Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Safety Score

This site may earn commission on affiliate links.
Scroll wheel reset is your friend on tough drives
Hopefully Tesla will monitor total odometer change since The Button was pushed on each vehicle, and compare to Safety Score logged miles, prior to handing out FSD Beta enablements.

Vehicles with major discrepancies can have their scores reset and allow for another 7 days of accumulation of scored miles, with no discrepancies, before reconsideration for admission.
 
Hopefully Tesla will monitor total odometer change since The Button was pushed on each vehicle, and compare to Safety Score logged miles, prior to handing out FSD Beta enablements.

Vehicles with major discrepancies can have their scores reset and allow for another 7 days of accumulation of scored miles, with no discrepancies, before reconsideration for admission.

for all the tricks for gaming the system, tesla can easily detect them IF they choose to look for them. I suspect though that they won't.

The only thing that I think they will do (that Elon hasn't mentioned) is weight the final score by miles driven. It seems too obvious to allow a 0.1 mile 100 score through.
 
for all the tricks for gaming the system,
I have no problem with people gaming the system with drives specifically meant to improve their scores. It’s good practice and still requires self-control, alertness, anticipation, and enthusiasm. As well as an understanding of what is being measured.

The reset seems like the only mechanism which is just against the spirit of the game. I suppose you could argue that someone who is able to rack up sufficient miles with a good score shouldn’t be penalized for a few bad drives which they chose to hide (it lowers their total mileage of course). So prioritizing by total mileage would help to offset the benefits of the scroll wheel reset. So hopefully they’ll do that - but I suspect they’ll just set a minimum mileage to ensure a good sample of driving behavior (maybe 50-100 miles), and maybe a minimum number of scored days.

If they are serious about using it as a screening system for a safe FSD Beta evaluation, they should definitely remove people who have a tendency to hide their behaviors - they will be correlated (not saying the same group, just correlated) with steering nag defeats, putting up blocking pictures to spoof the cabin camera, etc.

Obviously everyone is “hiding their behaviors” by driving differently when under evaluation - but I’m assuming there will be ongoing strict monitoring during the FSD Beta development period. I’m hopeful that you will have to consent to monitoring on every drive, to enable the feature, once the FSD Beta is rolled out (so you can drive normally without fear of being kicked out, if you choose to not use the FSD Beta on a particular drive).
 
Last edited:
You need to drive a bunch of non-AP miles WHILE FOLLOWING someone at 1-3s to rack up time spent following safely.
Very true.

Drove today in medium-ish freeway traffic, running about 75 MPH, low regen, follow set to 5 or 6. Turn on AP while on the entrance ramp. Find someone to follow. When you do, disable AP and only use cruise control, speed set to slightly higher than traffic to maintain the magical 2 to 3 second (not more, not less) distance. If someone starts to merge in front of you, throw AP back on. Repeat for the drive. Did this for a 65 miles drive today and it dropped my "following" dings significantly.

Using AP most of the time. But if coming up on curves, stopped traffic, or a light I disable it. I need those manual turn/stopping brownie points.

If you get the FCW (which I thought also disabled AP, but maybe not), immediately reboot the car. It appears to effectively lose the current drive. Better to lose some miles than take a FCW hit. I use AP a lot and an FCW just happens regularly, especially if the intersection is slightly out of square.

Now I'm not sure if we're done on Saturday or if we have to play the game another week. I'm just barely at 99%. I'd have to drive 550 miles tomorrow to get to 99.51% (round to 100%). Not going to happen UNLESS this goes on another week.

Also if another week, there's hope if you have a lower score. Answer is miles per day with good scores. Looks like they average the daily average (very illogical) so one bad day can be overcome. Serious rush hour traffic is impossible for me to score well in, but now that we better know the rules, it's otherwise possible.
 
  • Like
Reactions: FalconFour
SumProduct (Weight Average) in Excel or Google Sheets would seem to work.
Here is an example:
4UlrSGx.jpg
Thanks for this. Now I can see that I am 1 day's normal driving from a 100 clawback.
 
Hopefully Tesla will monitor total odometer change since The Button was pushed on each vehicle, and compare to Safety Score logged miles, prior to handing out FSD Beta enablements.

Vehicles with major discrepancies can have their scores reset and allow for another 7 days of accumulation of scored miles, with no discrepancies, before reconsideration for admission.
I don’t think that they will. There are still valid reasons to do a reset. On the previous software rev, my screen totally locked up and I had to do the reset to get home.
 
  • Like
Reactions: AlanSubie4Life
My wife and I together have been running a consistent 99 and luckily we've been rescued by a long weekend family gathering that is better reached by rail.

I've decided that this whole exercise is one big psychological experiment. It's somewhat like this research study that my dog is participating in:

 
  • Love
Reactions: Microterf
True. If they didn't you're have trouble navigating in tunnels and other locations where the GPS signal is weak.

But even if the IMU data isn't used in the Safety Score, the steering wheel angle and vehicle speed could be used in tandem to measure "safe" turning.
yes.

when done right, you 'fuse' all of that. you get gps and imu updates as a stream, you get wheel 'ticks' and you get wheel angle (not 100% sure about the last one, some brands have that).

there may even be more than one gps unit in the car (I dont know tesla internals).
 
