Maradona vs Messi comparision in different way

Discussion in 'The Beautiful Game' started by darek27, Nov 1, 2022.

  1. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    What you'd choose is irrelevant. Of course, you somehow twisted this entire post.
    First of all, it stands to reason that Maradona did not take that many shots over 10 games. The sample size for shots is wildly heterogenous over a 10 game stretch - so how much we can infer is dubious. Secondly, passing volumes from both players in WC setting is similar with Maradona producing more forward passes. Thirdly, passing over 10 games will naturally be a more stable representation of Maradona's playstyle. So yes, he is less one-footed as a player. Lastly, we've already established that Maradona isn't an all-time goalscorer, nor was he tasked to be his team's primary goalscorer. Perhaps he is inferior to Messi in terms of finishing, two-footedness shots. Perhaps not. I am not sure how on earth you can derive that conclusion from a 10 game sample. Maradona wasn't finding himself in scoring and shooting positions like Messi as he played a distinctively different role. So I am not certain about the pertinence of your observations, since comparing how one-footed a player is as a shooter vs. a passer is a comparison that well... is illogical in the first place.
     
  2. Sexy Beast

    Sexy Beast Member+

    Dinamo Zagreb
    Croatia
    Aug 11, 2016
    Zagreb
    Club:
    --other--
    Nat'l Team:
    Croatia
    My point regarding eye test is that xGoT, the way its measured, uses only few measurable factors in consideration like placement of a shot and power(?) and lacks nuance that i would describe as decision making part (timing, set ups, etc.)

    I am not entirely sure how is xGoT determined in the first place, maybe that would reveal flaws in the metric, but I am pretty sure it doesnt capture all of it.

    That is why i say eye test for a single case is better than xGoT, but in a larger sample size, at scale, I agree that eye test always fails to some extent. Human brain simply cant hold many information at the same time and accurately assess the average...

    Great example is the goal from Bernardo against Real the other day. The xG of the freekick is 0,03 while xGoT of the shot is 0,12.. this for me vastly underestimates decision making behind the shot and that is specifically because model such is xGoT doesnt account for ingenuity. Bernardo seeing the open angle and Lunin off of his line expecting a cross, took opportunity to surprise him.

    If Lunin was a bit more prepared for a shot and reacted better, shot wasnt particularly difficult to save, but decision behind attempting the shot is great. Even if Lunin saved it, you can be sure that in the next free kick in similar area, he would be closer to the line which would open up more space for cross. This level of nuance is something that any model will never be able to capture as it can not evaluate decision nor future consequences of an action.

    Another example are penalty kicks down the middle. If player anticipates that GK will dive to side for whatever reason, shot down the middle is a good decision as it brings zero risk of missing but in theory it is an easily "savable" shot down the middle.

    I suspect that if Ive studied examples more that I could see more flaws that models could never account for.

    And that is a reality of metrics such as these.

    Although in a bigger sample sizes, these things somewhat cancle out so as a general rule of thumb, they are very useful.
     
  3. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    Everything I have seen suggests that xG is a mediocre predictive tool. At best, some sort of xG overperformance metric as applied to a team over a season seems to have some correlation with league placement.
    I genuinely think the way it is commonly used in broadcasts to be complete and utter rubbish.
     
  4. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    I don’t know why you assume the variance of a low sample size is more likely to mean the real truth would support your point more. It could easily be the opposite. In any event, though, I didn’t know this was from a 10-game sample (I merely looked at the screenshot you posted and responded to it, without digging for the exact context that you’d not included). If it’s data that shows relatively small percentage differences over just a *10-game* sample, then we can’t derive any real conclusions from it at all. So the point you’re trying to make is even weaker than I’d thought.
     
  5. Sexy Beast

    Sexy Beast Member+

    Dinamo Zagreb
    Croatia
    Aug 11, 2016
    Zagreb
    Club:
    --other--
    Nat'l Team:
    Croatia
    I dont know what you mean by "predictive" but I think i know where you are going with it and i would disagree with that particular point.

    When it comes to a single game, knowing xG doesnt mean much but that is more down to unpredictability of game itself rather than flaw of xG per se.

    But in an example where you see that team over the course of a some period of time vastly underperforms their xPoints (calculated from xG and xGC), it is a strong indicator that in future things will turn the tie.

    Using just shots taken and shots recieved to predict the trend for a team is a worse predictor than using xG values.

    So I agree with limitations of the model and that some might overhype it, but it is like that for a reason. Indeed it is the most advanced, relevant, useful metric in the world football right now. Without a doubt.

    And i think something much better that accounts more nuance will unlikely to emerge any time soon
     
  6. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    It does not a rocket scientist to understand that the passing characteristics across a 10 game sample are far more likely to represent an actual trend (ie. the sample has sufficient power), than the shot sample in 10 games - especially of a player that is not a striker.
    This is exceedingly obvious and does not require further explanation.
     
  7. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    I agree with that but the passing data is still derived from too small a data set to be worth much of anything, even if it’s larger than the shot sample.
     
  8. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    These are probabilistic models. Their real use and application is to be able to predict ... something. That something could be the amount of goals over a season, league position, points accumulated. The point is to try to correlate the output of the model with real-life results and make predictions.
    The most egregious use of it is what we see with modern broadcasters - the xG of a match is presented as if a team "should have scored x amount". It's a complete misunderstanding.
     
  9. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    There is no way that can be a legitimate take. While 10 matches will not be representative of a career - a player's playstyle changes, tactics change, teams change. The particular variable - percent of weak-footed passes - I think can be very reliably derived from a 10 game sample as a general idea of the player's playstyle. I have no idea about his pass completion rate, the quality of those passes, but it's clear that his UTILIZATION of his weak foot is more than Messi's. Now to call it worthless is just your usual bad habits.
     
  10. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    The thing is that you’re talking about small percentage differences, such that the difference between 83% and 89% is not going to be many passes. It’s something that could easily vary with a larger sample. I’d also emphasize again that Messi’s passing volume is higher, so Messi the number of weak-foot passes per match is likely not very different. For instance, we know Maradona attempted 37.5 passes per match in the specific WC games that were in that sample. If 17% of his passes were weak-foot passes, that suggests he made about 6.4 weak-foot passes per match. Meanwhile, that article says Messi over the five previous seasons had averaged making 8-11% of his passes with his weak foot. And we know from the data that in the five seasons before article, Messi attempted 63.6 passes per game. Which suggests Messi made between like 5.1 to 7.0 weak-foot passes per match. So it’s really not all that different in raw numbers.
     
  11. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015

    Again, you are side-stepping the point. The percentage offers an idea of utilization. Nobody is discussing impact or volume. Maradona has a greater proclivity for using his right foot in the game - and this data supports that assertion. Secondly, you are again making the mistake of assuming that Maradona's 86 numbers are representative of his VOLUME across his careers.
    In other words, I am discussing one thing and you are discussing another.
     
  12. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    Yes, this data supports the assertion that Maradona was more likely to use his weak-foot when passing. It is not strong support for that assertion, since the sample size is small. It is also not support for an assertion that Maradona actually used his weak-foot in passing more often in absolute terms than Messi did. So it is a very limited point. And, on the other hand, there’s also small-sample-size data showing weak support for the idea that Messi was more likely to use his weak-foot when shooting. And that is also more concretely backed by the fact that, Messi almost certainly has a higher percent of his goals with his weak foot than Maradona. Messi scored 15.2% of his goals with his weak foot. I’m not sure of the corresponding number for Maradona (maybe @Trachta10 knows?), but I am almost certain it’s significantly below that.
     
  13. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    It is entirely possible that Messi shot more proportionately to Maradona with his right foot. I don't see your point.

    I think the idea that a 10 game sample is insufficient to gleam a perspective of Maradona's likely utililization of his right foot to be ludicrous. But as I said, you have a regular habit of contradicting yourself when it suits you.
     
