I believe you can see that on the main screen under "Scoring Factors" as it does seem to show a percentage Hard Braking as you described using data across all days. But yes the overall score seems to be a plain mileage weighted average of daily scores not directly calculated from the aggregate factors. You can verify by copying the scoring factors into the score simulator, and it probably will be different from your overall score.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Safety Score
- Thread starter ScottNY
- Start date
Knightshade
Well-Known Member
Sure they have. Right in the owners manual.
FalconFour
Member
More to the point: now I can goose the score by doing more "safe following" instead of relying solely on Autopilot to be the only thing that could satisfy the mystery magic. My score was really bad before (but unimpactful) because I probably only accumulated a total of 10 seconds within the 0-3 second window across all my milesAnything before today has been locked in. but moving forward, you can prevent unsafe following from ever being an issue again now you know what to do
It was all made of merges and onramps/offramps before enabling AP or decelerating below 50!I guess I'll take another shot at rephrasing the idea in a way I understand:
It's the amount of time spent in a 0-3 second following period behind a car, while time spent below (closer than) 1 second is considered "unsafe following". Time spent more than 3 seconds behind a following car is not considered "following" for this score.
It is not. Trips must be at least 0.1 miles long. That's beyond the range of Smart Summon.Has anyone checked if a Smart Summon is counted as a trip?
FalconFour
Member
The Cabin Camera update already starts producing "boop-boop" stage-2 alerts when it notices you using your phone (ask me how I know... not gonna lie, but it's very rare). You oughtta know by now how much information Tesla hides - through lazy programming or design choice - from the UI. There's so very, very, very much more information under-the-hood than you can see on the screen (have you ever seen Service Mode? Oh my god, lol - it calculates the solar heat load hitting the car based on the time of day for purposes of automatic HVAC!). You can bet your sweet butt that they're counting "phone use" classifications.
Check Out How Tesla's In-Cabin Camera Analyses Your Face
Videos of what a Tesla in-cabin camera sees and what it analyses were posted to Twitter, check out what it looks like.
insideevs.com
It's capable of so much more than we can see - and thanks to Tesla hackers like Green (above), we can see some of what's cooking.
I've basically stopped using Autopilot this past week unless the highway is straight with nobody close to follow. I normally have it active for 99% of miles even on city streets, but after getting the max hard braking score on my first trip, I've realized that any potential scoring events are excluded when on Autopilot. This means in the 1% where I drove without Autopilot for braking on a residential street on a hill with no lane lines, the denominator was basically only that braking scenario. So by not using Autopilot, I have more "normal" braking patterns throughout the day to maximize the denominator, so if there was actually anything in the numerator, it'll be easier to fix up.
HighZ
Member
HaHa - here we go. In the context of getting the FSD Beta, I do not think that is what he is referring to. They have enabled city AP and if they didn't feel it was appropriate, we wouldn't have that. He means something, but I have been on AP on city streets and have had events that would have scored against me, but they didn't count. I just had an FCW event on a 15-mile drive while on AP, and it wasn't counted. So I am SURE that that is not what is considered inappropriate use from that phrase. YOU can consider it inappropriate, but that is not Tesla's definition.
While you can argue (and I know you will) that no one should be using AP other than on limited-access highways and that Tesla leaving that in the manual (yes, I know that it what it says) is all-powerful, I will maintain that Tesla is not going to penalize our safety score by using AP on city streets.
EVNow
Well-Known Member
Interesting - I had not seen this. (I've stopped using twitter for sanity)The Cabin Camera update already starts producing "boop-boop" stage-2 alerts when it notices you using your phone (ask me how I know... not gonna lie, but it's very rare). You oughtta know by now how much information Tesla hides - through lazy programming or design choice - from the UI. There's so very, very, very much more information under-the-hood than you can see on the screen (have you ever seen Service Mode? Oh my god, lol - it calculates the solar heat load hitting the car based on the time of day for purposes of automatic HVAC!). You can bet your sweet butt that they're counting "phone use" classifications.
View attachment 716362
Check Out How Tesla's In-Cabin Camera Analyses Your Face
Videos of what a Tesla in-cabin camera sees and what it analyses were posted to Twitter, check out what it looks like.
It's capable of so much more than we can see - and thanks to Tesla hackers like Green (above), we can see some of what's cooking.
But none of this means anything when it comes to who will get the beta fsd - and thats the point of the discussion.
Yup unfortunate outcome of averaging daily scores. If you find yourself with a low score for the day, it's much easier to fix it up on the same day by increasing say time spent gently braking with maybe just a few extra dedicated miles to drop it down to 0.2%. Otherwise the only way to raise the score on future days is to average in days with higher-than-average scores -- likely 100 -- with a lot of miles.
Assuming you had 200 miles in total with 97 score before your 600 mile day with 93 score, this would result in an average 94 score. You would need to drive 9000 miles with 100 score to raise the overall score to 99.5 to round up to 100. But if you're looking to raise the average back to 97, that would be about 600 miles of 100 score.
