[ 2020 ] Aston Martin RedBull Racing F1 Team - Honda

[the EDGE]

Great read!
Pretty much how I felt when watching those on boards...
https://the-race.com/formula-1/red-bull ... ws.twitter

I could have told you the Merc will be the car to beat & the RB will look more nimble a month ago, so could any other F1 fan.

[Bill]

How would you come to that conclusion.last year Redbull was struggling with snappy unstable rear according to Max they solved it with these car but following other cars is tricky they lose a lot of downforce may something to do with cape .

Ferrari are still struggling with undesteer from last year so I don't get why u are being dismisseve of Gary

[seense]

Great read!
Pretty much how I felt when watching those on boards...
https://the-race.com/formula-1/red-bull ... ws.twitter

Thats what Johnny Herbert said as well on F1TV; That the RB16 looks amazing trackside.

Also interesting what Palmer said about setting the car up with harder tires. With harder tires you always set the car up with oversteer because you know the rear will be more stable with softer tires. I guess not that strange seeing lots of cars with oversteer on the white tires.

[GPR-A]

Great read!
Pretty much how I felt when watching those on boards...
https://the-race.com/formula-1/red-bull ... ws.twitter

I could have told you the Merc will be the car to beat & the RB will look more nimble a month ago, so could any other F1 fan.

These experts who stand at the track side to watch cars go by should understand they are using an age old technique to report the balance of cars to viewers, when there was no recorded video available to watch. That report seems like from stone age, especially with live coverage available with on boards. Such reporting becomes even more stupid, because with varying fuel loads, engine modes, track condition and what a driver is trying at a certain part of the circuit to learn about the car, complicates the visual behavior for a track side observer.

Red Bull ended as the second best car last year and this year's car is more of an evolution. Whatever the technical nuances are, which a driver experiences, but for an average viewer that car looks planted with so much of down force in these cars. So, it's easy to say a Red Bull is planted. No rocket science, no big deal.

While even with such high FPS videos available, it becomes difficult to observe the difference of behavior between Mercedes and Red Bull cars, someone distinguishing them with naked eye seems so ludicrous.

[nzjrs]

While even with such high FPS videos available, it becomes difficult to observe the difference of behavior between Mercedes and Red Bull cars, someone distinguishing them with naked eye seems so ludicrous.

This seems a little close to "If I can't see it, it can't be done". Another version of this "who needs drivers when we can simulate and measure everything".

The answer to both is, of course, that it's complicated. The human brain is good a integrating across scales of time and space, so it's not 'we have high FPS video why do we use our eyes', but something closer to 'its plausible that a human mind that has built a good model it's entire life watching cars driver around can come to a reasonable qualitative opinion more economocally that building a computer model of all cars from multiple high fps cameras and working very hard to isolate and measure those quantitative features that map to car performance'

One learns that the difference between qualitative and quantitative observation is quite a lot of work to work out what the interesting quantities are.

[godlameroso]

While even with such high FPS videos available, it becomes difficult to observe the difference of behavior between Mercedes and Red Bull cars, someone distinguishing them with naked eye seems so ludicrous.

This seems a little close to "If I can't see it, it can't be done". Another version of this "who needs drivers when we can simulate and measure everything".

The answer to both is, of course, that it's complicated. The human brain is good a integrating across scales of time and space, so it's not 'we have high FPS video why do we use our eyes', but something closer to 'its plausible that a human mind that has built a good model it's entire life watching cars driver around can come to a reasonable qualitative opinion more economocally that building a computer model of all cars from multiple high fps cameras and working very hard to isolate and measure those quantitative features that map to car performance'

One learns that the difference between qualitative and quantitative observation is quite a lot of work to work out what the interesting quantities are.

People underestimate how powerful the human brain computer is, the amount of information it parses is orders of magnitude more than what is available to your conscious awareness. Because there are so many processes in the body that depend on accurate signals that it's just better if the conscious mind isn't involved in these tasks.

[GPR-A]

While even with such high FPS videos available, it becomes difficult to observe the difference of behavior between Mercedes and Red Bull cars, someone distinguishing them with naked eye seems so ludicrous.

This seems a little close to "If I can't see it, it can't be done". Another version of this "who needs drivers when we can simulate and measure everything".

The answer to both is, of course, that it's complicated. The human brain is good a integrating across scales of time and space, so it's not 'we have high FPS video why do we use our eyes', but something closer to 'its plausible that a human mind that has built a good model it's entire life watching cars driver around can come to a reasonable qualitative opinion more economocally that building a computer model of all cars from multiple high fps cameras and working very hard to isolate and measure those quantitative features that map to car performance'

One learns that the difference between qualitative and quantitative observation is quite a lot of work to work out what the interesting quantities are.