  14. PDG1978

    PDG1978 Member+

    Mar 8, 2009
    Club:
    Nottingham Forest FC
    #214 PDG1978, Apr 16, 2024
    Last edited: Apr 16, 2024
    I'm tending to think a vague comparison, in terms of role, could be Eden Hazard 2018/19 with Maradona 1984/85
    Eden Hazard 2018-19 | Dribbling Skills & Goals - YouTube

    Maybe Messi 2011/12 would be vaguely comparable with Zico from 1983/84 (first Udinese season) I'm thinking. I guess even Messi's role from the 2022 World Cup could compare a bit to Zico's Flamengo role too (playing from a little behind the front), although I think in gameplay and the fact it's prime vs prime Flamengo-era Zico and false 9 Messi would be a little comparable too (but Zico having a striker in front of him as well as two wide attackers; nevertheless it's clear he plays combinations and arrives in the 'striker' position fairly often).

    EDIT - When I said 'prime' I didn't mean to imply necessarily the best version of Messi was when he played in the most goal-focused role and scored the most goals. You guys have more insight on him than me I guess anyway, but I'd probably think myself more of top-form 2015 Messi, or after that maybe top-form 2011 Messi, if I was to try to narrow down 'peak' version.
     
  15. PDG1978

    PDG1978 Member+

    Mar 8, 2009
    Club:
    Nottingham Forest FC
    #215 PDG1978, Apr 16, 2024
    Last edited: Apr 16, 2024
    Maradona does suffer 4.79 fouls per Serie A game according to that (in 1986/87 not 1984/85 which we don't know about of course). WhoScored has Hazard on a little under 3 for 2018/19 in the Premier League (2.8, which would rise very slightly adding in Europa League games, where he suffered 3 per game). He received more in 2014/15 (3 per game in PL, 4.7 per game in his few CL games; also 4.5 per game in the 2018 World Cup for Belgium...with Maradona on 7.6 per game in 1986 though as shown on Sofascore that has the same figure, taken from OPTA data, for Hazard 2018 - the roles they played were not so similar in those World Cups anyway though, with Hazard playing more from the left side).
    Eden Hazard - History (whoscored.com)
     
  16. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    #216 lessthanjake, Apr 16, 2024
    Last edited: Apr 16, 2024
    The point would be that you said Messi is more one-footed than Maradona. If Messi shot a good deal more with his weak foot than Maradona, that would obviously contradict your assertion. You’d then be left relying on a mere 10-game sample that indicates Maradona may have made a higher percent of his passes with his weak foot. But even then, we know that Messi had greater pass volume to the point where that data doesn’t really indicate that Messi actually made fewer passes per game with his weak foot. Taking all that together, it would seem odd to see all that and definitively conclude, as you did, that overall “Messi is in fact more one-footed as compared to Maradona.” If you want to come to that conclusion, then fine, but I don’t think it’s concretely supported by this data (and, I should note, the StatsBomb article the data came from didn’t come to that conclusion).
     
  17. lessthanjake

    lessthanjake Member+

    May 9, 2015
    Club:
    FC Barcelona
    Interesting thoughts! I’m curious what you think the differences are between Messi’s role in, say, 2015 and Maradona’s role in very early Napoli or prior to that. I know you’ve said a little bit about this already, but do you think they’re meaningfully different, and, if so, what do you think the differences were? You mentioned earlier that you think Messi on the wing is more comparable than Messi as a False 9, which is why I am using 2015 as the comparison point (since Messi played on the wing that year). Obviously no two players play *exactly* the same role (and, in fact, even the same player will have small role differences match to match), so I guess I’m curious for differences that we think are genuinely important differences.
     