Last edited:
CharleyBC
Active Member
FalconFour
Member
You could with 2021.24.4 - which I held onto for a solid month of daily "no, piss off with your updates" dismissals. I swore I wouldn't take another update until we either get The Button (which we did, thus I updated), or we get USB media album art back. If they ain't gonna fix anything, why bother updating?
There was a glitch in the "dynotest" mode that seemed to accidentally trigger/unlock Service Mode as well when it was activated. So I used it a few times to poke through and document everything it had to offer.
On a related note, I REALLY REALLY REALLY wish that "Disable battery heating by drive unit" option was available in the standard settings. It's absurd to be throwing away 10%+ of my charging energy on unnecessary battery heating at slow CHAdeMO stations. OTOH, real Superchargers heat the battery so quick it only spends a margin-of-error of time heating the battery, and quickly switches to cooling. So it's an option I'd leave on all the time to save energy & charge 10% faster or more.
Word on the street is, if you're within some small proximity to a Tesla Service Center (e.g. its parking lot), the code "service" will also get you there until you exit. But no good for anyone that's not in area of an SC
As implied, I lost access to this when I took The Button update... a fair trade, I say. Now the "dynotest" thing just, I guess, does what it's intended to do. It was definitely glitched in 2021.24.4
My score is so pathetic that my only hope of being offered to join the FSD beta test group is if this whole thing is an exercise in smoke and mirrors. Like just one of Elon's many seemingly practical jokes. Painful that I'll be turning in this second Model S in less than 2 months having paid for FSD for both and not really enjoyed the fruits. Well, of course there will be those who argue that auto lane change is worth something. Ok, well then I did benefit something.
If nothing else this has taught me how terrible a driver I have been. I quite honestly did not know that there's an art or be it science to proper braking or that turning rapidly is an insurance risk factor. I'm consciously making efforts to improve my score day by day not so much to improve my chances of getting beta FSD tester group status but rather improving my driving skills.
If nothing else this has taught me how terrible a driver I have been. I quite honestly did not know that there's an art or be it science to proper braking or that turning rapidly is an insurance risk factor. I'm consciously making efforts to improve my score day by day not so much to improve my chances of getting beta FSD tester group status but rather improving my driving skills.
Last edited:
I completely agree, I've seen no dings on autopilot at and I have experimented quite a bit over the last 6 days after having a terrible first day score.
After some careful experimenting, I have noticed that if I follow and break even more carefully than the way autopilot does I'll get dinged for following too close and aggressive breaking when approaching stop lights, even though I'm being much more smooth and subtle that AP does. Also, I have noticed on a very short freeway test, I was in Autopilot following a car on 1 car length for 5-10 minutes, got no dings after a quick stop/exiting the car and checking the score. Then, I did the exact same thing, but then turned off Autopilot for a mile and kept the same or slightly longer distance than AP - and I got dinged for following too closely.
I then drove straight through the town, 10-20 stop lights and let Autopilot do everything, with crazy aggressive following and stops at times, and one crazy fast curve it took without slowing down the way it should have - but I DIDN'T have to intervene so I didn't. Perfect score! If I'd driven like that (based on first day of driving much better than AP) my score would have been all dings for all categories.
Kind of crazy game to be sure.
This has been my exact experience. The more I leave it on AP, the higher the score. In town I take if off just if I have to turn, and then put it back on immediately. Also, I add following distance in town if I predict I'll need to take over from AP so that I can slow gradually and so it doesn't jerk the car around the lanes (like when street lines are vague or missing) which could force me to take over while swerving hard (which AP does all the time, without penalty). But if I have to take over and swerve because of AP not doing something correctly I do get dinged.My first safety-rated drive was about 50 miles, nearly all on the interstate, no AP. I drove like a grandma with Prius. 100% One-pedal driving. I expected a 100 rating and was surprised to see my first score was 86. I got dinged for "hard braking" (with regen I didn't think that was possible) and "aggressive turning" (on and off ramps, I guess).
On the return trip, I used AP as soon as I got on the highway and kept it on the whole way home. Telsa rated itself at 100.So now I have an average score of 93.
I hope when people start getting FSD through the button system they'll post their winning scores and state.
jebinc
Well-Known Member
I agree. In and outs are problematic; especially if you break out because it's being unsafe in your opinion.
mltv
Member
I am starting to think that this is an exercise of not really measuring your safety score but rather FSD beta tester safety score.
- This measure your environment - higher scores are attributes to your environment may not be too busy. You are able to take it easy and not accelerate thru curves. This allow FSD an environment that it can start easy for the new testers.
- You as a training monkey - This trains you to not follow other cars too close. Not going thru curves too hard. etc.
- Gauge your daily activity or lack there off - If you are busy drop off your kids, running thru drive thru. You are probably will not score 100. Those unsafe following dings happens often when you are bumper to bumper but could not use your AP for the purpose. This weed out the busy bees and give the ones that has ample of time to actually doing the FSD testing.
EVNow
Well-Known Member
Similar threads
