Jump to content
Sign in to follow this  
Thirsty Lizard

Can Plucky Underdogs Norwich Upset the Odds at Preston?

Recommended Posts

9 minutes ago, Jim Smith said:

was he getting the same level of service he's getting at the moment from the likes of Buendia and Vrancic though Bethnal? most (but not all) of his chances have been pretty easy chances which i would have expected him to score so i'm not sure he is necessarily finishing any better than he has in the past our creativity and his movement combined just seem to result in him getting at least 1 very easy chance in every game.

Expected Goals is the measure of how good the chances he received are, so the measure is relative - which is why it is a far better measure than shots data.

Expected Goals models are transferable from league to league as it is a dataset drawn across many leagues. Some of them have data from over half a million shots. 

  • Like 1

Share this post


Link to post
Share on other sites
12 minutes ago, nutty nigel said:

So did you expect Pukki's goals to have dried up before now? Do you expect them to dry up now? Or in 5 or 10 games? What should we expect?

I’ve only recently heard about his high achievement against ‘expected’ so didn’t expect anything personally.

I can’t predict the future but the evidence suggests that on probability Pukki won’t continue to score so many from the equivalent chances. That doesn’t mean to say he won’t continue to score a lot of goals though. 

As I said, hot streaks can go on for a long time.

 

Share this post


Link to post
Share on other sites
3 minutes ago, FenwayFrank said:

Is it possible to have a minus expected goals ? Just asking for a bloke called Lambert 

If you're an ips*** striker yes indeed.

Share this post


Link to post
Share on other sites
1 hour ago, nutty nigel said:

We need this sort of insight on Ray's Funds Beth. 

If this insight was gospel the bookies wouldn’t exist for you to place the bets!

  • Like 1

Share this post


Link to post
Share on other sites
3 hours ago, BigFish said:

Big difference between using analytics to improve performance to what @westcoastcanary appears to be infering - which is using analytics to predict results. If it was possible City's analytics staff would be retired and living off their profits from the bookies. 

As I explained in my original post to Thirsty, performance stats, of which xG is one, provide insight into why a team is getting the results it is. That's retrospective, not predictive. But it doesn't mean that the insight provided doesn't have implications for future results. As Bethnal says, clubs pay a lot of money and employ a lot of people to furnish just such insight; by doing so they get a better understanding of where their own and others's strengths and weaknesses lie, and what action is needed to maintain, amplify or nullify them. The unusually wide discrepancy between our GF and xG, plus the high proportion of our goals contributed by TK (30%; 3 times the number contributed by any of the next three most prolific scorers in the team), gives you an insight into just how great has been our dependence on one player in getting the results we have.

There are other factors too, as Experimental 361's E-Ratings illustrate. Although we have improved defensively, we are still in negative territory compared to half the league. People look at GD while ignoring how that difference is made up. Read the research on the relationship of goals to points and you will appreciate that the ratio of goals conceded to goals scored is a more reliable indicator than GD alone, irrespective of the fact that GD alone can determine league position. 

The perception that there is currently an element of mismatch between results and performance in our case, as opposed to e.g. Leeds and Sheffield, is not confined to me; I started posting on this thread precisely because people asked what this apparently widespread perception could possibly be based on. 

Share this post


Link to post
Share on other sites
17 minutes ago, JF said:

If this insight was gospel the bookies wouldn’t exist for you to place the bets!

I don’t think anyone has ever said it is gospel. The thing that is so wonderful about football and for me lifts it above all other sports is its marginality. 

A team can be thoroughly out played but still win as goal events are so low in football. That doesn’t happen in most other sports. 

  • Like 2

Share this post


Link to post
Share on other sites

3 hours ago, Bethnal Yellow and Green said:

Has Pukki suddenly massively jumped in ability? Or is he on a ‘hot streak’? Chances are the latter. Which isn’t a problem, all players get these streaks and it is a testiment to Pukki that he has kept it going.

I wonder this too, but part of it, I think, is the quality of the delivery. Buendia was immense with his balls into the box vs Ipswich. The delivery was exceptional, but also, it's gotta be a confidence thing too. His goalscoring as upped to another level for Finland since joining Norwich. Perhaps this is also down to fitness and training levels. 

Share this post


Link to post
Share on other sites
9 minutes ago, westcoastcanary said:

plus the high proportion of our goals contributed by TK

I'm assuming you mean TP, otherwise Klose is having quite the season!

