The daily 'points required' thread....

[LlcoolJ]

Correct me if I'm wrong but I think that "expected goals" model got things pretty much spot on over the later stages of last season?

 

[Moshe Gariani]

Not really 45pts from 21 games = 2.14pts per game x 46 = 98 pts, and I think we'd be safe with that

Of course. My point is that we might need another 45 points from our remaining 25 games and that is essentially form usually required by automatic promotion chasers.

 

[Trevor]

It's based on the 'expected goals rating' that the author uses for his modelling. He explains it on his site but in brief, it is a combination of the quantity, type and quality of shots a team tends to take, and how often those types of shots tends to go in, in general. So it's prioritising 1-on-1 attempts against the keeper much more than 30 yard speculative shots, penalties have the highest expected goal rating, but if your team has a lot of it's goals from penalties (like Reading) then actually the model will assume that can't be sustainable. All of that is aligned against the number of shots that opponents tend to get against your team and how often they score those shots, plus the type of shots, etc etc. Ultimately he ends with a rating for attack and defence and then a blended rating, which he calls the 'e rating'.

The chart is then created by running thousands of match simulations for each team (some sort of monte carlo simulation I guess, if that means anything to you), against their remaining opponents and then plotting the results out to form a 'possible' set of results for every single match. 90% of the time we get top two, according to the latest. It changes of course, but if you go back and look at last year it was much tighter at the top and he wasn't far off.

I like his charts showing attacking and defensive strength - Derby were hilarious after 8 or 9 games. I think they needed around a trillion shots for every goal - he was reporting though that they would come good, as over the course of a season, that many consistent shots are going to start coming good. Conversely, he's been saying that our attacking threat has been dipping quite worryingly as we hasn't been getting many quality chances off, although we were still winning so the difference was just a bit of clinical finishing saving us.

That's really interesting that it is a comprehensive analysis of what has happened so far - and valid for that. So rightly or wrongly some sides may be expected to get stronger or weaker as a result of the transfer window - There is a thought that Newcastle are likely to get even stronger. Also, a side that most people do not have strength in depth might be more likely to perform less well (if injuries kick in for example) - It's good though

 

[GT49er]

Regardless of points needed, if we can win our next two matches, having over 50 points at the half way mark would be a terrific achievement.

 

[FatSuperman]

That's really interesting that it is a comprehensive analysis of what has happened so far - and valid for that. So rightly or wrongly some sides may be expected to get stronger or weaker as a result of the transfer window - There is a thought that Newcastle are likely to get even stronger. Also, a side that most people do not have strength in depth might be more likely to perform less well (if injuries kick in for example) - It's good though

Indeed, I don't think it accounts for much in the way of depth or new signings, because it's about the teams performance. If I recall, it uses previous seasons data as well as current - though of course you'd imagine it yo be less accurate at the start. Also will be a lot of fuzziness for teams test have had wholesale changes ( IE Fulham).

TBH when it's positive about the Albion, I'm all for it. When it's negative I think it's a load of bull

 

[Insel affe]

Regardless of points needed, if we can win our next two matches, having over 50 points at the half way mark would be a terrific achievement.

Next target please

 

[GoldWithFalmer]

Next target please

5 straight league wins...