Okey, can you share what you have? Because I can tell some of your numbers are wrong. For example, you use the source of goal.com that give 47 assists to Pelé in the national team, but when I did the research I found that some of them are wrong and also I found other assists that are not counted there. Do you just use goal.com as a source or did you do your own research for the national team? For Cosmos, did you just take the assists from that book or did you do your own research for each game? Because you seem very concer about the "opta criteria" but I suspect that this source uses similar criteria to Transfermarkt. Also for the season 1959, you tell that you find 13 assists, but for me that number is too low.
Here is the detail of Pelé's goals in the Paulista Championship. There are 466 goals in 410 games. I don't know why the RSSSF registers 467 goals. In the national team, Pelé played 91 matches, not 92. FIFA does not recognize Brazil's match against an all-star team in 1968. In total, Pelé scored 756 goals in 818 official matches. https://futebolpaulista.com.br/Noticias/Detalhe.aspx?Noticia=16359
Here I found a very interesting piece of information. Maradona scored 4 official goals with his hand. The hand of God existed long before Mexico 86. TODAS LAS "MANOS DE DIOS" DEL 🔟Maradona no sólo metió un manotazo contra Inglaterra en México 86. Para la Selección, Argentinos, Boca y el Napoli convirtió una decena de tantos con la ayuda del brazo. ABRIMOS HILO⬇️ pic.twitter.com/STfxNa30s0— TyC Sports (@TyCSports) September 16, 2021
A colleague of mine reviewed the games and scored 466 goals and it was corrected, 467 was counting with an own goal in a 1967 game.
This is the second La Liga game in the Nou Camp: According to transfermarkt the last game he played before his illness was against Real Sociedad on 4th December 1982, and his first game back was on 11th March 1983. This match against Real Madrid was on 25th March 1983, so he was back for just 2 weeks after missing 3 whole months, and didn’t seem to be at his best to me, despite having a goal and an assist. Compare that one to his performance against Real Madrid on 26th November 1982, just before his illness: This one looks like a potentially “prime” Maradona to me. His dribbling and creativity is just devastating to Madrid, and despite the Bernabeu pitch being horrible compared to the Nou Camp pitch he still played much better than in the second game. I think this shows how important match fitness is even back in the early 1980s
@PDG1978 Hello mate. I was thinking a way to measure the "level" of the competitions the player participate, and then adjust the contribution percentage with that level and create some kind of rating. Maybe you can help me with some ideas. For example we can use the Elo system. The best way to do this is to see the Elo level of the rival the player faced in each game and then create an average for each season, but this method is practically impossible, first because is a titanic work, and then because Clubelo only considers the games in leagues and champions and europa league, and of course don't have data for southamerican leagues. So, the best I can do is just see the Elo average of each league in each season, this number represent a rough average of the level of the games the player participated, but at least is good enough to have an aproximate idea. First example: Michel Platini Season 1972/73 http://clubelo.com/1972-12-31/Ranking I consider the number in the 31 December of 1972 for the season 72/73, (maybe other date is a better idea) The average Elo rating of the french league is 1432 That is the Elo in each season. If I calculate the total average in his career, the number is 1564.13 Second example: Johan Cruyff one isue, there is no Elo for the NASL league, so I consider a totally arbitrary number of 1400 Elo pts. This is provisional and can be changed The total average in his career is 1499.18 Third example: Cristiano Ronaldo The total average in his career is 1708.81 and just to clarify, the Elo is the average of the league in each season, so I'm not taking into account the champions league games
Elo ratings of league is a dishonest way to judge Say in a league there are two/three teams with massive difference with others. The league's elo ratings will be inflated by by these top two/three team which would hardly even reflect the story Also counting opponents Elo will be not a good measure The actual measure should be rather difference in Elo Ratings of your team vs Opponent teams your team faced during that time This is only way it should be calculated if you don't have intention of promoting certain players
You can do this by only way i am telling you. But it is going to take lots of hours and effort. You have to find : 1) Elo ratings of their team 2) Elo ratings of the opponent Just before that particular match , then maybe create three Columns : 1) Team's Elo 2) Opponent's Elo 3) Difference Then probably take the cumulative and Average especially , season by season like you did for the goal and assists Stats And then , Compile them That would be a flawless and most correct way to do it.
