Under the subject line, “A brief note on one of your favorite topics: measurement,” Jonathan Falk points to this post from the evil sports gambling company Draft Kings:
Falk writes:
Best comment (among a few hundred idiotic ones):
Bad news. They later found that home plate, due to a manufacturing error, was 7 nanometers too wide, so the automated system, expecting the normal size for home plate, was actually incorrect.
The thing about the automated system that I’ve never seen written about is that it makes judging by a human whether a pitch is too high or too low impossible. Where is the top of the strike zone? If the system is in use, the height of the player is measured during spring training and some fraction of that – say 56% (not the actual number) – is defined to be the top of the strike zone. There’s no allowance for the stance the batter takes.
So an umpire would have to know for each player what that number is, and visualize where it is in space. That’s an impossible requirement. Nor can a pitcher know exactly where it is, either. No wonder some pitches are called wrongly! It’s remarkable that the umpires do as well as they do.
I used to play baseball and that came with the task of being an umpire at lower levels. I hated doing that, one reason being the task of calling a strike or ball with less than clear rules for my liking and with a less than optimal point of view. I was a pitcher so I also know about the issues about calling a strike or ball from the other perspective.
We know that people are susceptible to cognitive biases. Sports betting on a cell phone with all sorts of deals and parlays. The companies are gaming this using Thaler Nobel Prize techniques to encourage bad decisions. Isn’t it time to ban online gambling in light of 2001 and 2017 Econ Nobel Prizes?
Danielle,
While we’re at it, I wouldn’t mind banning the Nobel prize too!
Shouldn’t that be:
“While we’re at it, I wouldn’t mind banning the Nobel prize too! – nudge, nudge”
Stating the obvious: 4 nm doesn’t tell us anything about the accuracy of the system. If the measurement is repeatable, that would say something about the precision of the system, but not the accuracy. But by itself, with one measurement it just sounds like reporting insignificant digits.
I’m struggling to understand what is considered the ‘ground truth’ for this claim of accuracy.