The World Is Not That Predictable

Despite “deep learning” and “machine learning” the world is not that predictable. When I woke up this morning it was raining steadily, precipitation that had not been forecast. Again, maybe it’s the nature of desert weather, but the National Weather Service is not that accurate in predicting rain here.

When the Russian dictator launched an invasion of sovereign Ukraine in February, most observers thought it would not last more than a few weeks. Here we are, about seven months later and it appears (take all reports with some skepticism) that Ukraine actually has the upper hand, at least as of now.

Even in the context of a trivial, and seemingly simple, matter such as this blog, I cannot predict readership. Yesterday saw a strong number of views (59% more than the average of the last five days with a post) and visitors, including a fair number for this post–250,000th Corvette and Other Things–that I published almost three years ago. When I mention trends in readership in conversation I am often asked, “Why did that happen?” I always answer that I have no idea.

Data analysis is powerful, but not omniscient. I think of the remark attributed to Albert Einstein, “Not everything that can be counted counts and not everything that counts can be counted.” I don’t interpret that statement as we can’t ever know the structural equations of a model so the reduced-form equations cannot always be accurate. I think of the statement as a broader, almost metaphysical, acknowledgement that no one knows everything about anything. Einstein’s own futile search for a Unified Field Theory shows that not even he knew everything about Physics.

******************

Speaking of trivial…I have reached the three-quarter mark in my computer football season. This is the longest such season I have ever attempted (216 league games, plus playoffs; a real NFL season has 272 games plus playoffs). Perhaps that is why I am becoming a little bored with the whole thing. I just want the playoff races to be settled and to play the post-season and crown a champion.

Fortunately, the simulation allows me to set up any game to be played solely by the computer with the results becoming known in seconds. I hope this displays OK; below are, hopefully, the current standings.

Sorry for the screen effects. I tried to copy and paste from the actual file, but the WordPress editor cannot seem to handle fixed-space fonts, as opposed to proportional fonts.

I cannot vouch for the accuracy of Home, Divisional and Conference records. Despite being on the market for more than 20 years, this game still has more than its share of bugs.

Being the weirdo that I am, I would be happy if an AFC team makes the playoffs with a 9-9 or 8-10 record. The Buffalo Bisons began the season 0-5, but are still alive for the playoffs; that appeals to me, for some reason. The playoff format is like the original one post-merger (1970) in that the three division winners and one wild-card team make the playoffs in each conference.

No, I do not think that these results are important to anyone except me. No, I am not delusional and think that this is real life. Once again, I have been playing statistics-based (as opposed to graphics-based) sports games since I was 12; it’s in my blood, I guess. I even played these games while I worked in major league baseball although not as much as before or since.

Do any of you have any “lifetime” hobbies that you’d like to share?

#TheWorldIsNotThatPredictable

#somanyCARSjustonelife

#disaffectedmusings

If you like this blog please tell your friends and share the blog URL (https://disaffectedmusings.com). Thanks.