Forums

People who set Speed Limits deserve this Punishment

Sort:
llamonade
llamonade
Closed: Inactive
May 8, 2019
0
#41
EndgameStudier wrote:

I know what percentile means, I was asking how they determined that 85% was the best benchmark to split the difference.

Sorry, I'll delete that then. I misread your question.

EndgameEnthusiast2357
EndgameEnthusiast2357
May 8, 2019
0
#42
llamonade wrote:

No idea.

But again, it seems reasonable to me. If you posted no speed limit, then people's speed may be very different, so I think that would increase crashes

But letting people drive however they want would probably give you a good estimate for what the speed limit should be, because on average people probably pick a good speed.

I guess average is a little too slow though. It seems 85th percentile is good. Why 85? Maybe they wanted to test that because that's roughly 1 standard deviation on a bell curve. Other than that I don't know.

I see, anything within a Z-Score of 1 was a good estimate.

llamonade
llamonade
Closed: Inactive
May 8, 2019
0
#43

It's been a while since statistics class, I guess z score was like that (just another word for standard deviation?) I don't really remember tongue.png

EndgameEnthusiast2357
EndgameEnthusiast2357
May 8, 2019
0
#44
llamonade wrote:

It's been a while since statistics class, I guess z score was like that (just another word for standard deviation?) I don't really remember 

Z score is the number of standard deviations from the mean, a data point is. For example, if the average of something is 50, and the standard deviation (average distance from the mean) is say, 13. Then 76 would have a Z-score of +2 because it is two times the average distance from the mean (2x13). Supposedly 65-67% of data falls into a Z-score of 1, 80 something a Z-score of 2, and 99.9% within 3 Z scores. I learned all this bell curve stats stuff back in 11th grade, confusing at first, but then it made sense.