WTF! A Look At Copa America

Discussion in 'Copa América Centenario 2016' started by Asalieri, Jun 9, 2016.

  1. Asalieri

    Asalieri Member+

    Jun 29, 2004
    Mesa, Arizona
    Club:
    CA San Lorenzo de Almagro
    Country:
    Argentina
  2. PresidentClintonDempsey

    Apr 1, 2016
    Deuce States of Mexicanada
    Country:
    United States
    Little rant of mine: I HATE the pandering of West coast matches being held at roughly 7PT while here the east coast folk have to lose some sleep just to watch an admittedly subpar match for 1'45". It's a dick move; the only solace is the fact that here in "New York" (actually New Jersey because I live in Manhattan) the final will be held. Mexico just beat Jamaica 2-0: the game ended 30 minutes ago and right now at the time the comment is posted it's already 12:45am. Rant over. Mexico campeón!
     
  3. jass

    jass Member

    Oct 12, 2006
    Club:
    Parana Curitiba
    Country:
    Mexico
    Its Copa America. East Coast is not the target market, south america is.

    Amusingly, that means that a 10pm EST kickoff is 11pm in Brazil and Argentina! For reference, Libertadores games tend to kick off at 9:55pm local time.
     
  4. PresidentClintonDempsey

    Apr 1, 2016
    Deuce States of Mexicanada
    Country:
    United States
    wrong, the target market is the host nation itself. CONMEBOL could've held the tournament in Argent like it did 100 years ago. Obviously money was the biggest factor.
     
    MadridRojoRowdy repped this.
  5. jass

    jass Member

    Oct 12, 2006
    Club:
    Parana Curitiba
    Country:
    Mexico
    Wrong. TV market is what matters. Why do you think the world cup games in brazil had kickoff times that were designed for europe and not brazil?

    Do you think they gave any thoughts to Chiles tiny market when it was hosted there?
     
  6. Cliveworshipper

    Cliveworshipper Member+

    Dec 3, 2006

    You'd rather see a 4:00 West Coast start so you can pass out in your chair with a beer After dinner while West Coasters are still at work?

    Seem like kind of a dick move.
     
  7. Asalieri

    Asalieri Member+

    Jun 29, 2004
    Mesa, Arizona
    Club:
    CA San Lorenzo de Almagro
    Country:
    Argentina
  8. Abram Jones

    Abram Jones Member

    Jun 18, 2016
    Wisconsin (WI)
    You're basing your theories off of single games, or a handful of games, this is not scientific. FIFA and other elo rankings aren't perfect, and could certainly be improved, but they are definitely better than people give them credit for. Please see law of large numbers:
    https://en.wikipedia.org/wiki/Law_of_large_numbers
     
  9. Asalieri

    Asalieri Member+

    Jun 29, 2004
    Mesa, Arizona
    Club:
    CA San Lorenzo de Almagro
    Country:
    Argentina
    there's no science to football anyhow. And those rankings rarely if ever reflect the true efficacy of a team. IF they did, you wouldn't put Jamaica so high above a plethora of teams that regularly kick their asses and have better history against them.
     
  10. Abram Jones

    Abram Jones Member

    Jun 18, 2016
    Wisconsin (WI)
    How do you figure there's "no science to football" that doesn't make any sense. It is observable and measurable, so it can be evaluated in scientific fashion. Each game is like an experiment, but I think you missed my overall point. Generally such rankings are accurate (again, not perfect), various forms of elo rankings are the best rankings we have (i'm sure some are better than FIFA rankings). The reason why they are more accurate than other methods is because they collect data from a large amount of games and their results. Certainly there are overrated and underrated teams, but there are many factors. A short tournament like Copa America or the World Cup, though exciting, is not a mathematically sound way to measure rankings because of a limited data set. This is why in playoffs for NBA, NHL, and Major League Baseball there are 5 or 7 game series between 2 teams: the more games you have the more of a chance it reduces flukes. Especially a game like soccer this is a factor because of the low amount of scoring opportunities (this increases the chance of undeserved upsets). This is a simple mathematical principle. See the Law of Large Numbers, as I mentioned previously. A similar issue these different elo ranking systems place is the lack of games, this decreases their chance for accuracy, but are still working with more data so they are generally much more accurate than a single tournament in deciding true rankings. If internationals were played more regularly like league games it would be easier to rank teams more accurately. Law of Large Numbers :)
     
  11. Rickdog

    Rickdog Member+

    Jun 16, 2010
    Santiago, Chile
    Club:
    CD Colo Colo
    Country:
    Chile
    As how you finished this video, it seems one of your "future second placers" actually won the whole thing, vs. the number 1 ranked team of the world, .......for the second time in a row. :whistling::p

    One thing you are right though : FIFA ranks suck.
     
  12. posteador

    posteador Member+

    Dec 29, 2006
    Lincoln, UK
    Unfortunately we have to take FIFA rankings seriously now since they determine the seeded teams... and getting to be a group leader definitely helps.
     
  13. oliveandblue

    oliveandblue Member

    Jul 7, 2015
    United States
    Club:
    DC United
    The thing is, the USA - at #31 - just scored wins over #23, #13 (!), and #44 (and yes, Paraguay are badly underrated).

    #1 Argentina SHOULD trash them. #3 Colombia SHOULD win 2-o, and then 1-0 (which was more even of a game).

    Results-wise, the USA behaves like an anonymous World Cup level team. A team that gets to the big stage, and then loses to the REAL contenders.

    Back-to-back R16 finishes are quite strong results for a country that doesn't really know what they're doing.
     
  14. Abram Jones

    Abram Jones Member

    Jun 18, 2016
    Wisconsin (WI)
    #14 Abram Jones, Jul 2, 2016
    Last edited: Jul 2, 2016
    This is not how mathematics works. Elo rankings aren't attempting to decipher who is actually going to win in 1 game (or even more ridiculously what the final score is going to be), they attempt to decipher who has the probability of winning. What you are doing when cherry picking these examples is the equivalent of a scientist making a hypothesis and then running 1 half assed experiment, that's not how it's done. In order to prove something experiments have to be run many times in many situations. In other words, if Team A beats Team B in one game, that doesn't really tell us anything. Team A and Team B should play multiple times as well as multiple times against a variety of opponents. This is how elo rankings are decided. This is why elo rankings are generally more accurate than what a single tournament standings tell you (no matter what the common folk think). See law of large numbers. And again, I'm not saying elo rankings are perfect, they are just the best method we have to determine rankings. Another problem to consider that will make it harder for elo rankings, as I mentioned before, is the lack of internationals played. The lack of scoring opportunities in a soccer game makes these even more difficult (this creates a higher chance of undeserved upsets).

    A final note about the strength of the USA team. People under estimate the USA team all the time because of the perception that soccer is not big here because it is our 5th popular sport. This is an error in thinking because of the size/wealth of this country. We have approximately 13.1 million potential wealthy people contributing directly or indirectly to the infrastructure of the sport of soccer, this is much higher than most countries, this is whey we can out perform most smaller countries where soccer is their number 1 sport. This is why international sports are pretty ridiculous in their current form.



     
  15. oliveandblue

    oliveandblue Member

    Jul 7, 2015
    United States
    Club:
    DC United
    Elo ratings that I've seen set the USA in the mid 20s right now. Of course they're not a predictor of future results, though. It's a power ranking of nations.

    Also, there is no such thing as an "undeserved upset". A bad bounce of the ball is a FAIR event.

    You cannot quantify how well a team played. Possession, SOG, and pass completion rates tell parts of the story - but I've seen teams hold 70% of the ball, outshoot the opposition, and COMPLETELY DESERVE TO LOSE.

    Another problem that you have is you are trying to quantify something that can only be qualified. You cannot calculate an expected result in international football. (The bookies place odds to make money off of the opinions of bettors.)

    Also, due to tactical quirks and the vast variety of styles/climates/weather conditions it is utterly pointless to talk about the law of large numbers. A soccer game is not an event that can be repeated - every single game is a one-off event. Every game is entirely unique. You're trying too hard to calculate what can only be loosely judged. You need to be able to control more variables for your logic to ever hold up.

    Mathematics is not stronger than any opinion off the street for these reasons.
     
  16. Abram Jones

    Abram Jones Member

    Jun 18, 2016
    Wisconsin (WI)
    #16 Abram Jones, Jul 2, 2016
    Last edited: Jul 2, 2016
    Of course there is a such thing as an undeserved upset. You are making way too many assumptions. I did not say possession, sog, and pass completion tells a complete story, of course it doesn't. But that doesn't mean there is no such thing as an undeserved upset. If there are a limited number of scoring opportunities (other dynamics are required to get to this point), that means the experiment is incomplete. In essence what you are saying is playing one game is a more true result than playing a series of games to determine who is a better team.

    And no, it is not pointless to talk about law of large numbers at all. The reason why more matches and scoring opportunities are needed is exactly because of all these "quirks" you speak of, and those that you don't either (such as health/mood of players). I think you have completely missed the point and are having trouble understanding the concept. You could just as easily say "every science experiment is a one-off event" or in the other direction "every coin toss is a one-off event" It's complete nonsense. The law of large numbers applies to probability theory in our reality here in this dimension on earth :) for you to say that soccer games are an exception to this is laughable at best.
     
    oliveandblue repped this.
  17. oliveandblue

    oliveandblue Member

    Jul 7, 2015
    United States
    Club:
    DC United
    Everything in theory is a one-off event. We just try to control as many variables as possible in order to find trends. Scientific experiments do a better job than most at achieving this. Soccer doesn't.

    The problem in soccer is quite simple - with so few games, it is VERY easy to fall victim to the Clustering Illusion. https://en.wikipedia.org/wiki/Clustering_illusion

    International sides do not play often enough to even create a consistent baseline of performance - you don't know if you're looking at a random "positive/negative streak". That is why people use results in recent tournaments - ranking systems all suck! There really isn't a way to produce a good ranking system in international soccer, these are not club sides that play way more common games under common circumstances.

    You aren't wrong in thinking that tournament results are flawed and possibly even ineffective at determining side strength. I just don't think that FIFA rankings - or even Elo - do any better. By the time you have enough data points to build a trend, things shift and you're fricked.
     
    Abram Jones repped this.
  18. Abram Jones

    Abram Jones Member

    Jun 18, 2016
    Wisconsin (WI)
    Well you just sort of contradicted yourself here, because clustering illusion is related to the law of large numbers. the clustering illusion is exactly what can happen in a single soccer match because of the limited of chances (game laws make it too easy for defense).

    And yes you are on to something here, this is what I was saying before about too few internationals. But the elo systems aren't what need to be judged as much as this fact, their systems are doing the best they can with the limited data they have. So yes, there are many problems that the elo rankings must face and can't mathematically deal with, but still are the best ranking system we have. The real problem is that not enough internationals are played (as I mentioned before unless that was in a different topic).

    And of course a proper scientific experiment does science better than a soccer match :) but the same mathematical/experimental principles apply... and soccer matches are in a somewhat controlled environment, unlike military battles. And even in those cases we can use statistics to predict combat results as shown by Colonel Dupuy. He could even go so far as to show the Germans had better soldiers than the Americans, British, and French in WW2 despite all of the dozens of complex factors.
     

Share This Page