This is clearly wrong. Some teams are way better than others (as evidenced by the fact that they win more). The question is by how much.
Right, and why would this be surprising? Because generally, the opposite is assumed to be true? Why? Because we all think LAFC is better than TFA and Galaxy is better than Legends, so it stands out when results go the other way.
But this objection comes from a misunderstanding of how statistical models work. When a model predicts that team A will beat team B 3-1, this (obviously) doesn't mean that A will always beat B by that score. It's just that this is the most likely outcome based on past performance. The prediction will also come with a confidence level which describes mathematically how likely it is that this outcome will happen. When a specific outcome doesn't match the prediction, this doesn't mean there's something wrong with the model. You can't judge the model on individual outcomes, only on the aggregate of all the predictions it makes, so, even if we see anomalies like those mentioned above, if, over time, the model predicts the correct score more often than not, you have a good model.
Now, I agree that applying these methods to kids soccer is a little dubious, but only because kids themselves are less consistent than pros and (the bigger issue in my book) there's so much turnover in these clubs. However, if you look at
@Kante's predictions over time, they're pretty darn accurate.
@Kante do you have any stats on how accurate your predictions are?