  • Like 1

Share this post


Link to post
Share on other sites
6 minutes ago, westcoastcanary said:

The perception that there is currently an element of mismatch between results and performance in our case, as opposed to e.g. Leeds and Sheffield, is not confined to me; I started posting on this thread precisely because people asked what this apparently widespread perception could possibly be based on. 

This doesn't really address the question which started on this thread because a bookie had Preston favourite for tonight, a situation that seemingly has altered since we started talking about this. City's results are City's performance so by definition there is no mismatch. That these results are not supported by the analytics would seem to indicate the analytics are flawed if their predictive ability is the measure of their success. If the City's results don't match those predicted that is surely a failure of the model used rather than City outperforming.

  • Like 1

Share this post


Link to post
Share on other sites
8 minutes ago, king canary said:

I'm assuming you mean TP, otherwise Klose is having quite the season!

Oops! I keep hearing "puKKee" in my minds ear, ringing round Carrow Road :classic_biggrin:

Share this post


Link to post
Share on other sites
5 minutes ago, BigFish said:

This doesn't really address the question which started on this thread because a bookie had Preston favourite for tonight, a situation that seemingly has altered since we started talking about this. City's results are City's performance so by definition there is no mismatch. That these results are not supported by the analytics would seem to indicate the analytics are flawed if their predictive ability is the measure of their success. If the City's results don't match those predicted that is surely a failure of the model used rather than City outperforming.

I don''t think it fundamentally says they are flawed but I do agree there is limited predictive power in XG from game to game. For instance our XG may be lower than Leeds but that doesn't stop a shot deflecting to a striker 6 yards out with an open goal as happened v them. 

The potential predictive nature in my opinion is much more 'big picture.' Ie with us the law of averages suggests a regression to the mean at some point. We saw it with Ipswich last season (started brilliantly but they were basically taking every chance that came their way which was unlikely to continue and when that fell off they fell away) or even us under Alex Neil (getting away with giving up loads of chances which again regressed to the mean). However I don't think there is anyway of using it to say 'the regression will happen tonight!'

Share this post


Link to post
Share on other sites
5 minutes ago, westcoastcanary said:

Oops! I keep hearing "puKKee" in my minds ear, ringing round Carrow Road :classic_biggrin:

Yes it made me chuckle- I like the idea of Klose being responsible for so many of our goals.

Share this post


Link to post
Share on other sites
6 minutes ago, BigFish said:

This doesn't really address the question which started on this thread because a bookie had Preston favourite for tonight, a situation that seemingly has altered since we started talking about this. City's results are City's performance so by definition there is no mismatch. That these results are not supported by the analytics would seem to indicate the analytics are flawed if their predictive ability is the measure of their success. If the City's results don't match those predicted that is surely a failure of the model used rather than City outperforming.

I have no interest in bookies and their odds. I do have an interest in the difference between results and performance, which difference exists irrespective of any attempt to define it otherwise (a form of the No True Scotsman fallacy). You just need to ask Daniel Farke, who constantly refers to how unfair football can be, and pointed out in an interview pre this evening's game that you don't always get what you deserve (or, equally, the reverse).

Nothing I've said has really been about tonight's game. It's been about why people keep expressing surprise at our position in the league, and why they keep referring to us as outperforming expectations. 

 

Share this post


Link to post
Share on other sites
41 minutes ago, westcoastcanary said:

The unusually wide discrepancy between our GF and xG, plus the high proportion of our goals contributed by TK (30%; 3 times the number contributed by any of the next three most prolific scorers in the team), gives you an insight into just how great has been our dependence on one player in getting the results we have.

There are other factors too, as Experimental 361's E-Ratings illustrate. Although we have improved defensively, we are still in negative territory compared to half the league. People look at GD while ignoring how that difference is made up. Read the research on the relationship of goals to points and you will appreciate that the ratio of goals conceded to goals scored is a more reliable indicator than GD alone, irrespective of the fact that GD alone can determine league position. 

The perception that there is currently an element of mismatch between results and performance in our case, as opposed to e.g. Leeds and Sheffield, is not confined to me; I started posting on this thread precisely because people asked what this apparently widespread perception could possibly be based on. 

With respect westcoast it really does seem as though you are using statistics like a drunk man uses lamp posts - for support rather than illumination.  You compare us to Leeds and Sheffield Utd. Yet Sheffield Utd so far have been even more reliant on Billy Sharp than we have been on Teemu Pukki - Sharp has scored 22 of their 53 League goals - that's a whopping 41% compared to Teemu's 30% for us. Meanwhile Leeds ratio of goals scored to goals conceded is: 1.45 (51 to 35) whereas ours is a superior 1.53 (60 to 39). 

