Reader Submission #139 – The Official Mazda 787B

You’ve probably heard much rejoicing as of late from the Assetto Corsa community, as the PC version of the game has recently received a substantial software update that has been long-overdue for what has otherwise been a very incomplete racing simulator. Bringing with it proper pit stop strategy configuration screen as opposed to a Mario Party-like pit stall mini-game, the rudimentary implementation of driver swaps, and even a couple of new free cars from completely opposite ends of the spectrum – Mazda’s Miata and 787B Prototype, it appears the sim racing community have finally won out in the end. After years of staff members from Kunos Simulazioni angrily berating their users for “expecting too much” and “not understanding the purpose of Assetto Corsa” the team from Vallelunga are now slowly beginning to insert specific features and functionality sim racers have been requesting for years on end, indicating individuals the developers at one point labeled incessant whiners may have had actually had legitimate complaints about the direction of the simulator.

Regardless of how we’ve gotten here, I’d like to extend a thank you to all Assetto Corsa owners who risked multiple forum bans and being blacklisted by rabid fanboys for being very vocal about what the simulator lacked; it took a while, but Kunos’ recent additions to the simulator confirmed you guys were much more than just “trolls” and “haters.” Because of your diligence, Kunos are actually getting to work on making Assetto Corsa a much more feature complete piece of software. Good job!

However, with every twist, a turn. We have heard for several years that Kunos Simulazioni build cars within their simulator using an abundance of real data, often times pushing this element of Assetto Corsa to the forefront as a way to compensate for the shortcomings of the simulator – sure, there’s not been a lot to do until recently, but at least the cars are incredibly accurate, right?

Today’s Reader Submission notes that is not the case.

Hey PRC. There have been some posts on various forums about issues with Assetto’s quality of physics, or more specifically, the quality of the work pushed out by Aris under the Kunos banner. The fanboy army led by Stefano and his buttlickers seem to jump and try to dismiss legitimate discussions or questions. We have seen with many people, from banned users to the guy trying to find information for his mod based on his real life car. Having read a few of those hammered posts, I picked up on some aspects of what to look for thanks to the detailed info provided by the gurus and the nagging questioning brought up by certain users, including guys who DO release mods for Assetto. 

The Porsche from DLC pack 3 got postponed due to Kunos needing info, stuff missing, real life correlation, etc. Their words paraphrased. Well, how much of it is actually true? Do they really have the manufacturers go through everything and actually inspect the car? I call bullshit. That’s some yellow propaganda. Then to see them acquire mods and re-release them as holy grail content, as if the original mod wasn’t good or even superior, seems unfair. So with that information, the recent update and the possible flame coming up from the questions on the Porsche and the Mazda, I checked the following on the Mazda since it was freely available before. Note that all measures of CL and Downforce are in KG at 200km/h.

What you see in the picture above is the 787b with highest downforce achievable before the stupid loss that takes place. I’ve no idea how Aristotelis comes up with his stuff.

Next, we have the maximum downforce achievable while maintaining less shitty balance (still rubbish), so theoretically this is roughly the max downforce possible with 35% forward aero balance.

Third, I will compare everything to the other official prototype car, theCc9 they made which is Le Mans-specification. And remember, the 787B is supposed to NOT be Le Mans. Roughly this is the max downforce in a straight line.

Lastly, I will do the same as with the 787b, giving it a more functional 35% balance. The car actually makes a corner like Eau Rouge instead of just understeering off like a wooden box.

The value we have to look at is TOT CL: x.xx in the bottom of the app on the screen (I left the HUDs to be informative). The max for the 787b (1st image) is 2.8, the usable max is just 2.5cl. The C9 is 2.54 and the usable max is 2.39cl, so the range between the two cars (one Le Mans spec and one supposedly not) is 0.4CL at 200kmh, which equates to roughly 154kg of downforce.

Nowhere is an interesting thing that seems to relate to what the people are moaning about. The drag coefficient (CD) is much higher on the 787b than the C9 BUT the difference is the same as the downforce difference at ~0.4 CD (which = the .4cl range of downforce difference). With this drag you can say the car is not the LM-spec but if you go HERE and HERE (one of them was a link posted in the forums, I found the other from there. Great site!), the story looks wrong. There you find the downforce levels of comparable sprint-spec cars of the time. The C9 has cl of 4.47 @ 241km/h in 1989, the C11 has a cl of 5.36 at 241km/h in 1990. So the issue that follows is how the hell is the Kunos 787B, from 1991, performing at less than half of a car from the year before and much less than a car first developed 3 years prior?

