Actually it's quite the opposite. xG isn't calculated by asking some bloke how likely you think it was that a particular chance should have been scored. It looks at data of thousands of historical shots, comparing those with similar distance, angle to goal, type of shot it is (eg - header or volley), the positioning of players around them (most notably the goalkeeper) and generating an overall score of how likely that that particular shot would likely go in based on historical evidence. The reason why this stuff is important is because we all get sucked in to the emotions of the game - "he HAS to score that!" and our own personal biases depending on whether it's a team or a player that we like or dislike taking the shot, whereas xG is supposed to remove such biases and look at it objectively based on what usually happens in that scenario.
There are a number of flaws with xG models - most notably that most only actually look at the shots themselves - eg - if you play a beautiful passing move and carve open the opposition and have a player standing with an open goal if you can only pick him out with a simple final pass, but then you fuck up the simple final pass, it would count as 0 xG because no actual shot is taken - that felt like the case a few times yesterday. Where you could argue that opinion comes into it is the weighting that certain factors take in making the calculation of xG, so two different match reports might give two different xG scores if they use a different calculation system.
But...that doesn't mean it should be disregarded. Top PL teams and managers are all over data science now. Just take a look at the way Guardiola's teams play - it's all about the percentages and creating patterns of play that lead to the highest probability of a goal being scored from any given shot. IIRC they didn't score a single goal from outside the box in the first half of their centurion season, which is why that Vincent Kompany goal against Leicester became so iconic - it wasn't just because it was a late winner in a title race, but because a defender taking a longshot was so against the tactics/philosophy that Pep's data-driven approach had instilled in them that it really stood out. Another example - Arteta was very close to being sacked in late 2020 when Arsenal were hovering above the relegation zone but he used data science to "prove" that his team were doing the right things and that he needed a bit more time and luck to get things right - that is paying off for them now. And Liverpool also did some data driven analysis into Klopp's disastrous final season at Dortmund before appointing him - there was a red flag over him as many thought he'd lost the dressing room or that players had tired of his methods - but the data actually showed that they were still doing all the right things but had been particularly unlucky in the goals that they scored and conceded.
A good book to read on it all if you're interested is this one -
Come on, he knows that we are an injury/suspension away from Dier and/or Sanchez playing, and Ben Davies is not gonna rocket in those worldies every gameThe problem with yesterday's win is that Levy is rubbing his little mits together thinking we beat utd with what we got, so good to go
That's what I was thinking.The problem with yesterday's win is that Levy is rubbing his little mits together thinking we beat utd with what we got, so good to go
In fairness, there was that time that we beat Man City without Kane and he didn't get Nuno the players he needed, but we ended up having a super great season laden with success and everyone was super happy all the time and it 100% did not end up undoing almost all the good work Poch had done.The problem with yesterday's win is that Levy is rubbing his little mits together thinking we beat utd with what we got, so good to go
We were in the second half. To make sure we keep things in perspective, boy did they miss some sitters in the first 30.We really were quite good, weren't we?
This, we could well have gone into half-time two goals down and then it's a whole different scenario. Positive signs but we still need reinforcements up front and at centre half. It's a blessing in disguise that Ange will have loads of time to work with the squad at the training ground.We were in the second half. To make sure we keep things in perspective, boy did they miss some sitters in the first 30.
It's positive, forward steps, it's a great atmosphere, and the team is young and energetic. Relative to where we were, I'll take that for now.
AND....for the first time in a while, we're buying good players. It's made all the difference.
Sarr, Udogie, Maddison, Bentancur, Bissouma - basically for the same price as Caicedo. It's sensible stuff
That Bruno header was a massive turning point.This, we could well have gone into half-time two goals down and then it's a whole different scenario. Positive signs but we still need reinforcements up front and at centre half. It's a blessing in disguise that Ange will have loads of time to work with the squad at the training ground.
Actually it wasn't a first time cross. He dribbled down the byline.I think it noticeable that our first goal was a result of Kulusevski hitting a first time cross rather than cutting back onto his left foot. It makes such a difference.
Could have? Would have? Should have? All ifs and maybes that never happened. We could have been 2 up at half time as well as being 3-3 or any score for that matter.This, we could well have gone into half-time two goals down and then it's a whole different scenario. Positive signs but we still need reinforcements up front and at centre half. It's a blessing in disguise that Ange will have loads of time to work with the squad at the training ground.
Why spend money today when you can panic buy some unknown waste of space for an over inflated amount later?That's what I was thinking.
You'd love a chairman though that saw that great performance and instead thought to himself "Imagine how even better we'd will be if we bring in 3 more top players"
Forever resting on his laurels Levy.