Derby v Stevenage FC - Match Day Thread

[Blondest Goat]

37 minutes ago, G STAR RAM said:

Hopefully 👋 

Will just put all the insufferable p*&£s on ignore.   Guess who's first?

[DRBee]

50 minutes ago, ap04 said:

 

In response to what you said, you would struggle to find more than a handful in their rightful place at the end of a season, it's a complete mess. According to xPoints -just one of the estimates- last year it was one in 24.

Given the purpose of Xg and their ilk is predictive ability ( and not just based on one match but a sequence) do you not realise how your comment undermines the value of what you argue for?

[ap04]

49 minutes ago, ck- said:

I think I understand, but enlighten me if you can. 
My understanding is that the xG is derived from considering the outcome of similar situations. The model is built from as many instances as possible. What isn’t clear is whether it’s calibrated according to: 
- the skill level of the player shooting

- the skill level of the league the match is in

The first one in particular surely means that we should expect to do better than xG because we should have better players than the average - otherwise how are we near the top of the table. 
 

I’m genuinely interested, not trying any silly point scoring.

Correct, not factored in at all. There is a post in the Football Forum if you haven't read it already.

The first thing to counter this is that better players do not have much better goals to xG, there is some difference but not much and it varies. It's more the creating of the situations.

Second is that goals to xG is a much bigger reflection of luck (fewer goals than deserved, goals to chances not regressing yet due to the rare nature of goals) than skill.

You can see this clearly in actual goals vs xG league tables. For instance on this year's EPL City and Liverpool (just about the best sides) are only mid-table, even Wolves Forest West Ham and Luton are better at 'converting' (despite worse players).

Another example, last season we also had one of the league's very top sides, very similar to now if not better. We even had McGoldrick upfront. On the goals for xG not only we weren't near the top like now, we were fifth from bottom. We were unlucky, which fits in with the amount of results we deserved on the balance of play and didn't get.

Edited February 20 by ap04

[Ram-Alf]

57 minutes ago, ap04 said:

In response to what you said, you would struggle to find more than a handful in their rightful place at the end of a season, it's a complete mess. According to xPoints -just one of the estimates- last year it was one in 24.

Well they got 2 right

Plymouth as Champions of league 1 in 22/23

Orient as Champions of league 2 in 22/23

 

[Ian Buxton's Bat]

47 minutes ago, ck- said:

I think I understand, but enlighten me if you can. 
My understanding is that the xG is derived from considering the outcome of similar situations. The model is built from as many instances as possible. What isn’t clear is whether it’s calibrated according to: 
- the skill level of the player shooting

- the skill level of the league the match is in

The first one in particular surely means that we should expect to do better than xG because we should have better players than the average - otherwise how are we near the top of the table. 
 

I’m genuinely interested, not trying any silly point scoring.

 

My understanding is that the XG of a chance is based on the likelihood of an 'average player' scoring that chance.....I honestly don't know if the average ability is calculated for each different league.......but the principle is that a penalty has an XG of ~0.8 (meaning 1 in 5 is missed by an average player) and a 50 yard speculative punt will be close to zero.

The team gets a cumulative XG at the end of the match based on all the individual chance XG's being totted up.

Over the season our XG for is pretty low - meaning we don't create too many chances but our XG against is also low, meaning we don't concede many chances.

The team XG's within a league do not factor in whether a team or players are better than average.....it's simple totting up of individual chance XG's

But..........we are scoring more than our average XG meaning we have better than average finishers and we are conceding less than our average XG against, meaning we have better than average goalies or defenders who block chances.....like Adams v Exeter

I've read that Man City use the concept to deliberately create high XG tap-ins from the edge of 6 yard box......that's why they pass teams to death rather than speculative shooting

 

[DavesaRam]

4 hours ago, On the Ram Page said:

Our display in the first half was anything but toothless. Initially we had to match or better their aggressive play (which we didn’t do at their place) which was achieved. It didn’t “almost” work - it did. I didn’t think Nyambe was “gassed out” as you say. I think we had mainly won the physical battle which meant we could slightly more offensive. Warne has got things  tactically right, on the whole, for quite a while now.

I felt Nyambe, who had played well went off the boil early in the second half, for example he kept trying long clearances or passes, but ended up spooning the ball out for throw ins, and did look tired. He did his job, and did it well - as with the whole team. I am loving getting to the final whistle and not feeling that anybody has played poorly! Long may it continue!

[DavesaRam]

49 minutes ago, ramsbottom said:

Exactly this.  Warne and the players got pelters in the reverse fixture for not being able to cope with their style.  So on Saturday he sets us up to counteract it.  Going way back, everyone was up in arms with McClaren for having no plan B and only having one formation, now we have a boss who can set up in many ways, with whatever players he has at his disposal, and is still able to win games, yet still gets abuse from some quarters.  The bloke can't win...

He did on Sattdy!!!

[G STAR RAM]

1 hour ago, Blondest Goat said:

Will just put all the insufferable p*&£s on ignore.   Guess who's first?

Millennium Ram? 😂

[FlyBritishMidland]

2 hours ago, ap04 said:

Says who? I mean everyone says it, but have you got the numbers of how lucky a team usually is or can get after x amount of games and why, or when it eventually evens out, or is it just another cliché?

In response to what you said, you would struggle to find more than a handful in their rightful place at the end of a season, it's a complete mess. According to xPoints -just one of the estimates- last year it was one in 24.

I think you’ll find a certain Mr Clough had something to say about it.  Along the lines of you can win a cup with luck, but the league goes to the best over the season.

Are you seriously trying to say that only 1 team finished where they deserved last season??

[Caerphilly Ram]

11 minutes ago, FlyBritishMidland said:

I think you’ll find a certain Mr Clough had something to say about it.  Along the lines of you can win a cup with luck, but the league goes to the best over the season.

Are you seriously trying to say that only 1 team finished where they deserved last season??

Makes no sense to me either, downplaying what actually happened in previous season or the current one so far in favour of the statistical probability? My understanding of “luck” as a concept is related to the pure chance of something without your input, suggesting you (or in the case of Derby, the team and staff) have no control over the outcome. Saturday is the prime example of us “making our own luck”, set the team up to be stubborn and not concede initially then changed the personnel and went forward looking to win the game….which we did. 

I did share the dictionary definition of luck earlier but felt it was too argumentative for the sake of being so, so I deleted it. I realised it really doesn’t matter to me if one person would rather focus on the statistics they’ve gathered instead of the actual outcomes, when the team wins I’m happy for the result and the 3 points that come with it, that’s enough for me.

[Comrade 86]

3 hours ago, McArthur Park said:

Well doesn't this just show what a pointless statistic it is then?

Exactly. Not a terribly compelling argument, that's for sure 😂