Paulista and Carioca's elo

Discussion in 'The Beautiful Game' started by Ftblabsolute, Jul 13, 2024.

  1. Ftblabsolute

    Ftblabsolute Member

    France
    Mar 17, 2024
    In this post I'll explain how I applied clubelo's elo (http://clubelo.com/) system to the Paulista and Carioca to judge the level of both leagues from 1957 to 1965.

    How does Clubelo's system work

    Firstly I'll explain precisely how clubelo's elo system work, since I reproduced it to the leagues. http://clubelo.com/System on this page the system is explained but I'll clarify it here, and we'll forget about the Tilt part of the system because I didn't take it in account nor is it important for the elo.

    The first thing to do is to set a base elo for the teams of the league you'll analyze. I decided to have a base elo of 1600 for the teams in the Paulista and Carioca. If you find that elo high, on average (roughly) the elo of the bottom 10 teams in England 1st tier in this period is about 1600. I'm basically considering those teams to be mid-table teams or teams that would fight for maintain in England, I find that comparison fair if not even too low.

    Then, the win-probability for a team is given by the following equation:
    upload_2024-7-13_15-50-58.png

    Where Δelo is the difference of elo between the team we're calculating the win probability and their opponents. The elo of the team that plays at home is increased by the "Home Field Advantage", which I set at 50 in my calculations. However this HFA doesn't apply to international games but I'll talk about it later.

    The number of points exchanged (won/lost) according to the result of the game is given by the following equation:
    upload_2024-7-13_15-54-24.png
    Where R is the result of the game (1 for a win, 0.5 for a draw and 0 for a loss) and k a factor that vectors how fast will the results be accurate. Basically, a higher k factor will bring faster results but may be a bit dodgy while a low k factor takes more time to be accurate but will be more accurate overall. Clubelo uses a k factor of 20. I'll give the k factors I used in the next part.

    However, a win 6-0 isn't worth the same as a win 1-0, and that's why the goal difference is taken in account in the final points exchanged. The final number of points exchanged is given by the following simple equation:
    upload_2024-7-13_15-59-30.png
    Where GD is the goal difference.

    Now that I've explained how the elo rating system works, let's move on to the next part.

    Applying the system to Brasil

    As I already explained, the base elo I chose for teams is 1600, but how do we determine the number of elo points allowed in both leagues? The answer is simple, I took games played by Brasilian teams against European teams and applied the elo rating to those games, with the real elo of the European clubs. If Brasilian clubs underperform against European clubs compared to what they're expected as their elo indicates, then they'll lose elo and overall the league will lose average elo.

    Therefore, I applied the system to the top 4 clubs' in both Paulista (Santos, São Paulo, Corinthians, Palmeiras) and Carioca (Flamengo, Fluminense, Botafogo, Vasco da Gama) performances against European teams since 1946. I applied a k factor of 40 to all international games to have faster results as they don't play many games. From 1955, I apply the system to every league game with a k factor of 20, just like clubelo.com. I also applied the elo rating system to Torneio Rio-São Paulo games so that the Paulista and Carioca can balance eachother themselves. Overall, more than 5 thousand games were counted by the elo rating system.

    The results

    I'll now present the results that this system gave me. For better accuracy, I removed/added the points European teams lost/won against Brasilian clubs, so that the performance against them is taken in account even in their average ranking. Here is the ranking of Paulista, Carioca, Torneio Rio-São Paulo (Rio-SP) among the best European leagues every year, according to their average elo:

    *I added the average elo of Paulista's top 16 clubs for the first years because that's the number of teams in the LaLiga at that time

    1957
    Rio-SP - 1810
    Paulista (top 16 only) - 1666
    Carioca - 1656
    England - 1654
    Paulista - 1628
    Spain - 1625
    Italy - 1612
    France - 1623

    1958
    Rio-SP - 1835
    Italy - 1679
    Spain - 1674
    Paulista-TOP16 - 1667
    England - 1667
    Carioca - 1646
    Paulista - 1637
    France - 1579