Anyway - lets hope we win what appears to be a very tough game tonight. 😉

Share this post


Link to post
Share on other sites

I’m in BigFish’s camp (bowl/pond/sea?). If the predictive model was perfect, there would be no point watching matches as we would already know the result and we might as well all take up stamp collecting instead. What will those of you who believe in the accuracy of the statistical models say if we do end up in the top two at the end of the season? 

 

Obviously there are factors influencing games which the models don’t take into account and these are things that work in our favour. Perhaps the quality of coaching, which results in players improving beyond what their previous career stats would have indicated? Long may they continue whatever they are!

Share this post


Link to post
Share on other sites
59 minutes ago, Nuff Said said:

I’m in BigFish’s camp (bowl/pond/sea?). If the predictive model was perfect, there would be no point watching matches as we would already know the result and we might as well all take up stamp collecting instead

This is strawman in the extreme- nobody is making the suggestion that any predictive model is perfect.

Share this post


Link to post
Share on other sites
59 minutes ago, king canary said:

This is strawman in the extreme- nobody is making the suggestion that any predictive model is perfect.

I think we need to see the stats on how accurate it is 😉

Share this post


Link to post
Share on other sites
2 hours ago, Thirsty Lizard said:

With respect westcoast it really does seem as though you are using statistics like a drunk man uses lamp posts - for support rather than illumination.  You compare us to Leeds and Sheffield Utd. Yet Sheffield Utd so far have been even more reliant on Billy Sharp than we have been on Teemu Pukki - Sharp has scored 22 of their 53 League goals - that's a whopping 41% compared to Teemu's 30% for us. Meanwhile Leeds ratio of goals scored to goals conceded is: 1.45 (51 to 35) whereas ours is a superior 1.53 (60 to 39). 

Anyway - lets hope we win what appears to be a very tough game tonight. 😉

:classic_biggrin: There are so many of us drunks out on the street Thirsty that I can't get anywhere near a lamp post! The message isn't mine; it's in the numbers. It is graphically well illustrated by Ben Mayhew's E-Ratings, Scatter Graphics and Timelines which I suggest you look at again, not just the latest, but over time as well. And Experimental 361 is by no means alone, the message is much the same wherever you look.

I repeat the analogy I used earlier; it is perfectly possible that we will defy gravity and claim top spot and/or automatic promotion come the season's end. But "possible", as I said before, is not a synonym of "certain" or even of "highly probable". And as a further thought, defying gravity is a lot easier in the Championship than in the EPL ............... (though Stoke managed it for a good period under, yes, Tony Pulis).

Share this post


Link to post
Share on other sites

I'm still of the view of a top 2 finish. I have been since December. And for all the analysis (which I love), I have my intuition and eyes. Both show me we are one of two outstanding football teams in this league. And I'm enjoying our performances this year as much as any year in the past. It's fascinating. It's a cracking league is the Championship.

Yes, West Brom and Sheff Utd are thereabouts but just are not at our level. I could be wrong and no problem if so.

  • Like 1

Share this post


Link to post
Share on other sites

I'd argue last nights game kind of showed @westcoastcanary's point about XG. Last night Pukki had a couple of those decent but not great chances (when Buendia played him through and when the Preston defender underplayed the backpass for example) which we've been so reliant on him taking. This time he didn't and we struggled.

Share this post


Link to post
Share on other sites

Not for me Kingo. If Pukki took every opportunity like that he would finish the season on 60 goals or more. I don't expect that. Getting his one goal would meet expectations. Of course it would have been better if he'd scored the first opportunity but I don't expect him to. 

Share this post


Link to post
Share on other sites

21 minutes ago, sonyc said:

I'm still of the view of a top 2 finish. I have been since December. And for all the analysis (which I love), I have my intuition and eyes. Both show me we are one of two outstanding football teams in this league. And I'm enjoying our performances this year as much as any year in the past. It's fascinating. It's a cracking league is the Championship.

Yes, West Brom and Sheff Utd are thereabouts but just are not at our level. I could be wrong and no problem if so.

