Tesla will probably see that as a way or attempt to defraud the scoring system.....you probably cant opt in and drive badly, then reset to zero and start all over....people that di that may get put in a bucket that will not allow them to relieve the download
-
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
FalconFour
Member
Well, good for you. Opt out of the beta so that those of us that have been hoping to participate in the beta development for the past 2 years or so can inch a little closer to getting into the testing and development phase sooner than later
FalconFour
Member
OTOH it could be a legitimate way to slip out of the scoring while you rip around the track, and get back in. Probably more smiled-on than a wholesale "wow holy crap that is 100% definitely cheating" hack like rebooting after an FCW
AlanSubie4Life
Efficiency Obsessed Member
I’d recommend not using AP at all and using regen and manual driving to build your denominators.
I’ve decided using AP to mask unsafe following (the only utility it really has for the Safety Score) is too disruptive to driving flow for this beta and I don’t plan to use it anymore.
The only exception to that lack of utility is using it on surface streets to mask yellow light braking events. I’d recommend staying at or below the speed limit, watching light timing carefully, and if you’re in a “close call” situation, just go through the yellow rather than jam on the brakes to stop (probably safer anyway - assuming you are traveling below the speed limit). Full regen will not trigger braking events (except possibly on steep hills, but even there I am not sure it will since there is a limit to regen power and the pull of gravity will contribute to that power - so max regen will slow your wheel speed slower on a hill than on a flat - you REALLY can’t touch the brakes in those situations though) - and will build your denominator.
Don’t be afraid to roar up to traffic lights and completely let off the pedal and coast to a stop right at the line, without using the brakes.
No need to baby it.
I’ve gotten burned a couple times at yellow lights now, but there was really no safety reason I should have stopped for either of them. So I am recalibrating to just fly through them when the timing is too close (unless of course it is not safe to do so). If I can coast to a stop with full regen, I will stop for them.
Last edited:
EVNow
Well-Known Member
I think the concern is full regen + some braking (say for a traffic light at the bottom) will push you over 0.3g.
Basically every 20% gradient should add about 0.1g i.e. at the same speed and braking distance, you need that much more force to stop the vehicle. So, at 60% just to brake the car so that it won't start rolling (or start accelerating in general) you need to exert 0.3g of force. These numbers are very rough.
BTW, which g-force app did you use ?
I saw someone post this one either in this thread or another one and gave it a try. seems to work fine. There might be a better one; I didn't shop around.
I haven't used it much yet, but when I was testing it out, it seemed like on flat ground, regen doesn't peak until you're getting pretty slow ( <20mph ). Down the hill, I didn't see max regen generating higher g forces; it just stayed there longer. And I have some pretty steep hills here. All I wanted to know really is if I'm ever in danger relying on full regen. Seems like the answer is no. I shot down this hill at 40mph and let go of the accelerator abruptly. g forces were surprisingly low until I slowed some. The butt-accelerometer is deceiving.
I got it on the 23rd and it was just like any other app update. I know they do a rolling release of updates, but I didn't think it was his slow.
Glad to hear that everyone can get it - whether from the play store or via side loading.
EVNow
Well-Known Member
FalconFour
Member
Is FCW keeping your score low?
Yes -> good plan, miles driven on AP will dilute the FCWs-per-1000-miles score and increase it.
No, it's something else -> miles driven on AP will not add any good points to dilute the bad, and you'll just accumulate bad marks from driving manually only for tricky merges and immediately engaging AP which masks any good points.
To increase following score, drive manually whenever it's clear and you can follow someone while counting following seconds to stay in that 1-3 second window.
To increase braking score, just brake a lot with full regen. More seconds spent fully braking with regen = more good seconds to increase your score.
To increase turning score, general idea is to take road curves that have a yellow "curve ahead" speed limit sign at +5mph or so. Traffic turns I find best at 15-20mph. That's all about butt calibration... no good way to classify it, but if you take it too slowly, it doesn't count.
Enjoy the drive!
mackgriggs
Member
Is FCW keeping your score low?
Yes -> good plan, miles driven on AP will dilute the FCWs-per-1000-miles score and increase it.
No, it's something else -> miles driven on AP will not add any good points to dilute the bad, and you'll just accumulate bad marks from driving manually only for tricky merges and immediately engaging AP which masks any good points.
To increase following score, drive manually whenever it's clear and you can follow someone while counting following seconds to stay in that 1-3 second window.
