I will take Pep's word at face value. I care what he says. If you don't believe him that's your choice. I reckon that he's achieved more in football than you so really there's no contest.
Expected Goals. There and back again
- Thread starter RobinHood
- Start date
You just don't get it do you, that quote doesnt mean what you think it does.
I'll give you a another quote - seen as you have finally worked out how to use them:
On Pep Guardiola
Bayern Munich’s Head Of Analytics Tells Us How Pep Guardiola Has Embraced Stats
At Boston’s annual MIT Sloan Sports Analytics Conference, the main focus is on U.S. sports. But people come from all over the sporting world, and the globe, to …fivethirtyeight.com
He’s deliberately misrepresenting Pep saying players shouldn’t focus on their personal statistics (goals, assists etc) to pretend Pep thinks data analytics aren’t important.
He did the same thing with formations earlier, Peps famous telephone numbers line is not because he thinks formations are nonsense, it’s because modern football teams use 2,3,4 formations depending on the phase of play, so asking a manager about the formation Sky puts up on a graphic before the match is meaningless.
Strange that City / Pep want another first team analyst then isn't it? Its almost like you are talking absolute nonsense and making a fool of yourself.
Are you referring to me or Pep?
I hope you don't mind me using your post as a reference. You and others took the time to post good, even excellent material on the subject. Perhaps you shouldn't. All this is yesterday's news, obviously professionals have been using these tools for years, and the research for new tools and new methodologies is and will remain an open process: Football is a highly complicated sport. Not to mention, there are numerous sources out there for people to get relevant information and study the subject. As long as they are interested, of course...
I remember a few years ago, I used to visit regularly a site, infogol or something (it no longer exists, I think), for info on xG etc. I have spent hours and hours and hours studying findings on that site and many others. Disparities such as the ones in Don's post were not unusual. if you studied the material carefully, used additional info perhaps etc, you would usually solve the puzzle. I'm not talking about the obvious ones, others have already provided examples: You have a team that scores high on xG, yet does not get the actual goals -and does not get the results. Provided the finding is systematic, it could suggest that the team's game is competitive enough to regularly guarantee creation of chances, even good chances, even sitters. Yet those chances are not converted. Of course there is (always will be) room for improvement in terms of the quality of the team's game (which is an entirely different discussion), but it could suggest that the club in question has neglected, among other stuff, a fucking attribute called "finishing". Extremely important in a sport like football, which is not decided by performance related points (like boxing, for example). Nope, it's decided by the difference between goals scored and conceded. The reason finishing is considered such an expensive attribute, clubs could invest fortunes on a player that possesses it...
But cases like the above are rather easy, what I find fascinating are more sophisticated stuff. Such as identifying correlations between different metrics. Or relationships between individual and collective stats etc. Kevin, for example, would usually score lower than the team's average performance in terms of "pass completion" (constantly around or even >90% through the years). The manager's decision supported this, you knew it, you accepted it (do I have to explain why?). The aim would always be to analyze, rationalize, evaluate, elaborate on a series of findings. Next step would be making the correct interpretation and drawing the correct conclusions. Which would lead to actions and measures (technical and / or tactical) on the training ground so as to enhance individual / collective performance...
As you and others have said, this is common practice in the industry, long time now. It's also an area of information, methodologies and ideas that can help fans improve and develop their knowledge of the sport. Again, as long as they're interested in terms of working towards those ends, of course. We are referring to people seeking for the truth of things, lazy people should stay away from this field...
I have come across "discussions" such as the one in this thread numerous times, in social media etc. Always, I mean every fucking time I wonder. What is it that people do not understand. Why the fuck people do not learn. I mean, eventually you will have to accept that there are people who join forums not to contribute to a discussion (even though they pretend doing so), not to be informed on a subject, not to be educated. Nope. They join so as to fulfil other kinds of needs. You don't want me to go there, trust me, but I'm sure you understand what I mean. The thing is, they are bound to fail miserably, those needs will not be satisfied, they never do this way. Which means, their problem will not be solved. But they do not know this. So they will keep insisting. As long as there is a platform that enables them to do so...
A discussion can transform into a magical process, a concept well known since like 2,500 years ago. But if you and others actually think you're having a discussion, I have news for you: You're clearly not. He is not fucking interested, man. He has made up his mind -and nothing will change that, it's not your fault. I mean, there is a guy who has already posted fucking definitions of the metric, we are back in the fucking classroom. He has even provided the limitations of the methodology in question (direct quote: "Small sample sizes mean variance and randomness (e.g., wonder goals or errors) can distort match conclusions."). What the fuck else do you need? You and others have been posting in good fucking faith, he is not fucking listening, how the fuck am I supposed to express it?
