I don't know that noise is a good explanation for why the maximum xPts are consistently lower than real points. Noise should go in both directions.
Regardless, I agree that Barcelona being ahead of Real Madrid in xPts is probably meaningless.
Champions League — quarter-finals and semi-finals draw
- Thread starter Red Star One
- Start date
I don't know that noise is a good explanation for why the maximum xPts are consistently lower than real points. Noise should go in both directions.
Regardless, I agree that Barcelona being ahead of Real Madrid in xPts is probably meaningless.
The thing with xPts is that variance is baked into the model, because wins are worth 3 points, not a % of it based on win probability. For exaple, by the understat model arsenal's recent win over Brighton - 4xG to 0.5 - was worth 2.95 xPts
I tend to agree it's not a particularly good stat honestly, but i don't know enough tbh
It does. For example right now Bayern and Bayer are quite close in their xPts and Bayern even is slightly ahead. The team that's overperforming is noticed as such, but nobody cares about the underperforming team. It's quite natural to see an overperformer at the top because that's what brings them on top. The underperformers aren't noticed as much.
I don't know that this should be the case.
It does. xPts award less than 3 points per win
Most definitely. "Luck" is a far bigger factor in limited group states and knockout rounds than a league table setting. It's partly how a side nowhere near winning their league title can surprise and win a cup tournament.
xPts is perfectly capable of awarding a higher xPts than Pts for a champion, see here Barcelona having >94 xPts and 91 Pts.
Anyway there's a stats thread somewhere, I'm going way off topic here.
How about injuries? Are they somehow represented in the xPt approach?
They often play an important role during the season
They often play an important role during the season
If the expectation is that xPts has a linear correlation with actual points (and for a large number of the naive (no pun intended) audience, it is), then yes, you're right.
I don't see it like that. Here is how I view xPts and other xStats in general:
Points ~ xPts + xOtherBullshit
xPts is a statistically significant input to actual points. Which makes sense. The more chances you create and the less you concede, the more points you can expect, on average.
The xOtherBullshit just accounts for the fact that there is more to football than chances created/conceded and aggregated over 38 games. Some of it we have a good statistical handle on, and some of it is still a challenge.
So xPts being wildly wrong is true if that is your only input to actual points and you assume the model is linear. Maybe there is a non-linear relationship e.g. Points ~ xPts^1.05, where higher xPts yields disproportionately higher actual points (and vice versa). And maybe we are not accounting for other variables (examples like form, pace of play, relative level of injuries, relative fitness of squad, experience, etc) that play into the actual points. It's a combination of both I would guess.
It all comes down to a misunderstanding and misuse of stats amongst laymen. I do think there are more advanced models out there that correlate many variables. And unfortunately they aren't available to the general public.
Oh yes, the competition where a club buys the second chef of the referees for 20 years and Atletico plays at radically different intensity depending on who is the top rival is an unbeatable reference for analysis.
He's miles and miles beyond what Salah achieved at the same age. That's a pretty laughable comparison.
I agree. But in xOtherBullshit I include 'actual bullshit', i.e. methodological flaws.
Oh absolutely. I just couldn't find the epsilon symbol on my keyboard in time and was not drunk enough
Indeed, evidently I don't know anything about football. Fancy thinking that the best team finishes first in a league table. What an avant garde notion.
Yes mate.
This is a mental suggestion. He’s far better.
Why do you believe that? Do you think there is no such thing as an undeserved winner in football?
Not in a league format
Why do I believe that? Because it is what the league table is designed to determine. It's the entire function of it.
What team(s) that have won the Premier League do you think didn't deserve it, rendering the table invalid? Out of the 31 Premier League seasons?
Why not in a league format?
I'm not following the Premier League that closely. But I know that football teams win games they deserved to lose all the time and I also know that a season is too short for luck to even out. So even in a league format, the winner isn't necessarily the best team.
Blackburn. United were better that season and beat them twice
Because in a league you play every team home and away like everybody else and the number of games ( usually above 32) ensures less variation. Also good teams usually create their own "luck". Apart from Liverpool of course who have paid off the refs and the fa and use steroids and feck their own sisters.
But less variation doesn't mean no variation, no?
From a statistical/scientific perspective, 38 games simply aren't enough to eliminate the luck factor. That's not really anything opinion related but mathematics.But less variation doesn't mean no variation, no?
Why have the league table at all then, if it's not a valid way of determining what the best team is?
