Safety Score

[blecchus_rex]

After grinding the whole week to get to 98 from 96. Wife take the car out and score a 96 . Can't help it - venting. Currently 98 on 849 miles.

My wife is currently driving my car after having picked up a nail in hers.... and I made the mistake of requesting that she take it easy in mine.

 

[d21mike]

Haha, I know it is opt-in as I posted the screen where that happens and the TeslaFI changelog email a few post up!

Good idea about adding the firmware version but honestly I'm not sure there are any beta testers on TeslaFI and if 3rd party tracking sites are allowed (NDA?). I haven't noticed odd firmware version when I'm looking on TeslaFI OTA tables and such.

I was talking about this Saturday when they start adding 1000 per day starting with the 100% list. I would be surprised if the NEW FSD Beta users are required to sign a NDA and we would have to remove our cars from TeslaFI. Elon said he would start with 100 then 99, 98 etc. I wonder how true that turns out to be.

 

[SD Diver]

Here I had been at a perfect 100 for 5 days over 100 miles. Yesterday I drove to ucla from San Diego and scored 97 for the 220 mile trip. Most of it was on autopilot. Twice I was cut off and AP aggressively applied the brakes. I was dinged. I had the following distance set to the highest value of 7. Dinged for following too close. I even got an aggressive turn - can’t figure that one out. So, a week of driving like an octogenarian in chill mode was ruined by being on the 405. So I’m now at 97 total with no real way to make it up. I’ll happily let Tesla refund my FSD cost and call it a day. I’m dreading driving now…

 

[KArnold]

NEW FSD Beta users are required to sign a NDA and we would have to remove our cars from TeslaFI

Not sure I understand this. Originally I was thinking Elon didn't want commoners to know when an FSD releases came out. But now he tweets about it. What could be his motivation now? Maybe that ship has sailed.

 

[EVNow]

Good point - I think that they would have to cut off scores below a certain %, like 90 for now. We don't know how many beta testers Tesla wants to let in - Elon indicated that it would be 1000 per day, but for how many days? Maybe they will let in 10-20k? Is that too small of a group?

We don't know - the request button said its "early access". 20k early access is quite good (if total is 100k or so). But, if the idea is to expand to everyone who requested the button (and has a good score) - they will keep going until everyone has it. I think the latter might be the implication of all the build-up ... otherwise we'll have a lot of angry customers. Its possible the # of people who requested and access and have a score of 80+ is about 20k.

 

[S4WRXTTCS]

Here I had been at a perfect 100 for 5 days over 100 miles. Yesterday I drove to ucla from San Diego and scored 97 for the 220 mile trip. Most of it was on autopilot. Twice I was cut off and AP aggressively applied the brakes. I was dinged. I had the following distance set to the highest value of 7. Dinged for following too close. I even got an aggressive turn - can’t figure that one out. So, a week of driving like an octogenarian in chill mode was ruined by being on the 405. So I’m now at 97 total with no real way to make it up. I’ll happily let Tesla refund my FSD cost and call it a day. I’m dreading driving now…

The way the safety score is being tied to getting FSD beta early shouldn't even be allowed.

The simple fact that it's not safe to gamify driving, and the safety score does exactly that.

Does it really make any material difference if a driver has a safety score of 80 or a safety score of 100?

As an example I drove 9 miles to work today with the plan of achieving 100% for the day. I made sure to take corners slow, and not to hard brake. Everything was going well until a light up ahead turned yellow, and I slowed moderately to a stop right along side the car next to me. That was the only significant braking event, and that single event knocked the score for the drive down to 91.

Having people attempting to have a perfect score throws drivers off from what they should be doing. What they should be doing is driving, and not worrying about their safety scores.

The irony is the FSD beta test is really intended for City Streets, and yet the best way to get a good safety score is to avoid city streets whenever possible.

 

[S4WRXTTCS]

Here I had been at a perfect 100 for 5 days over 100 miles. Yesterday I drove to ucla from San Diego and scored 97 for the 220 mile trip. Most of it was on autopilot. Twice I was cut off and AP aggressively applied the brakes. I was dinged. I had the following distance set to the highest value of 7. Dinged for following too close. I even got an aggressive turn - can’t figure that one out. So, a week of driving like an octogenarian in chill mode was ruined by being on the 405. So I’m now at 97 total with no real way to make it up. I’ll happily let Tesla refund my FSD cost and call it a day. I’m dreading driving now…

My understanding is when you're on AP none of those things will count against you. The problem is that so little time is spent outside AP (and above 50mph) that anything that happens will likely score a lot against you. Like you likely took over when cut off, and as soon as you took over you were following too closely. Then when you put AP back on none of those miles were counted (for the following too closely) so it really exaggerated the bad score.

 

[d21mike]

Not sure I understand this. Originally I was thinking Elon didn't want commoners to know when an FSD releases came out. But now he tweets about it. What could be his motivation now? Maybe that ship has sailed.

You dropped my "I would be surprised" which changes my sentence. Maybe YouTubers and taking posting videos of them driving FSD Beta. Elon was just interview about NDA and said he did not think they need one since everybody is posting videos of it. So, I really do not think the new ones would have the restriction.

 

[d21mike]

My understanding is when you're on AP none of those things will count against you. The problem is that so little time is spent outside AP (and above 50mph) that anything that happens will likely score a lot against you. Like you likely took over when cut off, and as soon as you took over you were following too closely. Then when you put AP back on none of those miles were counted (for the following too closely) so it really exaggerated the bad score.

Many people have said this and it is in the documentation very clear. But my worst scores were on AP. So something like you are saying must have happened but being aware of what you just said I was extra careful and still got dinged so really not sure what I did. Anyway, I drive almost completely off AP and my current score is 99.