Or if it takes too much effort , you can do one thing as a shortcut (although it would be far less accurate) (Elo Ratings at beginning of season + Elo ratings at end of season) / 2 For all teams in the league. Then maybe find out difference with rest of the teams in the league , and calculate the cumulative and Average , season by season
But don't you think if I do this, that would probably destroy the numbers of players like Cristiano Ronaldo and Messi? My idea is adjust the contribution percentage of the player with the "level" of the oponents right? But the thing is that my contribution percentage already adjust the numbers in some way For example Maradona has 0.722 G+A-PK per90 in his career Cristiano Ronaldo has 0.913 G+A-PK per90 in his career (Cristiano with a better ratio) But Maradona has a contribution percentage of 48.47% compared to Cristiano that has a 43.93% Now is Maradona who has a better number. We can say Maradona played in relatively weak teams and Cristiano in more strong teams So if you think about it, is like I already adjust the "team´s level" of the player Imagine if I do that again, If I took the contribution percentage of Crisitano and Maradona and adjust that number with their Team's Elo, that will destroy even more the numbers of Cristiano That's why I think only take into account the opponent level is what I should do
Thanks for the tag Trachta - I'll take a look at your posts and give it some thought over the next days (just letting you know, since I logged in and posted on another thread this morning).
Yeah, having had a bit more time to look and think than I did yesterday, I'd say I agree with Legolas that this would inflate the calculations for players who played in 'top leagues' but largely in one-sided matches, and I think it's maybe a bit unnecessary if I'm honest about it, as your goal contribution numbers stand as they are and people can then interpret the circumstances they were achieved in themselves. I don't think you can make a calculation to really determine the best players (short or long term), or even the ones with the most impressive end product contributions, in a decisive, unarguable way, and when introducing any method with some subjectivity then it will always come down to individual choice about how to do it, and others can always argue it should be done a different way, with a different method, different weightings or whatever. I think your goal contributions numbers are good as they are, and you/we did end up with a few different variations of those already I suppose. taking out penalty goals for example, or even the one weighting assists more. I would say though (following similar principles to your goal contribution idea actually I think) that it is the relative quality of the teams lined up against each other that would be more important than strictly the standing of the opponents based on ELO (with perhaps different emphasis to some extent between different clubs of different nations on Europa League for example...not to take too much away from Spanish clubs who won that or suggest others would always stop them if they tried - in these days it's true that clubs that go further in the Champions League are not entered into that competition at any point though, so a large number of English, or German or Italian clubs say going far in the Champions League takes out contenders for the Europa League that in the past might play the UEFA Cup or Cup Winners Cup if they weren't league champions of their country...but now I go off track a bit maybe...and of course good Champions League results count towards ELO for those clubs that achieve them too). I think ELO is a bit delayed maybe as a method anyway - maybe it's be better to look in theory at a 3 year period - the end of the 'previous' season to the one where a game is played, the end of that 'current' season and the end of the next one too and then average them out? I just mean that ELO ratings don't consider how well the club goes on to do in the remainder of a season a game is played in, and they are built up cumulatively over a period of time so reflect the past more than the present potentially (although the post-game ELO rating is adjusted isn't it, from the pre-game one, so a great result against a top team does bump it up immediately in that sense).
Yes they reflect the past more , but also the present . They take upto the latest match you've just played in . Btw shouldn't a great result against a top team should increase your rating? It reflects your strength as a team (yes there will be always upsets but then your rating would go down in other matches you lose) so i don't see how there's a problem But what you said is also right, since a big portion of result is based on previous matches, so a team that did way better last season but had a huge drop off this season after say major managerial changes or transfers wouldn't get reflected in it immediately (would get eventually) just not now. But 3 year period is dishonest metric because it is a long time and in club seasons , things can change within 1 year literally , or even 6 months (although it would get reflected in elo more or less) but can be hugely bumped by result in 1st season of 3 (say) where your team was nowhere as strong ,bt turned to a superteam in next 2 and majority games you faced a team was in latter 2 seasons. Probably better way is to take Ratings before the season and Ratings after the season and take an average from it , which would reduce a lot of the problems and efforts.
Yeah, I haven't studied how it's calculated closely (or lately) enough really, but I meant just that it might be better to look at the trend of the ELO rating a little bit after the match concerned at least (because a team can just be starting to become really good, and because at any point in time the ELO number is indeed based on past results going back a bit, although yeah I wasn't saying it's not good that the rating changes positively after a great result...just stating that it does reflect these kinds of things in the post-match ELO if that is what is looked at rather than the pre-match one - although hypothetically Messi scoring a hat-trick in a 4-3 loss to Dortmund would show up better than in a 4-3 win against them maybe if using the post-match ELO rating to determine level of opposition...which actually could be a case of fine margins to some extent and a bit misleading?....but yeah in general the fact it does react positively to good team results is good of course).
