RandomSoccerFan
PREMIER
I’m not going to call out a team for sandbagging so I will not cite examples.
So it happens "all the time", you just can't think of one or don't want anyone to show how your perceptions don't match the mathematical reality. Got it. Here are the top 20 California teams. Which ones are gaming how they enter tournaments to effectively adjust their rating as you are accusing? 4 screenshots are 2010G California, 2010G US, Club CA (G), and Club US (G).




I get it, you want the soccer rankings app to be 100%.
Nope. But the same folks who continue refuse to understand why their team isn't ranked higher tend to be the same folks who want to discount it entirely. The way they do that is by saying "Yes, seems OK in general, but since the teams don't actually play each other I'm going to substitute my own biased judgement when convenient and nobody can prove otherwise". No ratings app is going to be able to predict the future with 100% accuracy. But the better one is, the more likely it is that a team with a higher rating will beat a team with a lower rating. Of course individual games are going to go differently than expected. Of course those variations are going to adjust the ratings of teams up and down. And of course when the ratings between two teams are close - the likelihood of picking the right winner, bottoms out at a coinflip.
As everyone is aware, there is no award for showing up #1 in these ratings, #20 in these ratings, or #2478 in these ratings. Inherently, they don't mean anything. The only thing they represent is that based on past game performances, this is how they would be expected to perform in future game performances. If the performance then differs from the rating, the rating continues to adjust up and down with every game. Keep in mind much of the value of a system like this is for predictions against teams that haven't played each other before recently (or ever). It's what gives it value - compared to just looking at an individual league standings table. That league table is completely sufficient if the belief threshold is that teams must see each other directly before a prediction can be considered real - it just doesn't do much good when looking forward to tournaments against teams that aren't in the same league/geography/direct game history.
It’s ok for there to be some type of variance of accuracy in any application or algorithm, that’s normal. Scientist would agree, mathematicians and even physicians would say to be skeptical and challenge everything until it’s proven to be Law.
Yes, completely true as stated. But many instead interpret it to mean "Scientist would agree, mathematicians and even physicians would say to be skeptical and challenge everything until
This app isn’t Law. There’s going to be some faults, and that’s ok. When you say Top Hat is #2 in the country and the only team we can use to cross compare results is Charlotte SA ECNL who they tied 1-1 I don’t put them at #2 in the country. Maybe top 10, maybe top 20, but not #2. Sorry.
And that's the ballgame. If you're never going to believe the results unless you can point to teams that have played each other directly, this isn't a resolvable discussion, and you misunderstand and/or don't value the primary benefit of what you're looking at. And in this particular case - until TH 2010G plays SW CA teams - minds aren't going to be swayed.


