Maybe not limited in the amount of data, but very limited geographically.

You misspelled Aldus Pagemaker. For spreadsheets, it was Excel (before Excel was available on the PC). Jazz was heavily hyped, but flopped.

Submitter is a tyro. It was, of course, Visicalc which added to the Apple ]['s success. And, before Lotus 1-2-3 on the IBM PC, there was Multiplan.

Your point is that if all your friends jump off a bridge into shallow water, you should too? Or is it that breaking the rules is fine as long as you don't get caught?

All aid to Israel should have been stopped the first time we caught them red-handed.

"Simpler." Accuracy and repeatability weren't conditions.

They learned it from the government, where when they raise taxes by $120 billion instead of the $180 billion they really wanted, call it a tax cut of $60 billion.

"It is Boeing's job to make the airlines whole (if it's determined that they are indeed unacceptable), and Honneywell's job to make Boeing whole."

Without knowing what the contracts say, that statement is pure conjecture. Both contracts may have acceptance clauses ("we think this meets spec, but you're ultimately responsible for testing it in your usage environment to be sure and notifying us in a timely manner") or simple warranty terms ("warranty on defects in material and workmanship only good for 90 days").

Huh? Points drop off at a constant rate without respect to age. So, how are points/tickets on record biased against younger drivers? Is a young driver who has had 1 ticket in the past year somehow counted differently than an older driver who has had 1 ticket in the past year?

Furthermore, you want to limit the tickets considered to recent ones - older ones are more likely to have been received while driving a car which is no longer being driven by the same driver. The average time of ownership for a car in the US is around 5 years. It seems in most states points expire after about 3 years. So, it seems there would be a reasonable and logical correlation between the car one is currently driving and the car which was being driven when the points were accumulated.

You are, of course, begging the question by stating that "younger are more likely to still have the points on their record," then claiming that as a bias for a result which shows the cars young drivers own get more tickets. That's not a bias, it's an expected result - cars driven by people who get ticketed more often are more likely to have been ticketed.

Insurance.com is a aggregator, they pull quotes from multiple insurance companies. While having a broader base for data would make things more accurate, that's different than trying to claim that there's a bias. In what direction does that bias work, and why? It's not like the data is coming from a company which caters to high risk drivers, or provides insurance of "last resort."

"Yeah right, with your system you need to wait at least six months to calibrate a thermometer."

Did you even read the question which I answered? Requiring a thermometer which needs calibration is hardly "simple."

I suspect that when an insurance company does a quote, they consult databases of driver's ticket/points records. The "self reporting" would only apply to matching drivers with vehicles, so they can make the indirect link between tickets and vehicles.

It's unlikely there's any significant "self reporting bias," as you seem to be implying, which would be caused by drivers being deceptive about the vehicles they're actively seeking insurance quotes on.

Whoosh. As opposed to 100 being boiling water? That works for science, but not humans.

Try to convey temperature to someone in a temperate climate, unfamiliar with any temperature scale. If you tell them 100 is the hottest summer day and 0 is the coldest winter one, they'll have a pretty good idea that halfway between, 50, is roughly the average temperature during spring and fall. They'll have an intuitive understanding of the scale, and be able to make reasonable estimates based solely on their experience.

Now tell them that 0 is where water freezes (good so far), and 100 is where water boils. What's the temperature half way in between in their experience? What would they guess the temperature to be on a hot day?

You're arguing that well defined scale points are needed, and they are for many things. But picking 0/100 vs 32/212 is just a choice - there's nothing inherently better or more accurate about the Celsius scale for scientific purposes. There is something better about the Fahrenheit scale for human purposes - it can work reasonably well without any instruments.

Both C and F temperature scales were developed by northern Europeans. Both are arbitrary scales, but the 0-100 range of F better matches the range of climate experienced by humans. In some ways F makes more practical sense - it was made so the range between freezing water (32) and body temperature (96) would be a power of 2, easily and accurately divisible on instruments.

Temperature is unlike SI distance and volume units, where just the change to decimal ranges made things easier (i.e. 1000 m/1 km, as opposed to 5280 feet/mile, etc), There's nothing inherently better or easier with Celsius. In fact, one might legitimately claim that the Kelvin (or Rankine) scale makes more scientific sense.

"Water freezes at zero and boils at one hundred. What could be simpler?"

0 is a cold winter day, and 100 is a hot summer day.