I feel sorry for you guys who have to commute in a busy urban environment. I consider myself a good driver and was rocking a 99 after 4 days but I had to drive into the city during rush hour today (30 miles took 70 minutes) and despite being on autopilot 80-90% of the time and driving as cautiously as possible, I got a score of 81, which dropped my overall to 93. And it was one of the most stressful drives I've had in a Tesla because I was trying so hard keep my score up. Two cars ahead of me both ran a red light, autopilot tried to usher me through but I braked late and got hit with Hard Braking. A woman was walking close to the road, all lanes bumper to bumper so I couldn't give her any room, and got dinged for FCW. And the few times I wasn't on Autopilot, several cars cut in front of me (because I was trying to maintain the 3 second rule), and I think that triggered Unsafe Following. Getting a very high score (>98) is likely impossible in heavy rush hour traffic. Jeez this is too much work. I'm going back to just driving safely and not caring about the score or when (or if) I ever get FSD Beta. Best of luck to the rest of you.

View attachment 716447
Opt out on the autopilot screen and let us know if the safety score keeps tracking or not. Thank you for your sacrifice!
 
I got a 7% unsafe-following driving on side streets & going through a drive-through. The whole trip is 10 miles. I was not driving 50mph and I wasn’t following anyone (early morning). I did get a bunch of sensor songs while crawling through the drive-through.

Since then I’ve stayed way back from the cars in the drive-through and haven’t had a problem.

B918462C-49BA-4A19-BA46-4785EEA37F76.png Chick fill a drive thru caused the 21.1%
 
Why would you have 99.53? You don’t have a single 100 day from your screenshots. Looks like mostly 99s and below
I'm talking about just today's score, not overall. If you do the math on the metrics I shared above, the calculated score for the day is slightly over 99.53%, so rounded correctly that should be 100% just for today.

My total score is 98% and that makes sense, though it's annoying I got an averaged out 94% on Tuesday for getting dinged on a FWC due to the car freaking out while a traffic person signaled me through (which I posted about earlier in this thread with a video).

Don't get me wrong, 99% is still good for today's drives, especially given the total mileage of 121, but I'm just questioning whether Tesla is rounding up or down. Looks like they're rounding down despite the number being closer to the higher integer.
 
I’m not so sure Tesla is rounding up daily scores. Per my calculations, the drives I took today should have yielded a score of 99.5357%, but my app is showing 99% (not 100%) for today.

@AlanSubie4Life any ideas?

View attachment 716599
That set of values ends up with a value between 98.5 and 99.5 (did not calculate it). Anyway it rounds to 99. And I think it will be weighted as 99*121miles in the overall score, regardless of whether it was actually 99.3 or 98.7.
the calculated score for the day is slightly over 99.53%,
It’s 99.34.
It looks like you just accounted for the first factor of hard braking.
 
That set of values ends up with a value between 98.5 and 99.5 (did not calculate it). Anyway it rounds to 99. And I think it will be weighted as 99*121miles in the overall score, regardless of whether it was actually 99.3 or 98.7.

It’s 99.34.
It looks like you just accounted for the first factor of hard braking.
I found the bug in my spreadsheet, turns out I was taking into account all factors EXCEPT hard braking (the cells column for that param had an extra letter, typo on my part). All checks out now, thanks.
 
  • Like
Reactions: AlanSubie4Life
Ironically, in spite of all the complaints here, I think the thing that is the least safe about the Safety Score system is that it incentivizes AP use in busy situations, and the characteristically unnatural and terrible driving it produces. A true 100 score for me in my 40 miles of driving today though. 🤷‍♂️

Probably one of the worst incidents I've ever had on AP. There's a reason I typically only use AP when there's light traffic! Yet with the Safety Score I must do the exact opposite. Sad.

As the description says (this was I-15S at the split to the 163):

If driving manually (which I would normally do in these circumstances; I was using AP to avoid unsafe following Safety Score dings) I would have signaled right and gotten to the right lane sooner, but the traffic in the fast lane was moving very slowly (it's always a disaster at this interchange, with people cutting in late, but normally traffic in the left lane does move the fastest, which is why I had picked it) and the lane to the right had a steady stream of cars - and Autopilot would have had a difficult time of it (but this is what I should have done, and just tried to have it find a spot - ironically I thought I would end up following too closely and wanted to wait). Going manual mode would have meant a Safety Score ding! The entire video is on Autopilot, with Lane Change Confirmation required, following distance 7. The only thing I did was reduce set speed (by about 10mph) immediately, as soon I saw the truck start to signal (I have no idea why he decided to get in so late - those familiar with this interchange know that you have several miles to get in the correct lane). And then Autopilot proceeded to ignore the truck and accelerate into it, after he was fully in my lane. I held my foot over the brake but did not use it (eventually AP applied the brakes/eased off accelerator); would have meant an unsafe following ding! :rolleyes:

The wife was not happy - and neither was I. It would have been bad, but acceptable, if it had continued to open up space after the truck changed lanes (the correct behavior would have been for AP to see the truck signaling, and preemptively allow 2 seconds following when the truck entered the lane, which it's not capable of doing) . In retrospect I could have kept dialing back the set speed to 40-50mph, but it surprised me (I don't know why this was a surprise) that AP did not see the truck at all for a good long time. I expected it to continue to fall back, and its behavior was exacerbated by the truck also applying the brakes. Radar really doing a great job.

Perfect 100 safety score, 0.0% for everything. About 0.3 seconds following distance.


Preceding this, I had been in the right lane, but there was a lot of slow traffic which I had shifted left to avoid, and there were also disabled vehicles parked on the right shoulder which I wanted to avoid. Then the lead car started driving slowly, and then all the traffic started passing us on the right. In retrospect, should have just dealt with the traffic in the slow lane.

I blame the Prius.

Tesla should have skipped the part of the Safety Score where AP masks errors.
 
Last edited: