But there was far less overtaking pre-DRS than there is now according to this site [1] which tracks such things. These stats don't include:

The overtaking figures for each race (across all data sets) do not include:

* Position changes on the first lap of the race

* Position changes due to drivers lapping backmarkers

* Positions gained in the pits

* Positions gained due to drivers yielding

* Positions gained when a car has a serious technical

problem; e.g. puncture, accident damage, etc.

[1] http://cliptheapex.com/overtaking/

There are more overtakes, but not _better_ overtakes. The vast majority are boring passes on the straights, with absolutely zero excitement.

Switch on DRS, breeze past the guy on the straight.

Yes, there were less overtaking. But whenever it happened, a large part of it was due to the driver's skills rather than sheer superiority of the car. The DRS rule says a driver can activate it only if he is within 1 second of the car in front. So when he does activate it, it's just like a free overtake and the car in front can't do anything to avoid because IIRC he can't activate DRS. This just doesn't make sense to me and takes the fun out of overtaking.

I don't think DRS was necessary in the old formula but with the new aerodynamic formula I think it is and will get much better. It's extremely hard to get within one second now so if you can do it you're much faster and only being held up due to the dirty air. Look at AUSGP, not much overtaking and I think it has two DRS zones.

Those restrictions on computing power (25 TFLOPS of CPU) for your CFD simulation cluster seem rather esoteric when it comes to racing cars...those hundreds of pages of regulations must be ridiculous to comply with.

In CAD they manifest themselves as a bunch of boxes in which you are allowed to put your shapes, so it's not overly taxing for the engineers to follow. However there is a risk that the 'box' layer is not done correctly at the beginning of the season and you end up missing something important.

The limit on computing power encourages you to think more carefully about what a certain component is interacting with the rest of the car. A few years ago I remember a CFD engineer running a massive DoE with hundreds of cases to get his stats up for the week when there was spare capacity in the cluster. If the limit were removed I doubt there would be any increase in innovation, just way more carpet bombing with parametric designs.

Yeah, the 25 TFLOPS thing is so weird. Is this just to ultimately force a trade off between simulation fidelity and cycle time for analyzing new designs? Can you not run multiple simulations on separate 25 TFLOPS clusters?

Also, how would regulators be able to verify how many simulations were run on how many designs?

It's an R&D spending cap, to keep the sport more competitive. Much like many pro sports leagues impose a salary cap, so the richest teams can't simply spend their way to victory.

While DRS may seem artificial, it may pay to re-watch Alonso's race in Abu Dhabi in 2010 where he was stuck behind a much slower car for what seemed like the entire race.

With the heavy reliance on aerodynamics and the issues with a car losing downforce in the wake of the one in front, you can view is as compensation for that effect and (if you stretch your imagination somewhat) closer to the pre-aero days of racing.

Since DRS has been introduced however, the number of overtakes in each race is steadily decreasing, so you may see a more radical solution in years to come. I personally would like to see wing-aero substantially reduced and grip / tyre size increased to compensate, but the massive investments in tech and people and aero's tendency to make the whole process mean-reverting make it a strongly vested interest for the top teams.