What are the chances?

Somewhat off-topic post tonight.

I pulled up my usual weather app this morning and noted an interesting anomaly.

Why would the probability of rain be exactly 65% for such an extended period. (I'm guessing it's really 66.7% or 2/3, but that the app rounds to the nearest 5%).

The "chance of rain" is actually complete fiction. Pin any meteorologist down on this if you don't believe me. Is it the likelihood that it will be raining at that very instant? Is it the probability that some precipitation will be measured within some fixed time interval? Is it the odds you would take or give if betting? Is it the historical percentage of observed rain when similar pre-conditions are observed?

No, it's none of those. Why? Because most people don't have a clue how probability works. Look at the last presidential election. Nearly all the polls (and certainly the average of the polls) indicated that there was a 20-30% chance that Trump would win. Yet, people still marvel about how "wrong" the pollsters were. They weren't wrong. A 20-30% chance is the same chance a major league baseball player has of getting a hit when they come to the plate. Do people think that getting a hit in baseball is some sort of bizarre rare event? Of course not. It's just less likely than getting out.

But, if you tell someone that the chance of rain is 75% and they cancel their picnic plans, they'll be mad if it turns out to be a nice day. So, the weather folks play this silly game of trying to come up with a "probability" that conveys what they really mean, even though it has nothing to do with actual probability.

I can't know for sure, but my guess is that the algorithm that generates probabilities for the weather app I use has some sort of harness that prevents the chance of rain going beyond 2/3 in certain conditions because people will interpret that as "It's definitely going to rain."

Interestingly, a couple hours later, they had changed it to look a lot more like a real super-accurate prediction (any honest meteorologist will also admit that the difference between 50% rain and 60% is meaningless more than a few hours out). Clearly, the same algorithm is being used as the shape below the bound is identical, but the bound has been removed.