Cardiff City away match thread

[Dollymarie]

1 minute ago, Fordy62 said:

Alright. How we playing? 

Unfortunately can’t watch the game… drinking my way round Barca!  

You’re not missing a belter, so far at least!

[1960maaan]

With a Bell and others on the shoulder, I wish we would knock it longer now and then. The runs are on and it would make them think.

[Mr Popodopolous]

Bit stop-start isn't it, from what I've seen.

Allsop, twice down and taking the sting out playing the system well. What was the injury to Williams like? Last thing we need.

We seem to be from what I've seen, the more enterprising side- albeit relatively even?

Edited March 4, 2023 by Mr Popodopolous

[Leveller]

1 minute ago, old_eastender said:

Pretty even tbh. As against Hull, not working out having Bell CF.

Cardiff have been much better than us.

[E.G.Red]

Wish the commentator on red button would stop calling Sykes O'Leary

[miketh2nd]

Scrappy game , letting Cardiff have the ball so we can counter them , sykes has put in two lovely back post crosses . We need to see more of Mehmeti and Bell though . 

[ProfitInMyPocket]

Says it all that their comms are saying they haven't seen certain players play like this before this season, they're definitely up for it.

We haven't settled yet and we've still created the best chances of the game.

Getting dragged into a bit of a fight man for man but we're holding firm and so are they, just need to be switched on when we turn the ball over.

[W-S-M Seagull]

Think some of thr criticisms are a bit unfair. We're clearly missing Williams. 

[IAmNick]

12 minutes ago, ExiledAjax said:

That Scott/Mehmeti double chance really was huge. xG of 0.23 for Scott's header and a massive 0.65 for Mehmeti's follow up. Huge chances.

That's interesting - and probably a good example of some issues with xG models. Looking how close the Cardiff players were (with odd body positions) I'm sure in reality it was a lot less than 0.65. Incredibly difficult to measure though and once you start getting body positions in there the sample size would drop massively.

 

Really tough chance for me

Edited March 4, 2023 by IAmNick

[Leveller]

2 minutes ago, Leveller said:

Cardiff have been much better than us.

Though on the occasions we’ve got forward we’ve looked threatening. Mainly Sykes.

[AshtonRobin21]

This is poor.

Bell is not as effective when played centrally. 

Sykes keeps cutting back onto his left foot and we lose all momentum.

Loose passes and slow build up.

They look to have us figured out at the moment.

[Ecko]

Pretty poor half to be honest. They are tight and compact at the back and midfield, which doesn't help. 

Williams going off and Scott playing deeper isn't ideal either.

Pring, Sykes and O Leary are playing well.

Needs a much improved second half.

[Andrew_V1]

Wells for bell half time please not in the game at all. As a spectacle this is absolute litter so far, passing is snail paced with no movement. Pring only one who seems to want to drive into space 

[marcofisher]

Need to bring the ball down a bit more, Cardiff loving it being headed around the midfield.

[Engvall’s Splinter]

6 minutes ago, Andy082005 said:

We’ve looked really poor so far today. Need to get Wells on second half

 

5 minutes ago, Leveller said:

Cardiff have been much better than us.

Interesting how people see the game differently. The game is far from a classic I agree. We’ve had some chances but Cardiff look a bit more comfortable. Relatively even. 

Edited March 4, 2023 by Engvall’s Splinter

[ExiledAjax]

3 minutes ago, IAmNick said:

That's interesting - and probably a good example of some issues with xG models. Looking how close the Cardiff players were (with odd body positions) I'm sure in reality it was a lot less than 0.65

 

Really tough chance for me

Numbers from infogol. It's the best of the free xG number crunchers, but a more sophisticated paid for provider may well give a different figure. The also may adjust it on review. 0.65 is the instant inplay figure. You're correct it's not an exact science.

Checked and FotMob are giving it just 0.25. They give Scott's 0.18. so infogol may well have gone a little mad there.

0.25 is still a very good chance from open play.

Edited March 4, 2023 by ExiledAjax

[Alessandro]

A tense, fairly even, generally scrappy first half. It has the feel of a bottom end of the the table contest.

We’ve looked bright at times and probably had the better of the possession.

Chances-wise, both teams have had them - ours probably the better, notably a cross from Sykes on the break off the line from a Scott header and a Tanner cross hitting the bar. 

