Bill Price
Active Member
That’s your opinion, and in MY opinion, you could not be more wrong!!
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
OFFICIAL BUTTON WATCH
- Thread starter Drrobmiller
- Start date
Maybe for some. For others, the statistical model is dinging them for things that aren't unsafe. But I am definitely interested in the subset of unsafe drivers who thought they were safe who actually aren't. I've seen some people admit that the score is making them aware of their follow distance and lessened reaction times when they go fast, etc.
It's your opinion that the facts are my opinion, but you are wrong.
GeoffEV
Member
When approaching a Stop Sign or red traffic light on AP, how come AP races up to well past the point a normal driver would and then decelerates aggressively until in Hold? Perhaps I was wrong, but up until this Safety Score period, I had thought the computer was attempting to get max regenerated power so as to be most efficient. So, AP taught me how to drive my Tesla completely different from how I drive an ICE vehicle (anticipate SS & Lights by coasting to slow down and then light&slow pumps of brakes so as to not heat them up too much).
[Aside: I got over 120,000 miles on a Grand Caravan's brakes - the transmission crapped out before the brakes&rotors! "Kars4Kids" towed it out of here with the original brakes&rotors in fine shape].
[Aside: I got over 120,000 miles on a Grand Caravan's brakes - the transmission crapped out before the brakes&rotors! "Kars4Kids" towed it out of here with the original brakes&rotors in fine shape].
Cheburashka
Active Member
Damn, so the earliest anyone with the button may get FSD is Oct 8, and that is only for those with scores of 100... I already blew my chances.
Going on a road trip next week too, I guess I won't see FSD until 2022 at the earliest.
Going on a road trip next week too, I guess I won't see FSD until 2022 at the earliest.
I always hate to be THAT guy to pile on here, but I do have to say I disagree with this in part for sure and certainly in THIS context of the Safety Score. Beta. In support of your thesis, I DO think that I am a vERY good driver and my 38 year driving (and racing) record proves that ourright.
As I drive (and teach others to drive) I am always aware and ANTICIPATE possible actions on the road, and by other drivers. This has allowed my often and always to AVOID accidents that others would not have been able to. Abel to avoid situations that other POOR drivers put the other drivers around them into jeopardy. There is a need to drive, understand all the conditions around you and anticipate possible negative scenarios which allows for more rapid response and often that has made all the difference.
But, there are certainly situations that CANNOT be fully anticipated. A deer jumping out from the side of the road. A tree falling in front of the car that requires either rapid de-celeration or avoidance. All the snow from the roof of a car in front of you (while driving at a safe speed and MANY car lengths of distance behind) flying up and into the air and landing all right on your windscreen. These, and many others like them are situations I have encountered and avoided. Safety Score would have dinged me at least 50% for the successful avoidance of them and given no credit for doing so.
My scenario from over the weekend, was a driver who pulled out unexpectedly in between me and the car I was following. we were all traveling slower than the posted speed and with at least 3 - 4 car lengths In between us. this driver in a side parking lot (I’ll call him a..shat) decided that with enough speed they could get out of the parking lot and in between my car and the car in front of me, and accelerate FAST enough so I didn’t hit him, but then DE-Celerate fast enough so HE didn’t hit the car in front of him. This move was in MY anticipating matrix, as is the case with my driving behaviors. It was an as..hat move by them, but one that I’m sure we all have seen. As I commented before, all I had to do was an instantaneous lift off and let the Tesla decelerate with just regen braking. No accident, no hard braking, no sudden steering input require (although I was ready for that) not enough close. Yet, the tesla threw the impending collision warning. I get that - closing distance between my car and the car in front of me was pretty quickly HALVED. The computer is interpreting this as the car in front actually DE-Celerating RAPIDLY and the Tesla should take action. The tesla was certainly not doing any vision of the FULL environment (as was I) and anticipating the UN-Expected and preparing actually to take sudden action to de-celerate or avoid. But I was. Frankly, this is actually one of the overall concerns that I have with FSD going forward is that for a truly GREAT driver, it’s not only all about vehilcle control but aLao about overall situational awareness and anticipation to either drive optionally but also to avoid the unexpected as much as possible. I don’t expect this from MOST drivers and sadly I don’t expect it from FSD anytime soon.
But none of this do I interpret as becoming more aware that I am not the level and quality of driver that I have learned to be.
