Originally Posted by doublem23
This is literally the single worst analogy I have ever read because A) only a complete and utter moron would base probablity on mean and not mode. Like, I cannot comprehend how stupid someone would have to be to make such an error and B) obviously a dice roll is the definition of a random event. Even if you rolled a 6 on a 6-sided dice 10,000 times in a roll the odds you'd roll a 6 on #10,001 is still 1/6. Comparing a weighted, statistical model (such as say... PECOTA) to a truly random event shows a real and fundamental lack of understanding of statistics and probability.
Clearly you are unfamiliar with the concept of an "analogy". An analogy is not meant to compare two situations in every possible way. Obviously baseball is not random, and PECOTA is not merely a bunch of averages. That isn't the point. I am not trivializing the difference between "weighted statistical models" and random events any more than I would be trivializing the difference between fish and mammals if I were to say sharks:teeth::whales:baleen. Obviously, These two animals are very different in many ways. The analogy would be pointing out nothing more than the fact that teeth and baleen are located in the same place on the animal's anatomy--AND NOTHING MORE. It does not imply that they serve the same purpose or that they are made of the same material (they don't, and they're not).
I realize that I've just used an analogy to explain the concept of analogies, but bear with me.
Likewise, my die analogy does not imply that PECOTA is nothing more than arithmetic means. I will spell it out for you again: The analogy is that in both the case of the die roll, and the case of executives who place a lot of weight on PECOTA projections, you have a person who is basing an important decision purely on numbers, where there are other factors at play. In the case of the die, the other factor is that 3.5 is not a number on a die. In the case of PECOTA, the other factors are various--managers, team chemistry, fan support, etc, etc.
In a way though, you did grasp my point--guessing 3.5 as a die roll is obviously absurdly stupid. I'm arguing that basing baseball decisions PURELY on numbers is also pretty damn stupid, if not as stupid as the former.