I think that method you used where you looked at opposition goals too @Trachta10 does do something to reflect opposition level (in comparison to that of the teams the players are playing for), as well as perhaps reflecting the general Goals per Game rates a bit better like you were trying for with it I think.... On the other hand I can understand there would be reasons to not just make that the primary goal contribution calculation method, as then it'd be players of lesser teams and/or who played in teams with leaky defences or in leagues with several great teams of similar level that might get unfairly represented (rather than players who played for great teams who scored a lot of goals from various positions as with the original method arguably). I don't know whether an aggregate calculation could be made or something (considering the numbers with opposition goals and without) - maybe that is too complicated and also becomes a bit subjective and open to debate about how it would be done.
@PDG1978 @Legolas10 I found this website that have Elo rating for clubs not only from Europe,(data since 1911 for Argentina for example) http://soccerverse.com/stats_centre/overview I wrote to the owner and asked him about a way to know the general level of a league He sent me this image this is the raw data of all the leagues (since 1960) https://www.dropbox.com/s/thcflw01cz5hkbz/country_ranking.csv?dl=0# Do you think this data is useful enough to work with?
Interesting to see - I guess my question would be (in relation to the 'delay' I'm talking about), for example how much did the level in the Brazilian league/s really improve between the turn of 1960 and the mid 60s and how much is the spike related to what was actually happening around about 1960 itself. Secondly, would it be down to lack of opportunity (outside friendlies, such as Santos vs Benfica the European Champions in 1961) more than lack of capability to win against the best teams. It took Santos a few years to win the continental and intercontinental titles, but had their team improved considerably by the end of that journey compared to the start of it? Did Flamengo's success cause a similar spike (not related to quality of the average Brazilian team necesssarily) in the early 80s? I'm wondering whether what you are attempting could be more relevant for goals/assists per game than in terms of contribution % anyway? What do you think? I mean, it's natural that in an easier league the goalscoring will be higher, and also that in one-sided games the goals scored by the winning team and it's players can be considerable, but maybe the contribution % solves this in both cases in a way already? I think as well that it could be too easy to 'reward' players who were 'lucky' enough to play more years in top leagues too much at the expense of players who played less years in them but might have done just as well as they later did in the top leagues from an earlier stage had they arrived in those leagues earlier. Maybe Platini would be an example, although I don't know whether the French league was actually quite competitive with the Italian one in terms of ELO in the years before he joined Juventus (I didn't check, so I'm just not sure - I think the Italian league wasn't exactly ahead of the pack in that respect at that time anyway of course, generally).
This from that Excel file from the dropbox seems to have some ratings assigned to the leagues throughout the years. Nothing else It doesn't show how much was the ratings of a team withint the league and others i.e no display of the Local Maxima . So i don't see how can this be used to be honest Also it seems ratings of most leagues have increased now compared to back then. The Brazilian league which is much much weaker than say in 1960s have more ratings than even the best leagues in 1960s . So i am not sure how can this be used. Here you would find Elo ratings for almost all the clubs and every league since long back in days. But it doesn't contain data outside Europe. So unfortunately someone can't do it for players like Pele at Santos , Di stefano at River Plate , Maradona at Boca and Argentinos Juniors etc. But during Maradona's time in Europe , Di stefano at Real and other players in Europe, it can be done from what i see. http://clubelo.com/
Data start from 1977 for Brazil and that's Brazilian League probably, and Brazilian League format changed a lot around these times. The order nowadays is correct tho: The numbers say that the Premier League (not the Russian one) is the most difficult league to play in. (See y-axis.) Numbers never lie, so of course this is correct.Please harass @canzhiye if you disagree. pic.twitter.com/zMMZKuVa0l— Tony (@TonyElHabr) June 18, 2021 Take into account the BIG5 Euro Leagues + Portuguese league is full of foreign players (national foreign and non-europeans, including a lot of Brazilians) and Brazil has the equivalent of three Brazilian leagues of players around the world, and chances are that the Brazilian league had much more level before, specially because leagues like the Argentinian one was deeper depleted before. To be honest I only see the Italian League (since 1988) being superior to Brazilian League until the year 2000.
For all Football fans here- I invite you to come and select the top 10 Football clyb teams forever and ever: https://vote.sparklit.com/web_poll....=e20fa3b3ea2fde554f3521053689ce22b0&preview=1