1) Reversion to the mean. This is what the above person believes and that this likely won't happen again.
2) Indication of a trend. The pilot is actually incompetent, and this will happen more frequently with this pilot than an average pilot.
Regression to the mean is simply that any given datapoint is most likely to be the mean, or close to it. This means that any exceptional data point,up or down, can be expected to be followed up by the mean.
The example of this being misinterpreted that I am familiar with is that of a flight instructor's belief on training. When a pilot performed well, they wouldnt comment. If a pilot performed poorly, they would be punished. They believed this was better because when they praised a pilot, they would usually do worse the following run, and when they punish them they do better. This isn't technically wrong, they are just ascribing causation where there is none. I think I saw this example in Signal vs. Noise, but I'm not sure.
Basically, regression to the mean isn't a reason to pick someone who did poorly, it's a reason that that person will do no worse or better than they do normally.
https://www.google.com/#&q=%22how+to+win+friends+and+influen...
Flip 100 coins. Take the ones that 'failed' (landed tails) and scold them. Flip them again. Half improved! Praise the ones that got heads the first time. Flip them again. Half got worse :(
Clearly, scolding is more effective than praising.
While, of course, human performance is not always equal in the same way our coins are, the fact that performance (or whatever it is you are measuring) will still regress towards the mean still holds true. The coin example just gives an extremely obvious demonstration about what is happening when things regress towards the mean.
If labour is so abundant that there's little problem in replacing any given worker, people can and will be fired for trivial offences. If there's a shortage, or there's a considerable on-boarding process, less so.
The threat of a firing-at-any-moment also makes for a more tractable workforce, at least from a management perspective.
Humans aren't perfect improvement machines, but they surely beat a random variable?!
Neither the GP's quote, nor the OP, are making the statement "I understand that this mistake was an outlier, you'll probably never make this mistake again if you remain unchanged".
The claim being made is that acknowledging the mistake and learning from it can dramatically reduce the base probability of such mistakes happening again.
The question is: is the day-to-day random variation much bigger than the day-to-day average increase? If so, regression towards the mean makes total sense.
3) Out of remorse, pride, being shocked out of complacency or some combination of these, the man would do his absolute best the next day, especially for the very same pilot.
Not everything is a financial metaphor.
This is psychological insight, not statistical insight.
I think more appropriate in this context is learning though. And I would suggest here multi-armed bandits (with every arm representing a pilot).
https://en.wikipedia.org/wiki/Multi-armed_bandit
There is tradeoff between exploration and exploitation that has to be made.
The goal in this case corresponds to a particular type of bandit. It is to postpone death for as long as possible by pulling the right arm. I actually didn't find this type yet (a mortal multi-armed bandit has a birth-death process of the arms themselves).
Edit: This is only about the learning process on the non-pilot side, as one of the other commentators already articulated.
It would seem to me (somebody who knows squat about stock trading) that a warning screen would be much more beneficial for stock trading software than it is for an online video game.
I'd imagine in the case of the stock trader, where quick reaction is valued, he'd have quickly entered whatever key combo is necessary to dismiss it the warning and made the same mistake.
Alarm fatigue is an important design consideration, but not a reason to not put warnings at all (given that this isn't amenable to an "Undo" button, which is preferable when possible).
You shouldn't fix an issue like this with UI validation alone - it needs to go between the component that creates orders and the stock exchange, so that it also protects against software bugs. For instance, a possible bug is a developer multiplying the order size by the lot size in the backend, when it has already been multiplied in the frontend, causing huge orders to be sent. Sanity checks can catch this.
Automated trading systems have traditionally been under a lot of scrutiny, and nobody in their right mind would run one without sanity checks and a kill switch. That incident taught Mizuho that manual trading can, in fact, also be quite dangerous :)
"A perhaps-undesired recognition is the Hoover Nozzle used on jet fuel pumps. The Hoover Nozzle is designed with a flattened bell shape. The Hoover Nozzle cannot be inserted in the filler neck of a plane with the Hoover Ring installed, thus preventing the tank from accidentally being filled with jet fuel.
This system was given this name following an accident in which Hoover was seriously injured, when both engines on his Shrike Commander failed during takeoff. Investigators found that the plane had just been fueled by line personnel who mistook the piston-engine Shrike for a similar turboprop model, filling the tanks with jet fuel instead of avgas (aviation gasoline). There was enough avgas in the fuel system to taxi to the runway and take off, but then the jet fuel was drawn into the engines, causing them to stop.
Once Hoover recovered, he widely promoted the use of the new type of nozzle with the support and funding of the National Air Transportation Association, General Aviation Manufacturers Association and various other aviation groups (the nozzle is now required by Federal regulation on jet fuel pumps)." [1]
[1] https://en.wikipedia.org/wiki/Bob_Hoover#Hoover_Nozzle_and_H...
i am a huge fan of the turn of the century dancer isadora duncan and her lover who she first had a child with, the theatre set design theorist, edward gordon craig who she affectionately called endymion
a complete aside, but if you have, or anyone reading this has, yet to read duncan's autobiography 'my life' i highly recommend it to anyone and everyone
she is a brilliant writer, lived an eccentric life, and she was and her writing is imbued with a mad passion for expression and both life's hardships and joys
Or under external pressures they have no control over.
So I'd say it depends on your environment combined with the individual. Someone who is apathetic and/or lacks critical thinking skills will probably learn very little beyond avoiding that specific mistake again.
Shows the other meanings of the homophones - English huh!
A pattern of making mistakes can tell you somebody is careless, or sloppy, or in over their head. But a single mistake? That's just inevitable.
Or that you need to fix the process.
When that happens, I hope the people around you are more forgiving than you are.
I think it's especially true in tech jobs where there is a lot learned on the job. I would be deeply skeptical of any tech company that takes a fire first attitude. That just tells me they 1) treat devs as disposable and 2) they get rid of everyone that has a chance to learn from their mistakes
More generally it is very difficult to build highly reliable systems of any kind without having a very open culture where you focus on exposing problems and fixing them rather than blaming the message.
Everyone was chipping in with their theories about why the U.S and Silicon Valley were so good at what they do when Jobs lost his patience and butted in in true Jobsian fashion.
"Listen! Take a look around this table. Everyone here has a massive failure in their past. Big, epic failures. In the US, we think thats a good thing. In the UK you think it's a bad thing. Thats it."
Or soemthing like that, you get the idea.
Anyway, i've always liked that outlook regardless of wether it's true or not.