People underestimate how powerful the human brain computer is, the amount of information it parses is orders of magnitude more than what is available to your conscious awareness. Because there are so many processes in the body that depend on accurate signals that it's just better if the conscious mind isn't involved in these tasks.

Eh... It's still the human mind that is looking at videos to more accurately derive the right information. You guys need to read my post again. How did you conceive this as a Human Brain Vs Something else?

Do either of you watch Cricket? There are things that happen at much slower pace in a run out, but yet, there is "Third Umpire" to review video footage to accurately make a decision, despite the on field Umpire standing there gluing his eyes to watch the run out happen.

Being old school is one thing, but denying that the technology is helping assess the information more accurately, is simply being ignorant.

[Bill]

If a car doesn't have enough downforce it slides around in fast corners that's just observation you don't need to be an expert a lot of cars in the midfield are a hand full.the Redbulls during the Vettel era were said to be on rails because they looked composed.

[ME4ME]

To be fair it's also easy to over-estimate the RB16's grip and stability when it hasn't been pushed closer to the laptimes the Mercedes did. We'll see next week or at the first Grand Prix.

[nzjrs]

Do either of you watch Cricket? There are things that happen at much slower pace in a run out, but yet, there is "Third Umpire" to review video footage to accurately make a decision, despite the on field Umpire standing there gluing his eyes to watch the run out happen.

(begin off topic)

Run-out, for example, and most other 3rd umpire interventions are binary decisions, so it's not really a useful comparison here. Perhaps a more analogous situation would be an automated system for classifying if Muralitharan or Malinga, or a new bowler's novel action is legal.

My point was equally pragmatic. Imagine you need to build an automated system for classifying if a new bolwing action 'more weird' than those two, are legal. First you would need to identify a quantitative measures that encode all aspects of a legal bowling action, and then engineer a system to classify it. On the other hand, an umpire who has watched bowlers their entire life has an internal mental classifier that would peform quite well at a first pass, and would be cheaper (free to build). It's not a perfect analogy, but it's not to bad.

So it is with watching cars on track. A life trained human has a pretty good model. It's better than noise, and it's error is probably better than an automated system initially.

(end off topic)

[scheffers]

Looking to be on rails in first testing week says zero. Ferrari was on rails during the entire two weeks last year.

[nzjrs]

Looking to be on rails in first testing week says zero. Ferrari was on rails during the entire two weeks last year.

This is false, but I'll play along and say it was true. Let's say that it looked like the Williams was also a dog on the tests last year. At the end of the year Ferarri was faster than Williams and it looked so on track from them in S3 at pre-seaon - so visual drivability of a car in pre-season does have some predictive power (it obviously says more than zero).

Now that we agree that it says more than zero, we really only disagree on how much it does say, particularly how much the opinion of a trained observer says precisely. That's an interesting discussion.

(my off topic point was that when one needs to build a quantitative model to assess the performance of something from data, for example, that model (often) needs to be developed in conjunction with a trained observer in order to ascertain what is quantitatively interesting. this means at the start of model development, just asking the observer can be more accurate. they are in some sense, the positive control for the model)

[etusch]

Newey is unhappy with 2021 regulations

"It makes it a little bit GP1 which is not what I think Formula 1 should be.

"But it's been pushed through regardless of what people think, so whether it's good for the sport or not, only time will tell," he concluded.

https://insideracing.com/index.php/form ... Hi7DhXjxYo

[scheffers]

Looking to be on rails in first testing week says zero. Ferrari was on rails during the entire two weeks last year.

This is false, but I'll play along and say it was true. Let's say that it looked like the Williams was also a dog on the tests last year. At the end of the year Ferarri was faster than Williams and it looked so on track from them in S3 at pre-seaon - so visual drivability of a car in pre-season does have some predictive power (it obviously says more than zero).

Now that we agree that it says more than zero, we really only disagree on how much it does say, particularly how much the opinion of a trained observer says precisely. That's an interesting discussion.

(my off topic point was that when one needs to build a quantitative model to assess the performance of something from data, for example, that model (often) needs to be developed in conjunction with a trained observer in order to ascertain what is quantitatively interesting. this means at the start of model development, just asking the observer can be more accurate. they are in some sense, the positive control for the model)

Agree. Zero was not the right word.
I got caught up in the red line of the topic, "whether rb16 will be a car to beat the mercs with or not".
Looking at it with that question in mind it says nothing, since the on rails comments were made last year about Ferrari too..

[Juzh]

Looking to be on rails in first testing week says zero. Ferrari was on rails during the entire two weeks last year.

Ferrari exhibited understeery behaviour from the get go last year, even if it seemed to be a stable car. It was obvious when during the fastest lap vettel managed he missed pretty much all apex' on entry.