So, thesis, good.. conclusion not so much. I will say as it goes, you’ll probably get into the actual FSD beta before I will as I don’t see a path to get back to anywhere near 100% in the time allotted. I’m going to try and appeal to Tesla that they should probably do more than just a top line % comparisson and throw out maybe ONE outlier. It would make more sense than totally disenfranchising someone like myself in this situation (bought FSD early, very early buyer M3), clearly no accidents, clearly no poor driving behavior) but based on the math and their now purported process for allowing new participants into the FSD beta, I’m probably not going to see if for months or possibly by the end of the year. That’s just bunk.
Last edited:
It depends on how you read the tweets. There is a possibility they'll still start adding people October 1st with 10.1 or a 10.1.1 (not a thing but I wouldn't totally rule it out).
My justification for this is that he stated they'd roll out together, then in his clarification only specified the 10.2 date, not the expansion of the beta date.
I'd guess its >70% possible you're right though which would make me sad...
Agreed, it’s not clear that fsd beta will start expanding with 10.2. Or if it could expand this Friday with 10.1. But when in doubt just add two weeks!
Elon didn't say you had to have a score of 100. We'll be graded on a curve. He said priority would be given based on score. He tweeted that on the first day you'd probably get 100, second day 99, third day 98 and so on. So unless there are 1000 people with a score of 100, some with lower scores will get it on the first day.
Just go drive on autopilot for 1000 miles, and your current score will only contribute like 1 percent to the total.
Last edited:
We’ll see.. I will say, I doubt that they will add 1000 a day for more than about two weeks.. then they’ll pause for maybe 2-4 weeks, and then some new amount will be added. I think they might be willing/able to increase the current tester numbers by ONE order of magnitude, but mostly likely not TWO orders. We’ll See. Maybe I’ll setup a pole.
GargantulaKon
FSD Beta Chill Member
So it looks like from what I read on the safety score documentation from Tesla that we don't have to drive lots of miles to get back up to 100, we just have to drive with the goal of getting a safety score of 100 for each day which will then average out with the daily scores that we have been getting for the past 30 days. To get back to 100, we just have to drive a small amount each day with no incidents.
Am I correct?
We still don't know the minimum mileage needed though in order to be considered for the FSD beta though.
Am I correct?
We still don't know the minimum mileage needed though in order to be considered for the FSD beta though.
So you think 14,000 people will have a score of 100? I don't think so. It's pretty easy to get dinged.
Cheburashka
Active Member
He said 10.2 will be a week from Friday.
And he said it as a followup to "wow, so much interest in people getting the feature they paid 10k for", so because of this I'm 99% sure that none of the button holders will get access to the beta before then.
Pretty sure each day is graded separately, then the days are averaged together with a weighting on the mileage for each day.
Ex. If day 1 you got a score of 56 but only drove 4 miles, that would be a lot easier to bring back up than if you got a 56 but drove 400 miles.
It's weighted daily by miles driven. So if you have a bad score you would have to have relatively high mileage with good scores.
GargantulaKon
FSD Beta Chill Member
Ah, okay I missed the part about mileage. So longer drives with 100 are going to bring up the aggregated safety score more than a 0.1 miles trip with a 100 score.
I saw this part "Your daily Safety Score is not impacted by the number of miles or hours you drive." And thought that the daily score was the only value that was used.
After that, it says "We combine your daily Safety Scores (up to 30 days) into a mileage-weighted average to calculate the aggregated Safety Score, which is displayed on the main ‘Safety Score’ screen of the Tesla app."
Last edited:
GeoffEV
Member
Be content! Those of us who got the button and rushed out to see what's what on the Safety Test shoulduh stepped and waited for more info and guidelines. Right outta da barn, I got an 88 bc I had no idea that slowing down by mimicking what AP does would reward me with >0.4g deceleration demerits. All my trips since then are at 100%, but I'm stuck at a 4 day average of 98 - and, that's only if someone doesn't swerve into my lane tomorrow or on day 6 or 7 a deer doesn't decide to eat acorns on the other side of the next blind curve!
Best of luck to all!
Before that he said the button would come out with 10.1, and then your driving would be evaluated for 7 days. So, I don't know why we have to wait for 10.2. Just give us 10.1. I suspect it's because 10.1 is really 10.0.2(no combined stack). So we will be getting 10.1.
Last edited:
Similar threads
