RandomSoccerFan
PREMIER
boys national leagues (U15)
girls national leagues (U15)
girls national leagues (U15)
Youth Soccer Rankings ?
- Thread starter dawson
- Start date
What is see is for u15....
For boys 52-56 regardless of league should be playing each other.
For girls 48-52 regardless of league should be playing each other.
But wishful thinking is one thing. Reality is another.
RandomSoccerFan
PREMIER
Agreed in full. In the past, it was much more likely to have all of the top teams show up in tournaments that meant something. With the way things have progressed, it's that much less likely that a top team in any of these leagues will choose to face a top team in one of the others. Even if we think we have a good idea about what the results would be / should be, the point is that the teams should still be battling it out on the field.
rainbow_unicorn
PREMIER
Best graphs I've seen to explain the difference and overlap of leagues.
rainbow_unicorn
PREMIER
@RandomSoccerFan - Can you do this for all U13 - U18 ages? Curious to see if trends change as they age.
RandomSoccerFan
PREMIER
He mentioned that they had run it for U17 as well, and the trends were almost identical. I'd imagine they will post that one up on their social media at some point as well - but it apparently doesn't provide any new or different information compared to the ones for U15 here.
It’s too early in the season. A lot teams are still “unranked”. Also in SoCal, ECNL teams are currently ranked higher than MLS next HD so not sure why nationally it is showing a different trend.
RandomSoccerFan
PREMIER
Go younger. But then again u15 and u17 there is almost no cross pollination (they don’t play each other and not even with teams outside of their own league), so the ranking validity is questionable to start with.
RandomSoccerFan
PREMIER
You complained that you were interpreting the national trend differently, then pointed out that you were looking at the wrong age group, and then complained that you don't believe the ratings anyway? You're a hoot.
You agree that u15-17 don’t play against anyone outside of their closed league anymore? So the 627 games between mls next hd and ecnl are mostly for younger kids.
The bottom line is teams need to play each other to have a ranking.
RandomSoccerFan
PREMIER
I agree with very little of what you post. I'd in fact say that most shouldn't, but that's up to them. SR believes there is enough interplay at all levels.
No, teams need to have enough games against other ranked teams to have a rating/ranking. Your bottom line for what you choose to believe can be anything you want it to be.
Use common sense. You can’t rank two populations who never even come across each other. Let’s say after u15, mls next and ecnl don’t play each other anymore in tournaments. Sure you can carry on with their ratings from u14 and continue to curve fit into u15, u16 to u17. You come up with a ranking. Who is gonna believe that?
RandomSoccerFan
PREMIER
It's your attempt at using "common sense" that's leading you astray. You have it set rock-solid in your head that until two teams play each other directly, using additional related information to try and predict the result is not to be trusted. That fundamental belief might sound reasonable to some, or even many, but it is entirely false. A prediction can be made within a set confidence interval, and even better, the quality of the prediction can be directly measured afterwards - by looking at the actual outcomes.
Put all soccer teams in one bucket. All ages, all genders, all locations. Keep track of results between them. It turns out that a surprisingly small amount of results, as low as 5-6, develop a model that allow the 7th game to be predicted well within the appropriate confidence interval. It feels good - and is easy to understand - when you mentally apply the model to a small group of 8 teams and follow it through. But youth soccer is so predictable, that it works just as well outside that model, across the entire population. It's just harder to wrap your head around.
You can split these thousands of teams into a variety of cohorts, whether age/gender/geography/league, and you can report out how they compare to each other internally and externally to that subgroup. It turns out that you certainly can rank two populations who never come across each other. And the results will still be pretty good. But - the small amount of interplay that does exist, serves to both verify ratings across these subgroups, and also can help calibrate however necessary. You have an incorrect belief here that there must be a very significant amount of interplay, or the results can't be "trusted". It's just not true.
Look, you can be skeptical as much as you want. I'm done trying to convince you of anything - it's pretty clear you're never going to change your mind. These discussions are for those reading to come to their own valid conclusions.
Put all soccer teams in one bucket. All ages, all genders, all locations. Keep track of results between them. It turns out that a surprisingly small amount of results, as low as 5-6, develop a model that allow the 7th game to be predicted well within the appropriate confidence interval. It feels good - and is easy to understand - when you mentally apply the model to a small group of 8 teams and follow it through. But youth soccer is so predictable, that it works just as well outside that model, across the entire population. It's just harder to wrap your head around.
You can split these thousands of teams into a variety of cohorts, whether age/gender/geography/league, and you can report out how they compare to each other internally and externally to that subgroup. It turns out that you certainly can rank two populations who never come across each other. And the results will still be pretty good. But - the small amount of interplay that does exist, serves to both verify ratings across these subgroups, and also can help calibrate however necessary. You have an incorrect belief here that there must be a very significant amount of interplay, or the results can't be "trusted". It's just not true.
Look, you can be skeptical as much as you want. I'm done trying to convince you of anything - it's pretty clear you're never going to change your mind. These discussions are for those reading to come to their own valid conclusions.