So the main problem highlighted here is the downforce. The 787 is within .4cl of the Kunos C9 Le Mans specification but it is listed as a standard, non Le-Mans spec. So it is much closer to the C9 Le Mans spec than it is to the data suggested by the websites linked above showing the C9 Sprint (non-Le Mans) and C11 Sprint. Do they really pursue and get the information for the cars? If they do, why is it off in the game? What the hell are they doing to the cars to recreate them this way? I wrote all this for the Mazda but imagine the can of worms from the 2017 Porsche, being so different to real life according to mclarenf1papa? How can we trust that developer when they are consistently caught out with “alternative facts”?”

Kunos, in my opinion, likes to spin their information around with support from their fanboy army to portray an image that their content is always better, including the free mods they acquired. Their stance on waiting for data and a data sheet appears to be bullshit because you can right away check the downforce levels of the cars and how the diffuser makes no sense. Often the ratio varies wildly with higher ride heights generating over 100% of downforce. So when you feel the car understeer weirdly it’s because it went below the magic ride height number.

I personally doubt they have numbers for the latest Porsche as they made the claim. They probably had the company give the green light on the model and maybe engine, nothing beyond that. Meanwhile, modders get access to team manuals with legitimate air tunnel data and measurements. They are actually able to recreate the aero map very well (credit where it’s due) but Aris has no clue (modders words) about what he is doing. I don’t have time right now but if you extract the ACD from the cars, you’ll see the optimum heights and how it makes no sense how the downforce relates. Aris makes the diffuser have the wrong impact and instead of letting it stall at some point, it makes it not work.

People are circle-jerking over the latest update but I’d not doubt the 787B is much worse now than before IF they actually went over the original numbers made by the best guys. The Le Mans C9 had that issue of going below the magic ride height and losing nearly 100% of downforce. Now, the main thing we all know is you want the car as low to the ground as possible, just before scraping…. Not in Assetto.

Thank you for your very in-depth research, I must admit I’m a bit over my head here, but what you’re saying, as well as the data (and real-world tables) makes sense. I’d like to know as well how Kunos are claiming to have real data for cars, but the sprint variant Mazda 787b inserted into Assetto Corsa with the recent update has roughly the same downforce levels as the Le Mans spec Sauber C9. Obviously, it’s not right, and I hope it gets rectified. It also calls into question what other phantom numbers have been thrown into other cars, but we knew they did that already.


75 thoughts on “Reader Submission #139 – The Official Mazda 787B

  1. To be fair, the 787B is one of the most overrated cars in sportscar history and people just wank over it because it has a rotary engine and the sound it makes is good.

    It didn’t win Le Mans because it was the best car, it won Le Mans because all the main competitors crashed out or retired for mechanical problems. Pure performance wise, the car was 15 seconds off-pace from the car that was on pole which, surprise surprise, was a Mercedes C11, an evolution of the already good predecessor, the C9.

    I agree that the behaviour of the Kunos version of the 787B is average at best and subpar at most, but that doesn’t mean that a 1991 car should automatically be better than a 1989 car just because it was made 2 years later. That is just a retarded argument, no matter who you’re pointing your finger at (Kunos in this case) to prove something.


    1. yeah it won on reliability but having less than half the downforce (in game) than the real life C11 would put it on another planet in terms of performance. It was generally 5 seconds slower at tracks other than le mans than the C11 from some super casual research and there is no way it had that small of a deficit while producing the same kind of drag and half the downforce with less engine power (700 for the mazda and 730 for the C11) 😀


    2. There’s nothing to say the year of build should be the ultimate dictator in performance. Just look at that monstrosity Bugatti unleashed in the 1930’s that would have been competitive two decades later. The 787B won Le Mans but due to a bunch of factors outside of its pure performance, that is well known. The car is beloved for its looks and engine sound, never praised for performance.

      The point is how the Mazda 787B released by Kunos is not really a Mazda 787B. Period. The guy has proven that.

