can you include me in the group chat?
I'm always on there correcting source feeds and adding missing info... I'm like an unpaid intern...
Youth Soccer Rankings ?
- Thread starter dawson
- Start date
can you include me in the group chat?
I'm always on there correcting source feeds and adding missing info... I'm like an unpaid intern...
RandomSoccerFan
PREMIER
I do wonder how many of us are out there doing the same, whether it's a handful, dozens, or hundreds. I have to imagine the new usage limits have to be hampering some of this, but I imagine it's a necessary tradeoff to combat other datascrapers like all of the new AI bots trying to siphon pretty much everything online.
A 48 cutoff is way too high for a league. Only 31 teams make that line, and seven of them are from socal.
Can you imagine the travel for everyone else? You're talking 1000+ miles for a conference game.
They dont need to play each other all the time. The issue right now is because of closed leagues some of the best teams never play each other.
Who do you want to play?
Are you talking about teams like Sporting, Rebels, and Pats? Or are you talking about Surf, Legends, and Koge?
If it's the first three, you've got a point.
If it's the last three, get in line. Those guys look down on anyone who didn't make the top 8 at ECNL finals. And they have the record to back it up.
Same, and I keep getting my edit access disabled. Seems like they should white-list users after a few back-and-forths.can you include me in the group chat?
I'm always on there correcting source feeds and adding missing info... I'm like an unpaid intern...
Karlton Athletic
SILVER ELITE
I’ve decided the rankings app is cursed. We were playing a highly ranked team today and were predicted to lose 2-1.
They were obviously very good, but it was close and by every usual metric - shots on target, possession, touches in final third etc… we were ahead. We deservedly went up 1-0 with 20 to go, and they’d run out of ideas. Looking good.
Then a fluke freekick somehow went in and we ended up losing 2-1, even hitting the crossbar while trying to equalize.
I blame the rankings app.
They were obviously very good, but it was close and by every usual metric - shots on target, possession, touches in final third etc… we were ahead. We deservedly went up 1-0 with 20 to go, and they’d run out of ideas. Looking good.
Then a fluke freekick somehow went in and we ended up losing 2-1, even hitting the crossbar while trying to equalize.
I blame the rankings app.
RandomSoccerFan
PREMIER
That's the thing that always makes me chuckle as well. The app has no idea about any of the things you listed. All of them are clearly important to actually playing the game competitively and giving your team a chance to succeed. A team that can't do any of those things well isn't likely to be a very successful team.
But all the app can ever do is just look at final scores, and based on them predict future final scores. Anything/everything else is irrelevant. The games that tend to be the most interesting to watch are the ones where all of the stats say one thing, but the score says something else.
What always gets me is if the team is predicted to lose, but ekes out a tie or win. Of course I'm very happy for the team to succeed - but there is that slight twinge of annoyance that the app led me astray.
the dang app is always right!!! And then when you beat a team you were supposed to loose to 3-0 you still don't get a green!!!
RandomSoccerFan
PREMIER
I reported an interesting bug awhile back, and the updated version shipped this week seems to have fixed it. They had an unexpected and interesting explanation. The bug was this: in games that were predicted to be extremely close, like between teams that are 1 spot different nationally, the prediction percentages were weird. Like sometimes it would say that team A should beat team B 1-0, but also team A had a 37% chance to win, team B had a 39% chance to win, with a 24% chance of a tie. It didn't make sense. It turns out that since a tie is much less likely to happen than a win for either side, it would randomly choose one of the teams to be the winner, and sometimes it was the one with even slightly less percentage to win. That was a higher likelihood outcome than a tie.
They've decided to change the app behavior, so now it will be more likely to suggest a tie as the result (even if mathematically it should be significantly lower), and it will no longer suggest a winner that has a lower percentage to win than the other side. I don't think I've ever seen the bug prior, but I have come across it twice in the past few weeks - sounds like I shouldn't expect to see it again.
They've decided to change the app behavior, so now it will be more likely to suggest a tie as the result (even if mathematically it should be significantly lower), and it will no longer suggest a winner that has a lower percentage to win than the other side. I don't think I've ever seen the bug prior, but I have come across it twice in the past few weeks - sounds like I shouldn't expect to see it again.
I wonder if this was some of the weirdness I was seeing. It sounds like in close tie situations the ranking app was giving a predicted win to the team less likely to tie. When it should have reported a tie because less likely to tie doesnt really make sense.
Unfortunately Im not looking at the numbers that closely anymore so I cant confirm. But I do like simply showing a likely tie vs giving one team a predicted win just because theyre less likely to tie. Teams shouldn't be getting a psychological win when the reality is a tie is the most likely outcome.
RandomSoccerFan
PREMIER
That's the conundrum. Even if teams are exactly matched, a tie isn't the most likely outcome. It will more likely end up with one team beating the other. But I don't think it was the right answer to choose between them randomly if it really was equal whether either would win. I think predicting a tie, or failing that, choosing the team that is ranked higher (even if infinitesimally), is probably the better option - I'm glad those are the options they went with.