Premier League Table 2019/20 (as predicted using big data)
- Thread starter Pogue Mahone
- Start date
How can they do this when they don’t even know the clubs final squads?
No way we concede that many goals if Maguire comes in. Probably no way we score that many either if Pogba and Lukaku are both gone.
So signing AWB and maybe Maguire isn't going to sort our defence out?
Someone save this for seasons end. I’m pretty confident it won’t end like this.
Cant complain really. I fully expect us to have similar season to last. We havent improved with the exception of the RB
Can a loss be a win?
6th "again"
Brutal.
Brutal.
Everyone should post their predicted final table so that we can all laugh at all the wrong people at the end of the season.
Comfortably ahead of everton and wolves.
Get in.
Get in.
Ole out.
1 point between Chelsea, Arsenal and United... not exactly wrong as of now. 6th but basically means nothing since all 3 are the same.
Someone have to bump this thread at season end next year. 
We're going to score more goals than we have at any point since Fergie left. That would be fun as something we've struggled with and while not impossible I don't know what would necessarily lead them to that conclusion as we haven't brought in any top attacking talent. Ole factor? Well from when he took over until season's end we scored 36 in 21 @ 1.71 per game, under Mourinho it was 29 in 17, also @ 1.71 a game. 65 to 72 isn't a huge gap to bridge and there's always variance anyway, just not seeing the signs I suppose.
We will definitely not concede these many goals. On the positive side, 1 point difference between 4th and 6th which could easily change over the course of time.
I would say Leicester, Saints and West Ham will pick up more points and Sheffield won't concede as many but other than that it looks very plausible all things considered.
Are we?
I reckon that we will concede significantly fewer goals than last season; certainly more than the three suggested above. AWB will obviously help, but I'm hoping for better organisation and a regaining of the confidence that Jose destroyed last summer by telling the world our defence was shit. Despite it quite obviously not having been in the previous season when we finished second. I'm not opposed to the signing of Maguire but don't believe he'll make that much difference.
I just don't see where all these extra goals are going to come from. We should score more simply by virtue of playing more expansive football, but with a goal shy midfield it's a hell of a lot of pressure on our young forwards.
According to them, not me. That's why the rest of my post was sceptical about it.
Sure, it can happen, but I don't know why they'd necessarily predict that. After giving it some further thought the only thing I can come up with is that perhaps our Europa League involvement might mean we'll heavily rotate in that competition. If so, and we were to find ourselves 2 or 3 goals up in certain league games ahead of a EL match there's no need to conserve energy for the midweek game as the same players wouldn't be playing so they could continue to go for it.
In general though, not really seeing it as we didn't score more per game under Ole than we did Mou last season and we haven't signed any top forwards either. Don't really get how they've decided that.
Why is everyone expecting Chelsea to do well. They've lost Hazard and replaced him with Pulisic which is a definite step down quite a few levels.
With the added 66 goal difference, just to rub it in.
This has Chelsea topping the table at Christmas. What is it based on? Don't think many would pick Chelsea to be leading the pack at Christmas.
This is utter crap. Don't know why it's a thread. Just even more negative threads to add to an already depressed forum. Sort it out.
That's an interesting point, though Ole's rotation seems to have been all rather random so far, rather than with that sort of plan in mind.
I think I'm expecting a 16/17 type of season (plenty of chances; woeful finishing) than a 17/18 (feeding on scraps but tending to convert them). Quite honestly, I would be happier with the former.
These sort of analyses give Analytics a bad name
Are you being serious?
If the results of an independent, impartial data-based experiment are too negative for you then you might want to sack off watching games next season and instead find somebody who can let you know each time we win so you can follow United without all that annoying negativity.
There's no way the 20th place team will finish with that many points
So you think the results of an independent data based experiment is what we should be basing our thoughts on? Since when has data or stats told a full story?
Under their algorithm it will be to do with Rashford and Martial being a year older I'd guess. Expected goals would drop if we were to sell Lukaku.
Hehe. Sesame Street are even devoting that number to you.With the added 66 goal difference, just to rub it in.
And the letters C and L.
My issue is with their definition of a simulation, which looks to be deterministic (no probability included) or a single point stochastic run (one possible outcome is displayed as their prediction).
A sensible analysis IMO would state what the probability is that United finish 1st or 2nd or 3rd or... and so on for every club.
A sensible analysis IMO would state what the probability is that United finish 1st or 2nd or 3rd or... and so on for every club.
Do they update the data and associated predictions according to changes in squad personnel that happen during the remainder of the transfer window?
Statistics is never the full story. It provides us with a good estimate of the full story though.
You'll only get the full story come May next year. Until then...
There's very little information in the article on their methods. They just say Machine Learning, as if that is self explanatory.
Now I really hate them.
You should base your thoughts on whatever the hell you want, they are your thoughts.
It's your assertion that posting an independent data-driven simulation of the upcoming season is negativity unworthy of discussion that is silly.
Does it 'tell the whole story'? Well no, the season hasn't happened yet. Unless you can invent a time machine and travel a year in the future I think you will be out of luck if you're expecting to learn the story of next season before it's happened.
The lack of transparency in their method is disappointing. Still, the inner workings of models like this tend to include IP valued by companies as being worth million of pounds. Being deliberately obtuse and protective of the actual formulas and metrics that produce outcomes like this mirrors all of my experiences with private sector 'big data's companies thus far.
It's a shame it didn't come from a university or some PHD project as I quite enjoy drilling down into the nuts and bolts of how they arrived at their outcome.
Nevertheless, their final table is pretty close to what I would predict if I was basing it on the squads as of today - city and Liverpool as the top two, Spuds third with United, Chelsea and Arsenal in some order rounding out the top 6.
Probably should have put some sort of stats abuse trigger warning in there when I shared the link. Apologies for the oversight.I'm in my happy place..
whoever takes that league table seriously needs to have their head checked. I will like to see how/what data was parsed in order to get this report because I bet that data is insufficient and doesn’t take in a lot of factors.
Etherway, I am quite sure that is not how the table will finish as United has a better more mature squad than last season, plus Chelsea is without hazard and have an inexperienced league manager. Liverpool has also looked poor in pre-season and if they don’t get back the motivation they had last season, United, Spurs and Arsenal could finish above them. Overall, I don’t think any algorithm will be able to predict the league table because those type of analytics are not dynamic/real time, unless the table above will be upgraded as the season progresses.
Etherway, I am quite sure that is not how the table will finish as United has a better more mature squad than last season, plus Chelsea is without hazard and have an inexperienced league manager. Liverpool has also looked poor in pre-season and if they don’t get back the motivation they had last season, United, Spurs and Arsenal could finish above them. Overall, I don’t think any algorithm will be able to predict the league table because those type of analytics are not dynamic/real time, unless the table above will be upgraded as the season progresses.
There’s a complete lack of transparency to their methods which they haven’t explained. Unless you can explain it to me? How did they get to this? Or do you just believe anything you read? Because although it looks realistic based on last seasons finish I still stand by my comment about this being absolute garbage. Good day pal.
Good day to you too sir!
I don't know their inner methodology and like I said in a previous post, it's a shame they aren't more transparent about it. Still, there's nothing to believe or disbelieve - it's a simulation of the upcoming season.
You can choose to put as much or as little credence in the results as you like but nobody is advocating that we sack off actually playing football this season and just use analytics to crown the league champions now.
Last edited: