Harry Kane is a worse version of Van Basten Kane still shines and scores many goals because he's playing in the easiest era of all time with protection from the referees and VAR Kane's performance in decisive games, based on his career average, is absolute shit. like Ronaldinho Gaucho or Roberto Baggio were also in Big matches at averages !
I'm sure he has some we can dig through the record. But the point is he's not an elite goalscorer given he doesnt score against elite teams.
Just for the record, I don't necessarily disagree that Kylian Mbappé has already done enough in his career to be regarded as one of the greatest players ever. However, what do you make of this remarkable fact? Kylian Mbappé smashed the all time World Cup goalscoring record before scoring a single non penalty goal at the European Championship. We're not talking about a player who simply scored a lot of goals at the World Cup. We're talking about a player who surpassed every goalscorer in World Cup history before managing even one non penalty goal at the European Championship, whether in the group stage or the knockout rounds. The exact same player, representing the exact same national team yet producing two completely different goalscoring records.
Many things can be true. Mbappe's performance in 2018 was key to France winning, but nothing that we would put up on the historical record as sublime or extroardinary. Mbappe's performance in 2022 was more in keeping with a player of his stature, topped off with an insane hattrick in the final to comeback from the precipice - that final match alone secures his status forever. Club football aside - where Mbappe has excelled, though his UCL remains evasive - that alone 2 out of your 2 first world cup as a young player, the second as the protagonist, is enough to secure the status discussed. Most great players that we can think of have had some form of poorer showing in one or multiple international tournaments. As for why Mbappe has a poor showing in Euro - there are several contributing factors: 1. his worse form, especially the last one - he was not fit and was clearly worse. 2. worse team in those editions with the exception of the benzema led one - but they got bounced vs switzerland due to their arrogance and deschamps mismanagement 3. deschamps 4. better opposition
Have you ever used ChatGPT, Gemini, Claude, Grok, DeepSeek, Perplexity or any other LLM model? Have you found them valuable? In principle, they work basically the same except football models are not as advanced as LLMs yet due to incentives. Well, off the bat I can already tell you are confusing context. 9 rating vs Cape Verde in a group stage match is not equivalent to 9 rating in the world cup final. Football models do not attempt to contextualize performance based on a match significance, stakes or broader sense of difficulty. It focuses entirely on objective, measurable markers and indicators of performance within a context on the pitch. So the comparison is a false equivalence. Ratings of football models are not created to evaluate greatness in such manner, so a failure to do so is not really a failure. Obviously getting a 9 by football model in the wc final is much greater than getting a 9 in a friendly match for example. Not only is that not a failure, that is actually one of the most valuable characteristics of football models and why they are necessary tbh. Human evaluators are notoriously bad at objectivity, because we are terrible at isolating actions on the pitch from all the noise and narratives. We are easily persuaded by a broader context when we shouldn't be. When we should be objective about it. I like this disucssion. It is very important. There is a lot more to say. I will continue later.
For what it's worth, France Football's retrospective Ballon d'Or had Pelé winning it in '58 but not in '65.
You have to ask why performance in international tournaments matter to your view of his greatness. Is it because it shows he can perform outside of a system (this is the view that many had for Messi pre-2022)? Is it because it shows he can perform at the highest level of competition (the argument many have for those who fail to perform in the UCL)? Is it because it shows he can perform under pressure, in what is in many ways a different sport as far as team dynamics are concerned? You can ask a lot of questions like those. At 2 Euros, the sample size is just too small to draw much in the way of meaningful conclusions. PS: I say "you", but this is really for anyone
I don't think it just Mbappe who struggle to score goals in euro despite scoring a lot in world cup. There seems to be a trend amongst many great european players, of course there's a lot factor we need to take into account but I still think it's interesting that many of these players who scored lots of goals in world cup can't bring that same prowess to the Euro. Why do World Cup top scorers struggle to find the net in EURO? pic.twitter.com/0kEtvYtKOB— UEFA Cristiano League (@UCR7L) July 19, 2026
Nah, 1958 belongs to Didi. Pelé was only 17 and wasn’t even a starter for the Brazil national team at the World Cup. He wouldn’t be named best player in the world at that point. On the other hand, Didi was an established elite player, widely regarded as Brazil’s most talented footballer, and he was the one named best player of the World Cup that year. Pelé only settled himself on the team during the competition, he was still a rising star. Pelé is unquestionable from 1959 to 1965. Then he has a drop in level in 1966, after getting injured at the World Cup and even drops out of the national team for 2 years, until 1968. The 4-year gap between 1966 and 1969 is when you can have players topping Pelé as the world’s best (Eusébio, Best, Charlton, Tostão, Jairzinho, Ademir da Guia, Pedro Rocha, Gérson, Rivera, Riva, Albert…). And, of course, he gets topped by Garrincha in 1962 because of his World Cup injury (even though he dominated with Santos at club level).
Doesn't change the fact that Ronaldo is a narcissistic RAPIST. Even Portuguese players and people hate him. Anyone who is a fan of a rapist is no different from a rapist.
Yes, they hallucinate regularly are massively oversold and people use them to build complexity where simplicity will suffice. They are also designed and trained by humans yet people will claim they are objective in a way humans arent when those humans have built the very systems and decided what data to input etc. You can't tell me I'm confusing context because I'm not confusing context at all, thanks for you telling me I am though because you can clearly read my mind. You explained how the model worked and I gave my thoughts on why it seemed silly by citing an example of how the ratings quite obviously don't tell the story you imply. So now I'm confused about context, according to you, because I understand that Pele's stats from game against Italy are far more likely to lead to a victory than Pedri's against Cape Verde. It has nothing to do with one game being a final and another not being a final so you're grasping at straws bringing that up and tweling me I'm confused. I'm not at all confused and explained quite clearly what my issue was, which you have completely failed to address. Pele's stats from the game I cited are much, much more likely to lead to a win than Pedri's stats from the game I cited. They could be reversed in terms of which one was the final and it wouldn't change my thoughts or my point in the slightest so maybe it's actually you who are the one that is confused. I could use plenty of examples from other games that aren't finals to make the same point. A rating of a 9 in any game is equivalent to a rating of a 9 in any other game, how could it not be? It's quite literally the exact same rating so telling me they are not the same is quite obviously just factually wrong. You explained that the model takes the data and produces a prediction about how the stats will likely lead to a win. It converts that prediction into a number. Pele and Pedri got the exact same number and it matters not that one game was of more importance because the model isn't accounting for that so my point remains completely unaddressed. Any model that tells us Pedri's performance against Cape Verde was as likely to lead to a win as Pele's against Italy is silly unless I'm missing something. The same logic applies to Messi's rating against Spain and Ronaldo's against Germany and both of those are WC finals. So since the model is supposedly objective the point you make about me being confused is even more bizarre. The model is using an objective criteria apparently and doesn't account for game importance hence the ratings are exactly equal and you're telling me I'm confused for laughing at the ratings being equal when its quite clear, objectively, that Pele's stats are far more likely to lead to a win than Pedris and yet somehow their ratings are the same. Now I am actually confused, BTW what makes the stats objective? Are humans, who lack objectivity per your post, the ones who categorise the stats being fed into the model? Are humans, who lack objectivity, responsible for developing and designing the model? What makes it actually objective? For all the great value you see in the model I see something that tells me Pedri's performance was as likely to contribute to a win as Pele's and I laugh at the notion of this being an objective or worthwhile measurement. While humans may be bad at being objective, including when building LLMs presumably, it doesn't change the fact that if you presented me both sets of stats and asked which was more likely to contribute to a win I wouldn't need any computational power to tell you it was Pele rather than Perdri. The model is telling us they were equally likely to contribute to wins hence why they have the exact same rating. If we look at what actually happened it turns out the magic model with all its so called objectivity was actually completely wrong and I was completely right. Pedri's 9 performance wasn't as likely to contribute to a win as Pele's 9 performance. We know this because Brazil actually won 4-1 whereas Spain drew 0-0. I don't mean to be combative at all but I find your response surprising and it completely fails to address the points I'm making so now I am confused about that. I'm not at all confused about the original point I made which has been completely unanswered. Btw telling me that a 9 for a final is not the same as a 9 for a friendly after telling me the model doesn't account for significance of the game and is supposedly objective is wildly contradictory. If its simply an objective measure that doesn't consider context but only the measurable stats for the game then yes a 9 in any match is the same as a 9 in any other match, how could it not be?
I have just one question for you: in 2026, in what aspect is Ronaldo on Messi’s level? Let’s even say you’re right about the goalscoring argument and that Messi only scores against weak teams. But Ronaldo couldn’t even score against those teams at the World Cup. So I’ll ask again: in which factor is Ronaldo on Messi’s level?
You are absolutely wrong about this. The xA stat is never even calculating xG. It is calculating the chance that a pass will become an assist, not whether a shot from that exact spot will be a goal. These are very different things. If you’re looking for the passing counterpart to xG that would use the same model and same data, you’re thinking about xGA, not xA. I promise you that you’re wrong about xA. It is a completely different model that measures a completely different thing, uses different data, and does not take into account the same factors (including not taking defender location into account). And the thing about Son can’t actually explain this either. For one thing, Kane passed to other players a lot. They just didn’t score as much from those passes because they weren’t good finishers. If a player makes good passes to lots of players, the best finisher is naturally going to make up a disproportionate amount of the assists. And even talking about just Son, he overperformed his xG in the relevant years and competitions by 43.7%, while Kane’s assists more than doubled his xA. So even if he literally only passed to Son (which is of course not even close to true—he made plenty of passes to much worse finishers), Son’s finishing would not come close to explaining why Kane’s xA undershot his assists so much. Nor can Bayern’s finishing explain it in Kane’s Bayern years, as I already explained. There’s just no way to reasonably explain away the gap between Kane’s xA and his assists with teammate finishing. I also just want to step back and point out again that this isn’t theoretical. We literally see lots of Kane matches with less than 0.1 xA for the whole match despite a big chance created (and I showed video I could find of one example). I think it should be obvious from that that he’s being undersold by xA, because people at OPTA obviously looked at what he created and thought he created something where the recipient should be reasonably expected to score while xA clearly doesn’t reflect that at all. The fact that the assist to xA gap is also very consistent with xA underselling him and that I’ve provided an explanation of a potential mechanism for why xA undersells him just further backs up the point.