I agree with you 100% sonyc, and last night has if anything strengthened my belief that we are good enough for a top 2 finish

 

I'll try to post a report but have a bit of work to do today... basically I think our problems last night stemmed from missing both Vrancic and Leitner, so starting with Tettey and Trybull meant we struggled to create openings with our normal fluency in the first half, combine that with a goal conceded at a set piece and then another poorly given away penalty and failing to take our own penalty chance meant we were 2-0 down at halftime which didn't fairly reflect the fundamental quality of the two sides.  In the 2nd half Farke re-jigged our formation and it meant we dominated the half.  we failed to score until injury time for a mix of reasons, being : Pukki had an off night; Preston packed their box defensively; just one of those nights when things don't fall for you; a lot of players being out of their natural positions meaning our attacks lacked their normal fluency and passing moves broke down much more than normal; Rhodes looking ring rusty and struggling to get in the game; a bit of tiredness for some players from Sunday's emotional experience. And of course lacking Vrancic who is superb at the moment both shooting himself and setting up attacks. 

The risk when you're going for it at 2-0 down is you are open to a break, but Preston's 3rd goal did come from an amazing shot.

 

I love the fact that we were scrapping away for it right to the end and alright Max was foolish to push Rudd over, but I could totally understand his frustration. 

 

On another night, we would have gone in at halftime a goal behind and I'm pretty sure then we'd have won, or if we'd got a goal early in the second half, which easily could have happened, then ditto.  There were moments in the 2nd half when we played some great stuff but couldn't get the breakthrough.  Sometimes the breaks just don't go your way.

 

But overall I still felt we looked a better quality team than Preston, and it makes me feel much more confident that with the fixtures we have coming up, we are well placed to win games against the weaker teams in this league and keep on target for another 27 points from our remaining 14 games to reach 87 points and hopefully automatic promotion.

 

 

 

Share this post


Link to post
Share on other sites
48 minutes ago, nutty nigel said:

Not for me Kingo. If Pukki took every opportunity like that he would finish the season on 60 goals or more. I don't expect that. Getting his one goal would meet expectations. Of course it would have been better if he'd scored the first opportunity but I don't expect him to. 

I think you're misunderstanding my point. I'm not saying I expect him to score every chance or even that he should have scored more yesterday. What I'm saying is that Pukki has been exceptional at taking low % chances and turning them into goals (his first v Ipswich is a great example), last night those similar low % chances didn't go in, which is entirely to be expected.

Share this post


Link to post
Share on other sites
3 minutes ago, nutty nigel said:

It is confusing Kingo and I'm lost now. 🙃

When people talk about expected goals or 'XG' it is a stat based on the % of a chance becoming a goal. Generally this number will have been settled on by looking at a large number of similar chances in terms of position, distance, covering defenders etc etc (all quite complex). 

So the goal first Pukki scored against Ipswich would have a pretty low XG rating as most of the time you'd expect it wouldn't go in- he hit it perfectly into the one part of the goal the keeper couldn't reach with pressure from a covering defender. Alternatively the chance he scored v Leeds would have a very high XG- basically an open goal from 6 yards, you'd expect him to bury that most times.

  • Like 1

Share this post


Link to post
Share on other sites

Personally I think if you look at his 2 goals against 1p5wich, the first is a really good chance - the ball has been laid through perfectly so he doesn't even need to control or touch it, he's able to run for quite a few yards while he picks his shot, and he's got the chance to hit it on his favoured left foot at a decent angle, there is a defender but not putting him under immediate pressure.  The second one is also 1v1 on the goalkeeper and trying to hit it straight under the keeper is a good option which came off in this case.  So I'd put both of them in the good chance category.

 

Not sure yesterday he had chances that were as good, part of our problem was that without Vrancic we didn't have the quality of pass to him especially in the first half and the balls Buendia tried to put through for him didn't ever really come off, compared to the 2 which he put away on Sunday.  That's how I saw it anyway.

Share this post


Link to post
Share on other sites
4 minutes ago, It's Character Forming said:

Personally I think if you look at his 2 goals against 1p5wich, the first is a really good chance - the ball has been laid through perfectly so he doesn't even need to control or touch it, he's able to run for quite a few yards while he picks his shot, and he's got the chance to hit it on his favoured left foot at a decent angle, there is a defender but not putting him under immediate pressure.  The second one is also 1v1 on the goalkeeper and trying to hit it straight under the keeper is a good option which came off in this case.  So I'd put both of them in the good chance category.

This is why XG exists though- to actually try and put some form of qualifier on what is actually a 'good chance.' For me I completely disagree about the first goal v Ipswich- he has to take it first time from the edge of the area, under some pressure with basically one corner of the goal to aim at. 

 

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×
×
  • Create New...