Strictly speaking, this isn't even true. Teams are allowed to sign players a few games into the season, then again in the transfer window, and they are allowed to replace their manager at any point in the season. So you are not even really facing the 'same' teams just as everyone else.
Then there's the issue of playing in cups or European competitions, fixture congestion and delay, all of which can have an impact.
Last edited:
Years ago Spurs lost out on a CL spot because their players got diahrrea and people are really going to tell me luck doesn't factor in a league competition.
I have never seen anyone be more unbelievably pedantic than this. I'm gobsmacked. You absolutely 100% must be trolling now.
Yeah, it's odd to claim such.
Salah at 25 had bounced from Al Mokawloon to Basel to Chelsea to Fiorentina to Roma scoring 68 goals in 226 games plus 31 in 52 for Egypt by June 2017. He had won a Swiss player of the year award and a few other notable achievements. He moved to Liverpool shortly after turning 25 and that's when his career truly took off.
Mbappe had played in two World Cup Finals by age 24, scoring a goal in a Final win and a hat-trick in a Final loss. He had recorded more than 250 club goals and 46 country goals by the time he turned 25 last December. His honors list is ridiculous.
I never used the word same. Strictly speaking my statement that you bolded is absolutely true.
But less variation doesn't mean no variation, no?
It does not mean no variation, like Iker said there are plenty of factors that go into it. This is a sport played and reffed by humans after all. However in all sports the league format has the least amount of variance. That's why for eg city have won the league plenty of times recently but only have one European trophy.
I think almost everyone agrees that more games is less variation; the argument started because someone said luck was irrelevant in a league which is a bolder claim.
Maybe but a simpler explanation for why City only have one European trophy is that they're not playing the same teams. There's other league champions in the CL and they also don't win that many European trophies, it's more competitive.
Thankfully they have ditched the away goals rule because today's ties would've been so boring and defensive
I wonder if BVB stands a chance, but I guess the late goal they scored can give them some hopes. I fully expect Barca to win tonight and hoping for an exciting game
Dortmund ist just weird this season. They lose when they should win, they win when they are expected to lose, all while playing poorly and yet at least in the CL the results are absolutely fine.
Lovely stuff, classic amongst supporter groups this season.
Looking at those teams, PSG have Skriniar, Soler and Ugarte on the bench. All of whom would walk into our line up and improve it. All of whom were available for decent fees, or out of contract over the summer.
Not saying we should have definitively gone for those particular players, but in terms of looking forward to this coming summer; it should be clear that there is always value in the market if one recruits sensibly and with forethought.
I am buoyed by the changes to our structure, and getting the right football people in, but in the short term worried about our dealings this summer. That work should be taking place now, but the key players aren’t yet in place (not through our fault), and I don’t trust anyone currently at the club to make good choices. Might be a difficult summer, and for us to more reasonably expect real movement come next year instead.
It’s a shame we won’t have CL football to offer, but assuming that we stick to the words of Berrada, and don’t overpay for players (in fees or wages), it might mean that we recruit players who actually want to be here and “buy in” to the vision, rather than players coming for the money and CL opportunity.
Not saying we should have definitively gone for those particular players, but in terms of looking forward to this coming summer; it should be clear that there is always value in the market if one recruits sensibly and with forethought.
I am buoyed by the changes to our structure, and getting the right football people in, but in the short term worried about our dealings this summer. That work should be taking place now, but the key players aren’t yet in place (not through our fault), and I don’t trust anyone currently at the club to make good choices. Might be a difficult summer, and for us to more reasonably expect real movement come next year instead.
It’s a shame we won’t have CL football to offer, but assuming that we stick to the words of Berrada, and don’t overpay for players (in fees or wages), it might mean that we recruit players who actually want to be here and “buy in” to the vision, rather than players coming for the money and CL opportunity.
They certainly do, they got a bit of momentum towards the end of the first leg and hit the woodwork a couple of times as well. Could've easily gotten a draw.
They are a vibes team so definitely can be up for it at home. 1 goal difference isn't a lot.
Both BvB and Atletico are weird and you don't know what you'll get from them. Both sides are more comfortable when they play on the counter
Atletico does well when they sit deep and counter superior attacking teams but Dortmund is hardly an attacking team, and they even have less attacking qualities compared to Atleti
I still expect Atletico to qualify
I’m at the Barça game and the atmosphere is wild, give me a goalfest
So wild that Barca fans attacked their own team bus mistaking it for the PSG one
Ugarte definitely
Last edited: