This reminds me of the models I made a few years ago to predict where a Frisbee will land after you throw it. I made about 14 different models, some of them co

This reminds me of the models I made a few years ago to predict where a Frisbee will land after you throw it. I made about 14 different models, some of them correctly predict where the Frisbee will land. That's right, I have a model that can tell you where the Frisbee you are about to throw will land, it correctly predicts Frisbee throws. Here is how it works: You input all the factors of your Frisbee throw. Then you throw the Frisbee, and that's it! Well, I guess there is one more step, since all 14 models make wrong predictions about where the Frisbee will land. But you can just go back and adjust the models and input better data after you already know where the Frisbee actually ended up, and then some of them correctly predicted where it will land! Don't down-vote me, this is definitely what the phrase 'correctly predicted' means, and I'm not being misleading at all.