To increase braking score, just brake a lot with full regen. More seconds spent fully braking with regen = more good seconds to increase your score.
To increase turning score, general idea is to take road curves that have a yellow "curve ahead" speed limit sign at +5mph or so. Traffic turns I find best at 15-20mph. That's all about butt calibration... no good way to classify it, but if you take it too slowly, it doesn't count.
Enjoy the drive!
If I'm reading the safety score page correctly....then it doesn't matter if all of his miles are AP tomorrow vs manual. Basically, you can't "fix" the previous safety scores. Each day has its own safety score, which gets a miles weighted average to your final score.
"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."
So today, lets say he had a bunch of hard braking, and got a 96. Tomorrow....it won't matter if he tries to manually do a bunch of good braking....its not going to "improve" his 96 score. Lets say he drives 500 AP miles tomorrow, and gets a 100. Great...he got a 96 one day, and a 100 the next day. The two scores (96 and 100) get miles weighted averaged to compute his final score. The main safety score app page might show the same hard braking % number, and if he drives AP the entire rest of the time that percentage doesn't go down, but that is not what is used to calculate his final safety score value.
FalconFour
Member
Insightful, and indeed Tesla did write these descriptions in a very suspiciously "programmatic" style of language... describing the entire formula calculation in a detailed enough way for people to be making spreadsheets. 
But the way I see it: each day's driving, shown on the daily page, isolates each day's contributions to the score and scores them individually (or by drive), but the overall score (page) is just the combination of all points together for a time period (currently infinite, since it started a week ago and continues to this day).
I've significantly improved my overall "following" score, for example, which initially had trouble because I'd rely on AP to mask what I didn't understand about the following scoring. Started off at 38% or so, now down to 2.1%. It keeps going down with every drive.
Then there's the score simulator, which just lets you punch in percentages to get a score. That would mean that the values, as displayed, dictate your score - so, if you can improve the displayed values, your score will follow.
Thus, keep racking up the good-points-farming
But the way I see it: each day's driving, shown on the daily page, isolates each day's contributions to the score and scores them individually (or by drive), but the overall score (page) is just the combination of all points together for a time period (currently infinite, since it started a week ago and continues to this day).
I've significantly improved my overall "following" score, for example, which initially had trouble because I'd rely on AP to mask what I didn't understand about the following scoring. Started off at 38% or so, now down to 2.1%. It keeps going down with every drive.
Then there's the score simulator, which just lets you punch in percentages to get a score. That would mean that the values, as displayed, dictate your score - so, if you can improve the displayed values, your score will follow.
Thus, keep racking up the good-points-farming
mackgriggs
Member
Insightful, and indeed Tesla did write these descriptions in a very suspiciously "programmatic" style of language... describing the entire formula calculation in a detailed enough way for people to be making spreadsheets.
But the way I see it: each day's driving, shown on the daily page, isolates each day's contributions to the score and scores them individually (or by drive), but the overall score (page) is just the combination of all points together for a time period (currently infinite, since it started a week ago and continues to this day).
I've significantly improved my overall "following" score, for example, which initially had trouble because I'd rely on AP to mask what I didn't understand about the following scoring. Started off at 38% or so, now down to 2.1%. It keeps going down with every drive.
Then there's the score simulator, which just lets you punch in percentages to get a score. That would mean that the values, as displayed, dictate your score - so, if you can improve the displayed values, your score will follow.
Thus, keep racking up the good-points-farming
The safety score simulator page...at the top says:
"Make Adjustments to the factors below to see how they impact your daily safety score"...not overall score.
FalconFour
Member
Nice job filtering out only exactly one thing you can argue against.
Meanwhile, my overall scores keep improving and I DON'T KNOW WHY, THEN
It'll remain a mystery, I guess!
mackgriggs
Member
Your overall score keeps improving because your daily individual scores are better than your past days.
Edit: Yes, the main safety score page numbers for bad braking/cornering/etc are getting better for you...but that is unrelated to your actual overall safety score, which is simply improving due to your daily averages getting better, since we all know better how to improve our daily scores.
Edit: Yes, the main safety score page numbers for bad braking/cornering/etc are getting better for you...but that is unrelated to your actual overall safety score, which is simply improving due to your daily averages getting better, since we all know better how to improve our daily scores.
No luck. I wonder if Teslafi has computed it’s own safety score. It’s not showing up in the Tesla app. Maybe I’m still in the queue for the FSD beta.
Similar threads
