Poll: current safety score

[GWord]

Lots of angst in other thread about having something less than perfection.

Participate if you want… anonymous poll so you can feel free to put in your real numbers, good bad or ugly.

If you only have a couple of miles of driving, maybe consider not responding until you get that up a bit. We all tend to hit 100 on the first mile or two.

I checked the option to allow for vote changing so maybe we can update as our scores evolve?

 

[GWord]

I will say, just by way of commentary that:

1. I started at 100 after a 50 mile drive with mostly autopilot
2. dropped to around 92 after the 50 miles to return home and fighting Houston traffic, mostly with hand driving
3. clawed back to 97 with 50mi of granny driving, by hand in a suburban, surface street environ

 

[Terminator857]

Wife is scoring in low 80's. I just got my first 100. Woohoo.

 

[macmee]

9 miles driven 99 score so far

 

[AlanSubie4Life]

Looks like we need more resolution on the 95 to 100 range. (95, 96, 97, 98, 99, 100). Could do even finer resolution, but would require a link to a calculator and is a lot more complicated to figure out. Currently at 98.5 or so. Hard braking so painful & punitive.

 

[jebinc]

Pullin’ a 98 right now….

 

[TresLA]

99 so far, after 2 days of 10 finished drives totaling 45 miles. Only one incidence of hard braking and Im pretty sure it was pulling out of a parking space (ugh). In any case, I think I’m keeping the score up (all 100’s except for that one short drive) by staying on autopilot as much as possible (which except for the one factor keeps you safe from safety score dings), including on surface streets which is how I drove before anyway.

I am driving differently (than pre-button/safety score), by disengaging AP long enough before a turn to gradually (mostly just regen) slow the car before the corner and I granny-drive my way around the turn, then re-engage AP soon after the turn to let AP bring the car back up to speed. I know it’s gaming the system, but I’m definitely more attentive as to when I should disengage AP to slow the car more gently than usual in preparation for coming to a turn or lane split or lane change or etc. that I know AP will have trouble handling.

 

[aja2460]

97 after two days and about 100 miles... but I'm 75 years old...

 

[jebinc]

Up to 99 today.

 

[fiveten]

Sitting at 96 right now:
Day 1 - 110 miles at 93
Day 2 - 72 miles at 100
Day 3 - 33 miles so far at 100

I think I can hit 100 every day until Friday with a goal of getting my overall score to 99.

 

[gearchruncher]

So 75% of people have a score of 95% or above.
Elon says " If driving behavior is good for 7 days, beta access will be granted."
Are they really going to roll out the Beta to 75% of people that request it? Or will just a random sample get it, and will we see some people with a score of 96 get it while some with 100 do not?

 

[MrSlippery]

So 75% of people have a score of 95% or above.
Elon says " If driving behavior is good for 7 days, beta access will be granted."
Are they really going to roll out the Beta to 75% of people that request it? Or will just a random sample get it, and will we see some people with a score of 96 get it while some with 100 do not?

That is only 46 people as of now and we have no way to verify if those 46 people who voted are being truthful or if it will be maintained.

We also do not know the exact criteria, I doubt it is just "If driving behavior is good for 7 days, beta access will be granted."

 

[gearchruncher]

That is only 46 people as of now and we have no way to verify if those 46 people who voted are being truthful or if it will be maintained.

Lots of people over on Reddit saying they have scores of 100, and acting quite superior about it.
It's a pretty easy metric to maintain at or near 100% for 7 days, especially since EVERY SINGLE PERSON BEING MEASURED KNOWS THEY ARE BEING MEASURED AND HOW.
Won't surprise me at all to find there are lots of 100%-ers.

Will also not shock me one bit when someone has a 100% score and doesn't get it while someone with a 97% does. Do you really think the algorithm is so precise and correct that a 100% means you should get the beta but a 97% means you shouldn't?

 

[nvx1977]

yeah I would think of this like firmware releases. They are going to have other criteria, like geography, population density, weather, etc. We're probably obsessing too much over the score. Tesla's goal is to collect more data that will help them improve FSD functionality. Safeness of your driving is just a sliver of that objective.

 

[srs5694]

It's a pretty easy metric to maintain at or near 100% for 7 days, especially since EVERY SINGLE PERSON BEING MEASURED KNOWS THEY ARE BEING MEASURED AND HOW.

Given the number of people on this forum who state that they're driving at or below the speed limit, and accelerating gently, to keep their scores high, I'm highly skeptical of your all-caps statement. (Neither acceleration nor speed-limit compliance are among the criteria Tesla is using, unless Tesla is lying about that.)

 

[Eskimo YoYo]

I’ve been driving 10 over and sometimes 15 over on local roads and have a score of 100

 

[gearchruncher]

Given the number of people on this forum who state that they're driving at or below the speed limit, and accelerating gently, to keep their scores high, I'm highly skeptical of your all-caps statement.

Just being aware that the car is recording how you drive and using an algorithm to evaluate your "safety" over a 7 day period is more than enough to completely invalidate any data you get from that sample set.

"This study is to determine how fat you are. To calculate that, please give us permission to record what you eat in the next 7 days. If you pass, you get a cookie."

 

[EVNow]

Just being aware that the car is recording how you drive and using an algorithm to evaluate your "safety" over a 7 day period is more than enough to completely invalidate any data you get from that sample set.

"This study is to determine how fat you are. To calculate that, please give us permission to record what you eat in the next 7 days. If you pass, you get a cookie."

But, knowing you will be kicked out of Beta if you don't keep driving well is effective.

 

[EVNow]

Will also not shock me one bit when someone has a 100% score and doesn't get it while someone with a 97% does. Do you really think the algorithm is so precise and correct that a 100% means you should get the beta but a 97% means you shouldn't?

Won't shock me either. But not for the reason you stated - it would be because of the way they pick a subset of people who are "good drivers". Could be geographic distribution, # of miles driven, exact model of the car they drive etc.

 

[gearchruncher]

But, knowing you will be kicked out of Beta if you don't keep driving well is effective.

We have no evidence that will happen, just more random Elon tweets.
Look at the current FSD beta testers that post on YouTube. Are they driving safe? Many of them have hands off the wheel for a long time, and they get to keep it. Does anyone actually know anyone who was removed?

I am getting a kick out of all of this though. Those of you that paid for FSD are now trying to drive differently than you normally would, in order to satisfy a blind algorithm, in order to get access to software you PAID for years ago, that was supposed to drive the car for you, and you're worried that once you have it, if you drive badly in a car that is FULL SELF DRIVING that you'll loose access to the FULL SELF DRIVING you paid for, wherein it shouldn't be possible to drive badly since you will be using FULL SELF DRIVING all the time you drive and Tesla doesn't evaluate safety when you're using FULL SELF DRIVING.

NO REFUNDS.

It's crazy.

And the people now putting there "safety score" in their signatures. OMG. You have fallen for Elon's game so hard.