Yeah, direct comparison among players of today vs 20th century ones in non-sense, the formers are so ridiculously favored by a plurality of factors. Indirect comparison only does make some sense, and in this respect in my humble opinion Pele remains above Messi by a margin. No one in the history of the sport has ever appeared as untouchable by his contemporaries as O Rei. It is not even close.
what a cluster ******** of a thread i'm surprised these ancient forums are still around and of course you have your typical larping brasilians who are 'retro' watchers trying to say pele is better messi. the man played 1000 years ago in one of the easiest eras in football and you want to say he's better
I dont really understand this attitude. All players play to the conditions and circumstances of their era. They are best viewed relative to their peers and to describe the career Pele had as being among the easiest eras ever seems to fly in the face of reality. I think both Messi and Pele and have incredibly strong resumes and are the leading GOAT contenders. It's very natural to compare them. There are factors beyond their control that can weigh for against them overall but are not necessarily to do with their merits as a player. Taking just one example when Barcelona lost against Inter Milan many Messi and Barca fans insisted, some still do, that it was because they had to take bus to get to Milan. Compare and contrast the horror of having to travel from Barcelona to Milan with the schedule that Pele played and how he travelled with Santos/Brazil. In some matches they literally had to get changed in their bus on the way to the stadium. Doesn't sound particularly easy to me. Of course there are also things that Messi faced which Pele didn't such as just a vastly higher number of players simply due to population growth. Its fine to try and work out how these factors and different variables impact a players legacy. The issues arise, imo, when we try to selectively impose variables without due consideration of context and dont account for other factors. For me the discussions are interesting and can help to inform, illuminate and give insight into players past and present. I like football and I like history so no surprise that I like football history. If your attitude is just to say Pele's era was easy and there's no debate or discussion to be had then what's the point of engaging in a discussion?
You're mixing up shot based xA with models like xT and then acting as if Opta is using completely unrelated logic for each one. In Opta's system, when a player makes a key pass that directly leads to a shot, the xA assigned to the passer is derived from the pre shot xG of that resulting shot. They aren't isolated worlds with entirely separate rules. xG measures the likelihood of the shot becoming a goal and xA credits that same probability back to the passer. You're treating shot based xA as if it were expected threat or expected goals against. There's also a problem with your Son example because the math actually supports my point. Son overperforming his xG by roughly 44 percent absolutely matters when we're discussing Kane's assists versus his xA. If Kane plays a pass worth 0.08 xA and Son receives it, carries the ball another 20 yards, beats a defender and scores a fantastic finish, Kane still gets one full assist while his xA remains at 0.08. That means Son created a huge amount of the final value through his carry and finishing. The same logic applies when other elite finishers repeatedly turn low probability situations into goals. Actual assists can massively exceed xA without proving that the model is broken. Your Olise example runs into the exact same issue. You keep pointing to the big chance tag as if that automatically means the pass itself should receive huge xA value. Those are not the same thing. A human Opta logger can classify something as a big chance because the attacker eventually finds himself in a clear scoring situation, but if the receiver still had to control the ball outside the box, run a long distance and finish under pressure, the pass itself may still carry relatively low xA because historically those situations do not convert very often. The model is isolating the value of the pass itself rather than crediting the passer for everything that happens afterward. What you're ultimately asking for is for Kane to receive credit for the carry, the dribble and the finish that happen after the pass. But the whole purpose of xA is to separate the passer's contribution from the receiver's subsequent work. If Son or Olise do extraordinary things after receiving the ball, that's exactly what xA is designed not to attribute to the passer.
If you were referring to me, overlooking the laughably childish accusation of being a larping, you incurred two huge misunderstanding 1) I am italian for generations, never even been in Brasil 2) My point was exactly that one cannot directly compare players over eras. You can at most do it indirectly, by measuring players against their contemporaries, and then trying to decide who was the most dominant. So saying that Pele played in one of the easiest era not only is false, but even more is irrelevant here, cause the environment would have been favorable for many others, not only for Pele'. And in this framework, in my humble opinion, Pele was superior to Messi, cause he was more outstanding in his time. I accept different opinion of course. But dismissing completely Pele is something that only a misfit like a fan of Paredes or a Milei voter can have the guts to do.
Besides, it makes no sense, as a River Plate fan, to dismiss history. The club's golden era was in the 1940's, with the "La Máquina" squad. It makes ZERO sense to dismiss history, when your own club (and your own national team) had its peak 20 years before Pelé.