The underlying data

[weasel]

From the BBC



Leeds have been playing a like a mid-table side with their eyes closed. Pretty good football, but inefficient.

A lack of goals is a real issue for Daniel Farke's side. Only Wolves' tally of seven is fewer than Leeds' 11.

They are having shots, though. So far, they have registered 146 shots, which is more than 11 other Premier League clubs, including Aston Villa and Crystal Palace.

It is converting those shots to goals that is letting them down - Sunderland, for example, have taken 31 fewer shots than Leeds yet have scored three more goals this campaign.

And it is only bottom of the league Wolves (6.2%) that have a worse shot-conversion rate than Farke's side.

Defensively, only three teams have conceded more goals than Leeds this season.

Yet Leeds have given up fewer shots than 12 clubs in the league, including reigning champions Liverpool, Bournemouth, Newcastle and Brentford.

Both of Leeds' goalkeepers have been among the worst shot-stoppers in the league - Lucas Perri has let in 1.6 and Karl Darlow 1.8 more goals than they should given the quantity and quality of chances they have faced.

Only Wolves have a worse goal difference, but Leeds have a better xG difference than Tottenham, Aston Villa, Sunderland and Fulham.

So it comes down to where blame lies for Leeds being in the relegation zone. How much should be laid at Farke's door and how much at the performance of the players or the summer recruitment?

[weasel]

So our way of playing, how we have played, what we have done in terms of creating and conceding chances etc should see us 11th in the table. As such is Farke to blame or is it down to the recruitment in particular our signings of a striker and a keeper.

I know some people don't like the xg and all that but putting the xg stuff to one side I think all of us can see that we aren't clinical and have found ways to concede when we shouldn't.

Maybe we need to find a way to play which makes us create less chances but take a higher percentage of the chances we create. Maybe also we need to give the opposition more chances in the hope that we then save a higher percentage.

Stats like this often are the reason why a team 'suddenly' performs better with a new manager. It isn't the 'new manager bounce' but that these things revert back to an average so we are due to score a few more and concede a few less regardless of who is in charge. A bit like how Bamford would go 10 games without scoring then get 6 goals in 8 games. His average would be 1 goal in 3 matches but he at times would look far worse than he was and at other times far better than he was - but played out over enough games the average settles and maybe that will happen with us.

[Cjay]

I don't think you can entirely blame Daniel Farke, same as most struggling managers is always some context usually stemming from the natural lack of quality as you go down the league.

At different points most managers probably have a point.

At one point last season Ipswich were underperfoming their xG , Jesse Marsch wasn't wrong in his assessment about us either at one point.

The issue is first of all, the data tends to drop as confidence goes etc, teams struggling get progressively worse.

It's rare a manager can turn that data around especially down the bottom end.

The other issue is it lacks context in terms of when a team deliberately allows a team to have the ball rather than being pegged back and the goal peppered.

So for example when we played Tottenham the xG and shots was about the same up until Tottenham scored again after that they dropped off, we had 7 shots they had 1, our xG will have gone above them but we didn't create anything of note.

Burnley was the same, our biggest xG of the season 16% of it in fact in 8% of our games, but that was because Burnley dropped off, they let us have the ball, they let us shoot, but we didn't create anything of note.

That's the issue with xG it doesn't have context.

Arsenal could easily have scored more goals, their xG was less than 3.

Brighton could have had 5 or 6, xG less than 3.

Data is useful , I think it's very useful, but some things only your eyes can tell you.

And Farke knows that, every struggling manager pulls the xG card when it suits. You never hear it when you win.

So I don't think Farke is entirely to blame, but I don't think he helps either.

I don't think he is the sort of manager who can close those data gaps with coaching or tactics, which is probably what we need.

[weasel]

I don't think you can entirely blame Daniel Farke, same as most struggling managers is always some context usually stemming from the natural lack of quality as you go down the league.

At different points most managers probably have a point.

At one point last season Ipswich were underperfoming their xG , Jesse Marsch wasn't wrong in his assessment about us either at one point.

The issue is first of all, the data tends to drop as confidence goes etc, teams struggling get progressively worse.

It's rare a manager can turn that data around especially down the bottom end.

The other issue is it lacks context in terms of when a team deliberately allows a team to have the ball rather than being pegged back and the goal peppered.

So for example when we played Tottenham the xG and shots was about the same up until Tottenham scored again after that they dropped off, we had 7 shots they had 1, our xG will have gone above them but we didn't create anything of note.

Burnley was the same, our biggest xG of the season 16% of it in fact in 8% of our games, but that was because Burnley dropped off, they let us have the ball, they let us shoot, but we didn't create anything of note.

That's the issue with xG it doesn't have context.

Arsenal could easily have scored more goals, their xG was less than 3.

Brighton could have had 5 or 6, xG less than 3.

Data is useful , I think it's very useful, but some things only your eyes can tell you.

The thing is if we have 15 shots of low quality as we often do, but the opponent has 1 good chance, scores, then sits off the xG won't reflect that.

So I don't think Farke is entirely to blame, but I don't think he helps either.

I don't think he is the sort of manager who can close those data gaps with coaching or tactics, which is probably what we need.

Although I agree with most of what you say the highlighted bit isn't strictly true. If you had say the 15 low quality shots then maybe lets say they each only had a 5% chance of being scored, so those 15 shots would equate to an xg of 0.75 (so less than 1 goal). The opposition has a good chance, let's say an open goal with an 90% chance of it being a goal. As such their xg is 0.9 - so still less than a goal but it shows that their 1 chance was more likely to result in a goal than those 15 chances we created. So the xg does actually make sense in that regard. Sort of like if you hit a shot from 30 yards maybe 1 in 20 times it goes in.

100% agree with you though that the state of the game can often dictate the stats. Against Villa, as a recent example, they bossed the second half but after they went in front I bet we bossed the stats, had more possession, more shots etc because we had to attack more and they were happy to defend and hold onto their lead.

I do though think that it shows the flawed thinking in our recruitment. We needed player who were capable of getting near their xg, whatever end of the pitch. I think it may have been you who highlighted the fact that DCL vatly underperfromed against the xg. For me we would have been better off just buying 2 or 3 real differences makers rather than 7 or 8 'average' players.