Talking of Scott, he and Mehmeti very quiet today but that’s as much to do with them being marked out the game - in fact it’s nice to have a home team so obviously set up to counter our threats. 
 

They say if you double up on one player that brings opportunities for others…which brings us on to Bell. You can’t fault is work rate, but he’s just not ready yet for the central role. Virtually every time the play came to him it broke down.  I’ve said it before and I’ll say it again, I’m surprised NP continues to play him there right now.

Hope Williams injury isn’t a bad one, the reshuffle weakening us for me with Scott dropping deeper and Weimann still an bit out of form.

[old_eastender]

HT 0-0.

Sooner Wells is on for Bell the better, or even Wells for Mehmeti and play Bell in his better position on left. 

 

 

[Mr Popodopolous]

Hindsight etc but when Williams- I missed 1st part of 1st half- went off Sykes back into CM, Wells on, Mehmeti and Bell wide? Ideally you want as few changes to the shape as possible, stretched though we are squad wise.

[BUTOR]

We're defending very well 

[IAmNick]

3 minutes ago, ExiledAjax said:

Numbers from infogol. It's the best of the free xG number crunchers, but a more sophisticated paid for provider may well give a different figure. The also may adjust it on review. 0.65 is the instant inplay figure. You're correct it's not an exact science.

Do they adjust the result manually sometimes then, from what you're saying? Or adjust the input variables?

I'm a big proponent of xG by the way so no criticism. It's just interesting when you see examples where it perhaps works well/poorly.

Quote

Checked and FotMob are giving it just 0.25. They give Scott's 0.18. so infogol may well have gone a little mad there.

0.25 is still a very good chance from open play.

Yeah - the 0.65 stood out as that's close to a penalty in many models I think!

Edited March 4, 2023 by IAmNick

[Malago]

After Tuesday happy with 0-0 at HT.  Need to keep it tight for as long as poss and hopefully nick it at the end.

[Sturny]

1 minute ago, Bouncearoundtheground said:

We're defending very well 

That's something good to take from this. Going forward we're very disjointed. Need to mix it up and halftime will be good for us I reckon. 

Edited March 4, 2023 by Sturny

[Midlands Robin]

Average first half. 

We are really struggling to break down their back five. 

We've had nothing down our left and Williams going off has not helped. 

Cam Pring my Mom so far. 

Second half we could really do with Wells on and we need to go at them from both wings. 

In the centre of midfield we've not been pressing them hard enough. Most of their attacks have come from the middle of the park normally after one of their players has fallen over and been awarded a free kick. 

The Cider'eds are in great voice. I don't think the Diff fans have made any noise at all. Seems like quite a flat atmosphere. 

 

[Curr Avon]

Very hard to break down that back 5 and City's slow tempo is suiting Cardiff.

Can we play in Weimann breaking the offside trap occasionally?

For them, the loanee Philogene is giving Cam a real test and Ojo looks good in the breakdown.

With those 2 free-kicks conceded by Scott, we were relived to hear the whistle.

Just need to improve our passing in the final third and spring that offside trap, because Cardiff are trying to win by osmosis.

 

[Roe]

Not a great first half but I love how solid and organised we look in general

A couple of really good chances too

[Red_Alligator]

Agree some HT changes would be good to see. Definitely needs a freshen up if we're to nick a win. 

[joe jordans teeth]

Pretty even game,definitely goals for us in it,the sooner Conway is back the better 

[marcofisher]

We could do with pace on the wings, move Memehti central maybe and move Weimann up top? Or take Bell off for wells and move Weimann left

[ExiledAjax]

1 minute ago, IAmNick said:

Do they adjust the result manually sometimes then, from what you're saying? Or adjust the input variables?

I'm a big proponent of xG by the way so no criticism. It's just interesting when you see examples where it perhaps works well/poorly.

Yeah - the 0.65 stood out as that's close to a penalty in many models I think!

Yeh most models give a penalty around about 0.72.

I'm not sure if Infogol do a full human review, but they definitely adjust the numbers after a game. A few times I've looked at FT figures, then gone back a couple of days later and it's changed.

They have a detailed methodology here https://www.infogol.net/en/blog/education/expected-goals-shot-maps-explained-8102019

Their in-play model does not consider defensive players. It's mainly based on "the x, y location on the field where an attempt originates, whether the attempt is made with the foot, the head or another body part and a descriptive assessment of which type of phase of play preceded the attempt." So a shot like Mehmeti's, without the defenders, is basically a penalty.