    1959
    Rio-SP - 1834
    Paulista-TOP16 - 1690
    Spain - 1689
    Italy - 1655
    Carioca - 1651
    Paulista - 1647
    England - 1636
    France - 1575

    1960
    Rio-SP - 1851
    Spain - 1720
    Paulista-TOP16 - 1681
    Paulista - 1668
    England - 1657
    Italy - 1651
    Carioca - 1640
    France - 1605

    1961
    Rio-SP - 1829
    Spain - 1729
    Paulista - 1683
    England - 1655
    Carioca - 1635
    Italy - 1620
    France - 1593

    1962
    Rio-SP - 1811
    Spain - 1732
    Paulista - 1691
    Italy - 1664
    Carioca - 1626
    England - 1621
    France - 1566

    1963
    Rio-SP - 1824
    Spain - 1698
    Italy - 1697
    Paulista - 1695
    England - 1641
    Carioca - 1616
    France - 1581

    1964
    Rio-SP - 1813
    Paulista - 1697
    Spain - 1688
    Italy - 1686
    England - 1661
    Carioca - 1622
    France - 1584

    1965
    Rio-SP - 1821
    Carioca - 1727 (* Carioca reduced the number of teams in its first division)
    Paulista - 1702
    Italy - 1696
    Spain - 1695
    England - 1694
    France - 1561

    Average
    Rio-SP - 1825
    Spain - 1694
    Paulista - 1672
    Italy - 1662
    England - 1654

    Carioca - 1646
    France - 1585

    Conclusion

    Pelé played in the 2nd best league worldwide and Carioca wasn't far away, being 5th on average. The baseless claim that Pelé, because he played in a state league, played against amateur clubs or farmers, is simply false and inane.
     
    Runnec and SaidB94 repped this.
  2. PuckVanHeel

    PuckVanHeel BigSoccer Yellow Card

    Oct 4, 2011
    Club:
    Feyenoord
    The only thing this might show is Brazilian clubs performed very well in friendlies against European teams (plus a couple competitive matches).

    But it was Europe to pay Brazil, not the other way around. You don't pay Pelé to let him show his skills and draw the crowds to then kick him out of the pitch. The other European teams in a (visiting) tour would not like it either. Lots of stories about for example the friendship between Manchester United (Busby) and Real Madrid (Bernabeu) - I'd call that a cartel maybe, but those clubs don't like it if another team kicks all Brazilian stars in a hospital.

    Pelé did play again amateurs, as a matter of fact, while he himself had for that time world class facilities and preparation. That is not to downplay Pele, just see what he did at World Cups when he wasn't even at his best (even if also fortunate sometimes with a tired Italy and Havelange intervening), but it is reality.

    The age pyramid of even the richest Brazilian states of the time look as pure horror. It was pockets of excellence within a world of famine, severe poverty and third world standards. Argentina was already way better in this regard and how many had to live below 2 dollars a day.
     
    TerjeC and SaidB94 repped this.
  3. Ftblabsolute

    Ftblabsolute Member

    France
    Mar 17, 2024
    And if they performed 'very well' then that must means they were atleast on par with them, with some basic common sense. Essentially worse teams cannot perform 'very well' against other clubs supposed to be stronger than them continuously over hundreds of games. There is a reason that most of the time even the biggest clubs like Real Madrid were playing almost every player in the starting XI against Brasilian clubs.
    What is your 2nd point even meant to prove?
    'As a matter of fact', you still didn't give any evidence although I already asked for them. Paulista is a professional league since 1933 and players there became professional in 1933. Having 'world class facilities and preparation' at a time where those 2 elements weren't very present simply shows how professional, how developed and how good was football there back then.
    Again, wealth doesn't have anything to do with level of football. And as another user rightfully point out in another thread, São Paulo was by far the richest state in Brasil. Brasil had lower GDP per capita in 1970 than the likes of Djibouti, that absolutely doesn't mean that people could play football any better in Djibouti. Because again the most important factor is passion.
     
    Kaushik and SaidB94 repped this.
  4. Guts_1069

    Guts_1069 Member

    Cruzeiro
    Brazil
    Nov 6, 2023
    You could tell where the portuguese league ranks and how good they are?
     

Share This Page