  18. PDG1978

    PDG1978 Member+

    Mar 8, 2009
    Club:
    Nottingham Forest FC
    Yeah, I'd tend to think there isn't a big difference between Messi of 2015 and Maradona of 1985 in terms of how high they played relative to team-mates.
    Messi in 2014/15 did take more shots per game (4.9 in La Liga, 4.3 in the Champions League) compared to Maradona in 1986/87, but like I say 1984/85 we don't know for sure (I'm assuming it'll be in a similar ballpark for Maradona probably though, given more relative parity between teams and Napoli faring less well that season too) and Messi was fouled less too, but those differences might have to do with circumstances a bit more than player roles I guess (both like you said being mainly involved in advancing play, making chances, trying to score).
    Lionel Messi - History (whoscored.com)
    Messi's heatmap would I guess have more action towards the right side of the pitch, but for sure he came infield a lot didn't he so like Maradona in 1984/85 was in effect a playmaking forward.
     
    lessthanjake repped this.
  19. SayWhatIWant

    SayWhatIWant Member+

    Jan 10, 2015
    Obviously, the discussion pertains to playstyle / comfort with the weak foot. I've spoken at length about Messi's over-reliance on his left foot for passing. Shooting is far more dependent on rare situations and forces you more in terms of what foot to use.
    Again, you've built a strawman - nobody discussed volume - I don't understand the obsession. Footedness is a descriptor of playstyle and quality. But you've cleverly and deceptively shifted this about a conversation on VOLUME - which is UNPROVABLE and allows you to deter from the actual point at hand.
     
  20. PDG1978

    PDG1978 Member+

    Mar 8, 2009
    Club:
    Nottingham Forest FC
    In this post I should have pasted this formation too (for 86/87 - the other one on the page was for 85/86 and I think originally from Italian media as it's the same format as the one from the page about 1984/85)
    https://www.bigsoccer.com/threads/m...in-different-way.2123245/page-7#post-42146168
    [​IMG]
    The page calls the formation 4-1-2-1-2 (labelling Maradona as AM, but indeed in various places he's labelled FW in Napoli squad summaries so it's feasible to see him more like a 3rd forward - more like the old-fashioned inside forward I suppose, albeit not in a much-used system before that time - to me England did play kind of with a 4-1-2-1-2 in effect in the 1966 World Cup but in that case Charlton was more an AM than FW for sure I think, even if playing more advanced in certain games than others as alluded to in some commentary).

    My impression for the opening season is that he plays like combined role a little bit between AM/SS and left wing forward (although generally with freedom to go everywhere and appearing on the right too at times - that was where Bertoni was playing from though, while especially Penzo was more like a striker than winger in positioning I think, albeit he didn't play all the games).

    This was that other 1984/85 Maradona video, and I see now there is a 1985/86 one too (as well as some Cruyff World Cup 1974 videos on the channel for example):


    Available full 1984/85 Maradona performances (not the highest rated ones - in the Udinese game he got a 7.5 which is considered very good, though with the goals alone he'd be heading in that direction; Zico got a 7 as shown on Vegan10's thread which is also good for the Italian standards for sure and one of the best grades of the game too)


    Zico isn't playing in as attacking a role there as he generally did in 1983/84 I believe, but more like a classic playmaking AM....
    Maybe some of you have seen these already of course.
     
    lessthanjake repped this.
  21. PDG1978

    PDG1978 Member+

    Mar 8, 2009
    Club:
    Nottingham Forest FC
    #221 PDG1978, Apr 17, 2024
    Last edited: Apr 17, 2024
    The grades I refer to there were from Guerin Sportivo I think (I checked the other day on vegan's thread). The one for Maradona vs Milan from them was 6.5 (still relatively high/above average for Italian media) but Gazzetta dello Sport (from Milan as opposed to Turin) gave him a 7 in that one. The highest ratings he was getting in the season were 9s though IIRC.

    EDIT - I couldn't be certain Bertoni was nearly always playing from the right side in 1984/85 I suppose, so that comment was based on assumption partly. Also, I should have clarified the formation is labelled 4-1-2-1-2 from 1985/86 onwards, or specifically for 1985/86, as per the page linked on the original post I made....
     

Share This Page