-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
100 score but only drove 36 miles. Think I should stop driving today/tomorrow and lock the score in? Or do I need more miles?
- Thread starter macmee
- Start date
I had 100% before Friday but I did not have 100 miles, which he did not mention in this tweet that you just included. He added the 100 mile minimum caveat early this morning when it started rolling out which is why I said the goalposts keep moving.
Yeah I feel a little sad by this too. It makes sense that there’d be a minimum but I wish the rules were put out up front. Oh well. I’m going to keep driving until I get 100 miles and then we will just have to hope they reconsider us!
Babaron
Member
I have been struggling with this scoring thing for a week or so and was also puzzled as to why it was so sensitive to various things like fast turns, too hard braking and close following.
And then it hit me. We are not the ones being scored here. It is our driving ENVIRONMENT THAT IS BEING EVALUATED! Tesla needs to make sure that the earliest releases are for those driving environments that are best suited to train the system with the lowest risk. All of the so-called events that ding the score are those events that might be too challenging for the FSD software. In other words, they want the next wave of trainers to be in driving environments that have the least risk of a mishap.
So those of us that live in big crowded and congested cities with lots of starts and stops and turns as well as required tailgating are going to get dinged no matter how safely we feel we are driving. That’s the bottom line. But presumably, our turn will come once the initial wave of testing has resulted in further improvements in the software.
And then it hit me. We are not the ones being scored here. It is our driving ENVIRONMENT THAT IS BEING EVALUATED! Tesla needs to make sure that the earliest releases are for those driving environments that are best suited to train the system with the lowest risk. All of the so-called events that ding the score are those events that might be too challenging for the FSD software. In other words, they want the next wave of trainers to be in driving environments that have the least risk of a mishap.
So those of us that live in big crowded and congested cities with lots of starts and stops and turns as well as required tailgating are going to get dinged no matter how safely we feel we are driving. That’s the bottom line. But presumably, our turn will come once the initial wave of testing has resulted in further improvements in the software.
I know all of this. I don’t understand why someone would think 2.5 miles a day of driving is what Tesla is looking for in a beta tester
boonedocks
MS LR Blk/Blk 19”
Don't necessarily agree with this. My first commute day I scored a 97 with .4 hard braking 2.3 aggressive turning and 60.0 unsafe following. After that disastrous day I realized how the scoring was made and have maintained a 99 with 2,182 miles driving the city of Atlanta daily commuter traffic.
Is it a PITA to keep that score in a major city stop N go traffic? YES!! Is it impossible? NO!
It is definitely on the driver and not the environment. We all have those individual moments of other drivers causing dings but they are more avoidable than not by just paying attention. Otherwise I wouldn't be able to maintain a 99% after 2,182 mile of daily Atlanta commuter traffic.
Two drivers same city driving:
Because he specifically said “everyone with 100/100 will get it”.
Actually in hindsight, I wonder if he meant 100%/100miles? Otherwise he could have just said “everyone with 100% will get it”.
Anyway, someone could have driven 98 miles and missed the cut off. If the rules were clear you don’t think that person would have been willing to go drive an extra 2 miles?
While I agree rules were not clear enough, from an beta perspective its probably not a good idea to have a sample size too small or too large. There were folks with 30-35 miles over the period of two weeks - that's a sample size simply not large enough. On the opposite side, I am also presuming people who have (miraculously) maintained 100 score while hypermiling are also a bad sample size owing to many more chances of things going wrong.
Weird, since they give beta auto wipers to every Tesla. And beta traffic aware cruise control. And beta auto high beams.
Except one of the things you want out of a beta is an actual breadth of use cases. It could be that the people that drive a lot less also drive in environments that are different, such as dense cities. The instant you start adding any discriminators to your sample set, you run the risk of ending up with a non-representative population.
Yet again, we're at "What Elon meant" after what happened isn't anything like what his words said.
Who says 100/100 (always said 100 out of 100) when what they meant was 100%/100 miles? The safety score isn't even in a percent, on purpose, as it is purposefully meant to push everyone into the 80-100 range to make them feel good. It's not a percent and it would be inappropriate to refer to it that way. The median score in a % based system should be 50%, not the ~97% it appears to be.
+1 and two other points to add to this:
- The people with fewer than 100 miles are probably those that have other cars or don’t need to drive every day, I would think they would want some people in that group
- interestingly enough, a long time ago I got an invite to EAP, but never joined. So obviously there is something about my car or location that they would benefit from testing
Anyway who knows, this is not a perfect science by any means
Quoting my earlier post and debunking my theory - I read the tweet again, he didn’t say everyone with 100/100 will get it, he said “everyone with a perfect score will get it”. Nothing about minimum mileage required.
boonedocks
MS LR Blk/Blk 19”
I believe this is what was referred to...Elon is quite "cryptic" sometimes leading many to interpret what he meant even when there was none:
I had 93 miles, so I didn't get it! I understand the reasoning for a minimum miles requirement, I just wish they would have disclosed it. Hopefully they do another round of new testers soon, so I can finally get it (I've already gone well over 100 miles now to make sure I meet that requirement).
I hope we both get it!!It worked!! I just got invited to the beta just now. I can’t believe it!!!
Does anyone here actually think people who drove .1 miles for 100 and garaged it for 2 weeks would be eligible for the beta? There are at least a few criticism thinkers who questioned this before last Friday.
Congrats. Im seeing several posts of people getting it tonight…I wonder if they are just slow on the deployment….like you qualified for it last week but they are just now getting around to pusjing it to you
boonedocks
MS LR Blk/Blk 19”
I still LOL at seeing people posting their daily 100's with 7 miles.
Current score 99.04% with 2,429 miles
Similar threads
- Poll
- Article
- Question