I mean, you are trying to convince a guy who talks about "subjective data". Who questions the objective nature of -football- statistics. Take a moment to reflect on this, please. Do you understand what I'm saying? Do you, really? If you do, you and others know what to do...
The content in this thread is very interesting for a lot of people, trust me on this. My friendly advice to you and others: Keep it interesting. Keep contributing. Keep sharing experiences like the ones you've already posted. You know what to do -and you know what to avoid...
Last edited:
So when Pep says statistics are overrated you claim that he is wrong? And I am making a fool of myself? Obviously you know better than the best manager our club has ever seen.
You just bristle with supreme arrogance.
You are making a fool of youreself and you are misrepresenting what Pep said, which you have been told numerous times by different people but completely ignore. Just like you ignore all the questions which dont suit your narrative.
I'll try one last time why are we employing more analysts if Pep doesnt like analysts?
citizen_maine
Well-Known Member
- Joined
- 29 May 2011
- Messages
- 17,963
I'm not sure I see the logic in spamming a thread about a subject you fundamentally reject. Smells like WUMming.
Maybe best if people ignore it
Maybe best if people ignore it
You, and some others. Not Pep.
3 points about Pep and analytics.
1. His teams play like teams that have tactics based around analytics. Actions speak louder than words.
2. Sometimes he shuts down questions in interviews because he doesn't want to answer them. You should take what he says sometimes with a pinch of salt.
3. His saying there's too much stats doesn't mean he doesn't utilise analysis. It might just mean he feels he has enough and doesn't need more.
What do you expect? City 2 Liverpool 1, City xG 2 Liverpool xG 1?
If it were like that it really would be fucking pointless and add nothing.
They are accurate. They show which team created the chances with the highest likelihood of being scored. They are accurate at being xG. They are not supposed to tell us how many goals actually were scored.
Last edited:
I think the opposite. xG is probably the best barometer in the sport to show how a team has played and what the game should have ended up. It shows that teams have created goal scoring opportunities, and not just opportunities but the level of opportunities whether they be low goalscoring likelihood or high.
When City draw or lose a game and our fanbase loses its head about how shit we’ve been in the game but the xG (added to possession, shots, touches in the opposition box etc. and the fact the opposition happened to take advantage of one situation they could exploit because we play in the best league in the world and even teams down in the bottom are excellent teams so can exploit individual situations) is City 3.5 v 0.5 for the opposition… it shows we haven’t been shit, Pep doesn’t need sacking, X player isn’t shit, Y player isn’t a ****, Z player isn’t a liability and we don’t need to sell everyone and spent another £275m on new players… we actually played well and should have won but just didn’t put the ball in the net because sometimes even elite football players don’t quite connect with the ball quite right or that even teams down in the bottom half have excellent ’keepers.
Similarly, there are games where we’ve drawn or lost but the stats show we actually played better than some games we’ve won, and that’s a big reason why there’s no need to panic as much as our fanbase does after a draw or defeat.
Last edited:
Players like that managed so well and became such big names because of the lack of statistics in the game at the time. Opposition players had no idea how to defend them because they weren’t coached to. Watch back the full 90 minutes of the 1970 World Cup Final (that Brazil team still unfathomably heralded is the greatest of all time when in reality they’d lose to average Championship teams from today) where Italy defended worse than a Sunday League team on Hough End in the modern game and had no idea what they were doing.
Football is in another stratosphere now compared to even after Maradona retired because coaches base their philosophy of the game around stats and getting players to create chances to score from the averages of what players do in the game… or to stop it at the other end with teams now knowing how to actually defend based on the stats their coaches analyse.
It’s why Rodri won the Ballon d’Or rather than overrated Vinicius Jr with his fancy flicks and turns. Whereas in the 60s 70s and 80s, Vinicius-overrated would have won it because Ferdinand kept saying ‘Ballon d’Or’ over and over in the commentary.
Rodri: Much, Much More Than a Defensive Midfielder | Opta Analyst
When Rodri arrived at Manchester City, he was a number six. He is now far more than that, making key contributions at both ends of the pitch
theanalyst.com
^this didn’t just happen because Rodri got a bit more confident and improved, it’s because Pep and his coaching staff will have analysed every single aspect of his game and those of all the other teams we play against. From that they will have sat him and the rest of the team down and gone through in fine detail using videos and charts on a screen training him like a lecture in a university to become the player he has and not just clapping him from the sideline on the training ground and saying ‘go on Rodders lad get on the ball more’.
Last edited:
Robanompivenov
Well-Known Member
Exactly this. xG is as much of a statistic as possession, corners etc. It's a statistic based on probability. Can't for the life of me understand why people get so wound up by it. Strange "old man shouting at cloud" behaviour.