On 10/14/2019 5:55 PM, ~misfit~ wrote:

On 15/10/2019 2:48 AM, Brian Lawrence wrote:

On 14/10/2019 12:06, larkim wrote:

On Monday, 14 October 2019 10:05:06 UTC+1, Out Cider  wrote:

CS <crmstone@gmail.com> writes:

And I do like the new tyre wear graphic.  Wonder how they do that??

I don't. Mainly because I don't believe it is correct, nor do I believe >>>> that it can be. I don't think tire condition can be counted in
percents,
as it is not a single, quantifiable and measurable thing - except in
video games.

--
OutCider

Agreed in terms of accuracy.  Not sure how they factored differential
tyre wear for different sides of the car.  If it had any actual data
behind
it, it could be useful (e.g. if tyre slip, wheelspin, heat etc
sensors were
feeding back to the TV company somehow).  If it's just a straight calc
that rear right = highest usage, rear left = 90%, front left and right = >>> 70%, and then just apply expected tyre life to that formula then it
isn't
especially meaningful, though it does help as an aide memoire when
watching
to see where tyre life *might* be expected to be.

Looked at the graphic again today, and noted that it said 'Insights,
powered by aws' briefly when it appeared on screen. It seems that aws
is Amazon Web Services, which led me here:

     <https://aws.amazon.com/f1insights/>

    <https://aws.amazon.com/solutions/case-studies/formula-one/>

There's a link to a 10m talk by Ross Brawn:

     https://www.youtube.com/watch?v=phCSmKjBfEc&feature=youtu.be

Thanks for the links, interesting.

Personally I've not got a lot from the AWS on-screen stuff that's been introduced this year (so far). Then again I've been a student of F1 for decades and am capable of keeping a lot of the info about each of the
leading cars in my head in real time and can tell if a drivers likely to
be overtaken in pitstops etc. I can tell roughly how good tyres are by knowing compound, stint length, lap times, watching how it corners and
gets off corners and whether the car's been running in clean air or
closely following another.

So I guess most of this stuff is for casual or new F1 fans, to make it
more accessible to those who haven't seen every race for decades and
read a bunch of info on the technical side of things.

What I found interesting from Ross' talk was when he said that they have access to and are feeding the data into AWS that not even individual
teams have access to 'for the fans'. I assume he means the teams only
have their own data but AWS is being fed data from all teams.

Surely then the people that AWS (at least when the machine learning
matures) is going to benefit most are other teams? I mean if it's as
good as they're touting then rival teams are going to be able to infer
things about their competitors that perhaps they didn't already know
based on what's being displayed in the "F1 Insights" box on the world feed.

Maybe.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)