I'd love to see that! The chart that got posted to meme earlier with the models' predictions versus the pundits' predictions had the models in order of how they've fared so far, and @DTMAboutheart's is third. I believe the two above him also have only been "wrong" on one each, but the rankings reward/punish the "confidence" of the model (so a model that predicted 60/40 Sharks vs Blues ranks gets more credit than one that predicted 51/49).
@IneffectiveMath made a good point the other day though that the models are not really predicting winners, they are predicting probabilities. It seems a little unfair to say a model predicted "wrong" if something happened that the model said had a 45% percent chance of happening.
![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
Re: Game Day Thread: June 1, 2016
From: (Anonymous)@IneffectiveMath made a good point the other day though that the models are not really predicting winners, they are predicting probabilities. It seems a little unfair to say a model predicted "wrong" if something happened that the model said had a 45% percent chance of happening.