and 82% chance of completing the season with a play off position. (Soz about thread title)

Contrast Wendy - with a 7% chance of being in a play off position.



Here's the championship model prediction FYI

IMG_20181021_195827.jpg
 



Ben’s site is great, really interesting data on individual games and the season as a whole, definitely worth a regular look if you’re into that kind of thing.

His model is (quoting) “based on the combined quality of chances that clubs create and allow, rather than their results.” - basically, expected goals. There’s a longer explanation here: https://experimental361.com/explanations/explanations-e-ratings/

Naturally these %s will fluctuate as we progress through the season but overall this essentially shows what lots of us have been saying: we’re up at the top of league because we create better chances than most other teams and deserved to win most of the games we’ve played so far this season. By contrast Wednesday have been picking up points they didn’t really deserve and that’s why his model gives them such a low % chance of making the top six, even though they’re currently only two points off. Ditto Forest.
 
Just to add some meat on the bones of this, i have a similar prediction model and as with any football prediction tool the only data it can be based on is what has happened in past i.e. xG for and against in each game so far. How that in turn is then used to predict pts etc is indiviualistic.

Reason i'm saying this is in the EFL 5 game trends of xG change SIGNIFICANTLY for teams....i.e. a team in form for 8 games and having a good projection loses form for 5 games and then that projection changes massively. That is the EFL in a nutshell.

I have 5 game xG trends for all Championship clubs and approx 80% of clubs have such variation that using that as a tool to predict future points is just impossible, in the PL perhaps absolutley due to the chasm of difference between top 7 and rest. That said it is still good reference info as there isn't much else to predict on so its 50/50 really, just another indicator of how good we have been to this point but i wouldn't personally use the %'s as a reference point (at this stage of season anyway.

Michael Caley who basically introduced xG to football analysis said himself that in general xG as a team league positioning prediction tool is about as good as total shot ratio or shot on target ratio data, i.e. its just an indicator of a team's dominance.

Jeez this got nerdy fast.....feel like it needs blading up somehwat....CMON YOU RED AND WHITE WIZZAAARDDSS
 
So, correct me if I'm wrong :confused:, but what we need is Leon back on form, duplicating last season's form and giving us an even better percentage chance of promotion?

Seriously, being ruthless and efficient in and around the opposition box is where we're falling short.
 
And if mi uncle had a fanny he’d be mi auntie.
It’s all history driven, as GCB says we would have looked good this time last season, a computer can’t foresee injury, lack of form, referee errors etc
If it’s that good surely the owners of this system would bang £100 on a 10fold acca every weekend and be multi millionaires
 
Been following this guy a while on Twitter. He called Barnsley making the Play Offs in the early stages of their run in the 15-16 League 1 season - about January time - and had us as favourites for the League 1 title in 16-17 from November onwards. His model does often get it right - being the most likely to be promoted using his model is not to be sniffed at even if you’re averse to stats.
 



Bet he didn,t get Fulham right this time last year.

He doesn’t publish them every week but in January he had Fulham down as around 50% likely to make the playoffs. At that point they were about 10th in the table.

His model had them playing like the 2nd best team in the league after Wolves, it’s just that all the points that they’d already dropped meant they were more like a 50% shot than say the 85% that he has us at right now: https://experimental361.com/2018/01/21/return-of-the-e-ratings-championship-21-jan-2018/

We were in the top six then so had points in the bag but weren’t playing so well, so we were at about 63% to make the playoffs - not a huge amount above tenth-placed Fulham. If anything that shows how much better we’re playing this year, that’s we’re at a much earlier point in the season but looking like a good bet to make it.
 
Until we lose against stoke and we all of a sudden
“Haven’t got the depth to last the race” and “we’re only 3 losses away from relegation”
 

All advertisments are hidden for logged in members, why not log in/register?

All advertisments are hidden for logged in members, why not log in/register?

Back
Top Bottom