RE: Model Madness 2019 - Estimated Difference Likelihood Model
Firstly, a calculus post that got upvoted by @curie, what are the odds of that? Can we get a calculation on that? Thanks!~ Secondly, a post that makes a practical application of a theoretical equation and not only as a hook but a means to explain a thing? My oh my, now it just becomes getting juicier each and every second I type this comment out. Thirdly - okay let's cut the joking out now
Now I never really cared for televised sports events (mainly the anchor people that commentate on the event's proceedings) and so I just never paid attention to events. What I did knew of was how people decided to math sports (with good reason) and make sense of trends, which in all honesty should be looked into as this post doesn't make light off (and for good reasons that I appreciate that for). Also I must give gracious thanks for not only collecting a whole sum of data, a reference model but also a set of three different models - time-consuming for sure, thankfully computers can spit out code like the bad-machines they are!~
Also the topic of this post is rather interesting in my eyes, truly. Not because it literally is a prediction model based upon calculus, but more-so the fact of what is being strung together as I continue to look at it. To Calc one me, that would've looked like some hodgepodge of a mess before I realize how messy series can get and the trickery calculus nerds do to get any connection between data points.
The examples seem rather fine, as aforementioned, and seem reasonable - so do scratch the data collected on some part but keep it for the rest. Now I just do wonder if there's a model for models that calculates their range of error over time which then leads into meta analysis... which finally leads to fixing up models or making stronger ones and running the data again all over again. Fun with statistics being backed up by Calculus in the auxiliaries~