 

[S4WRXTTCS]

I agree - what's the point other than to create more of the 'game mentality' OR maybe they have some greater insight into what Tesla is going to do.

It certainly does penalize those 100's with few miles and rewards those 99-95 with high miles. It kind of makes sense but as you said, not much makes sense right now.

The problem with miles is high mileage = freeway = AP use = most of the miles aren't even counted except in the total miles (like for AP forced disengagement).

The people getting penalized are city drivers who spend most of their time driving in the city. The city is where a lot of the challenges come with hard braking, and aggressive turning

I'm going to experiment a bit to see the impact of using AP for a lot of miles on things like hard braking, and aggressive turning. If I interpret it correctly AP use won't impact those at all (either positively or negatively).

 

[Spuzzz]

…The irony is the FSD beta test is really intended for City Streets, and yet the best way to get a good safety score is to avoid city streets whenever possible.

I disagree with this. I’ve found it very easy to score a 100 on my city street only days. It when I take it on freeways that I get dinged. Usually for unsafe follow distance. On city streets, you generally stay under 50mph and at that speed unsafe follow distance isn’t recorded.

 

[Lowchrg]

I disagree with this. I’ve found it very easy to score a 100 on my city street only days. It when I take it on freeways that I get dinged. Usually for unsafe follow distance. On city streets, you generally stay under 50mph and at that speed unsafe follow distance isn’t recorded.

Yep this. ALL of my dings have been unsafe follow distance when merging with other lanes or faster traffic and can’t use the auto lane change yet. Any time I’m driving in the city I get straight 0’s all the way down.

 

[GalacticHero]

My score is down to 92. Two incidents were people suddenly getting in front of me when not on AP. Several were for hard brakes, accels, or swerves to recover from dangerous AP mistakes.

 

[AlanSubie4Life]

Today I have 13 miles, will add 13 more to go back home.

409 miles in total.

You're right, then.

In that case, as you said, you have enough total miles (and enough miles today already) that it just makes sense to drive as much as you can today, until you get to around 160 miles. It only makes sense to stop driving in very low distance (like, 1-2 miles) days (assuming there is a cap on the max value used for FCW/1000 miles). Because then that can be very low weight (1 mile with 1 FCW should result in 51 (assuming the score simulator reflects reality), rather than something lower, due to the cap) and you can easily make it better, with 50-100 miles the subsequent day. But that ship has sailed (and was never possible for you).

So you're right that given the situation, you do need to drive more miles today.

It's diminishing returns after about 160 miles though. At about 160 you'll be at 98.5 (99), and to make that to 99.5 you'd have to drive another 340 miles (500 total). So you may as well (given the way the rounding calculations work currently) just drive those miles the following day and get them to count and average to 100 (160 at 98.5, rounding to 99, plus 160 at 100, which averages to 99.5, which rounds up to 100). (That would be 320 miles total rather than the 500 on a single day, which is better.)

Anyway with all your prior 100 scores you just have to get to 160 miles today (you need to see the score go to 99) and then you should be ok.

For example, if you did just 26 total today, you'd be at 89 for the day, and to get that to average out you'd need to add 600 miles at 100 to get it to average out to 100 (99.5). That's worse (626 miles total). But if you did just 80 miles today (gives 97 score), you'd need to add 400 in the future at 100 to balance it to average out to 99.5. (480 total)

 

[mltv]

Yep this. ALL of my dings have been unsafe follow distance when merging with other lanes or faster traffic and can’t use the auto lane change yet. Any time I’m driving in the city I get straight 0’s all the way down.

Same here. Most dings are from merging lanes or cars from other lane changing lane in front of me. Not much I can do. Not sure this is correct, on freeway I use cruise control set to maximum follow 7. Let it keep distance from car in front. Does this work or I have to be off cruise control also.

 

[ajax85]

My wife is currently driving my car after having picked up a nail in hers.... and I made the mistake of requesting that she take it easy in mine.

I’m pretty sure I would get the kids and house in divorce after showing judge how wife dropped my 99 to 96 after taking car for one trip this afternoon…

 

[nvx1977]

I'm going to experiment a bit to see the impact of using AP for a lot of miles on things like hard braking, and aggressive turning. If I interpret it correctly AP use won't impact those at all (either positively or negatively).

Not going to stop you from experimenting and proving this for yourself (the world needs a LOT more people who apply the scientific method to their everyday lives and not just believe what they hear). But you are right: hard braking and aggressive turning are time-based measurements. mileage can't dilute those dings. Only more time spent doing good braking and good cornering can reduce the dings.

 

[sealbeach100d]

When I first started the "Safety Score", I thought I read that mileage driven would NOT be a factor.
Yet, as I've been following this thread (& others) mileage driven is talked about quite a bit (as in, if one doesn't drive xxx number of miles, they won't be considered for joining the FSD Beta group).
Did I miss something from Elon changing the mileage requirement?
Thanks.

 

[phantasms]

Anyone have a spreadsheet they can share with the weighting formulas already in it? Just got daily mile/score and total score. Got 2 mystery forward collision warnings and am crawling back up. Now at 97. I’d like to know how far I need to drive the next few days to get back to 99 if not 100. Past 4 days have been 100s.

 

[EVNow]

When I first started the "Safety Score", I thought I read that mileage driven would NOT be a factor.
Yet, as I've been following this thread (& others) mileage driven is talked about quite a bit (as in, if one doesn't drive xxx number of miles, they won't be considered for joining the FSD Beta group).
Did I miss something from Elon changing the mileage requirement?
Thanks.

Mileage matters for the score - because the total score is weighted average of daily scores, weghted by mileage.

Apart from that everything else is speculation / wishful thinking.