      Liked by 2 people

    3. Agreed the 787 was a winner but not a fast car.Because there are so many 787 fanboys I think the game makers fudge the numbers to make it a faster car than it would have been.Forza is only a simcade but the Mazda 787 has always been one of the fastest cars in game when a standard Peugeot 905c is about 9 secs faster around real life.


    4. That only adds to the beauty and hilarity of that win. Rotary winning on … wait for it … drumrolls … RELIABILITY AND FUEL ECONOMY.

      Only in real life. Authors can’t write fiction this good.

      Liked by 1 person

  2. Why is this autistic submission guy doing such a in depth analyze of some car in simcade Assetto Corsa when he could spend that time playing the real simulator aka rFactor 2.


  3. Assetto Hotlap had one unique advantage in the steering feedback and realistic feeling on dry track surfaces though others are now catching up. The rest of the game is mediocre compared to both PC and console alternatives. The arrogance of the Kunos development team makes this story no surprise. The data likely has to fit their game or is otherwise deemed unworthy and ignored. The console port debacle offers a different perspective on the supposed competence of this developer. One minute they are a small developer without the resources to fix basic problems, the next they expect us to believe they are the most accurate data-driven virtual car designers in sim racing? The sooner the next generation of driving games is released the better when they also include realistic steering feel Kunos will be a quickly forgotten footnote in the genre.

    Liked by 1 person

  4. “And remember, the 787B is supposed to NOT be Le Mans.”

    The 787B in the sim is the Le Mans version, you can tell because it’s missing the diveplanes and has smaller endplates on the rear wing. It’s not down much downforce-wise because both cars use the same undertray. TBH, I find it drives a lot like the 962C ST did before the aero was patched.


      1. Also, the 787B was slow as hell, but I’m sure you know that from the in-depth research. The only reason it won Le Mans was because all the competitive cars broke.


    1. so why is it not stated in game like the C9? and why do we keep getting le mans spec cars? and how do you explain the 25% increase in drag from the C9 to the 787b?


          1. “so why is it not stated in game like the C9?”

            No idea. Why is it not stated for the 919 or the 935/78 or the 908LH?

            “and why do we keep getting le mans spec cars?”

            Hell if I know. Probably because those are the only historic sportscars most people know, and also partially because they’re in Gran Turismo whose terrible content is what everyone apparently wants to emulate now.

            Personally if I were designing a sim all of the 80’s prototypes would be IMSA GTP cars which would make the point irrelevant. The NPT-90 and Eagle MkIII were generating so much downforce (8000-10000lbs) that they were bending their rims.



      1. The only 787B people know or care about is the Le Mans winning one, the orange and green one was immediately retired and has never been in any other configuration since.


  5. Not really a surprise you can’t understand the user submission cause it makes no fucking sense. Diffuser stalling when the ride height drops too much is a well-known phenomenon and is such a problem IRL that top end series all ban running the flat floor too low and including certain types of fences that make the problem worse.


    1. In a ground effect F1 car, you wanted to run almost-scraping with the skirts sealing off the airflow. In a GTP or Group C car you wanted a gap between the side of the car and the road because the air coming in from the side created most of the downforce.


    2. This is exactly correct. Unless I’m reading this article wrong (and the way it’s written, it’s nearly incomprehensible even to this Johns Hopkins grad with 4 years of physics), the author doesn’t seem to get the concept of diffuser stall.

      With a diffuser, if you lower it too close to the ground, the airflow from underneath the car starts getting cut off and you lose 100% of the diffusers downforce. Even worse, it happens intermittently (road surface is never flat) and then you get bizarre porpoising effects that start to harmonically resonate with the suspension. Heave springs ended up solving this by the late 90’s, along with high noses to allow inherently greater flow to the diffuser area.

      Why do you think active suspension was such a giant advantage in F1? It was purely for ride height management to keep the diffuser (and front wing) at optimum height off the surface.

      Seems like, from what I can tell, that’s what is happening with the 787B. You get increased downforce at lower ride heights, but if you exceed a critical threshold, you get *no* downforce.

      Downforce (lift) isn’t some magic thing. It’s just Bernoulli’s Principle: As the speed of a moving fluid (liquid or gas) increases, the pressure within the fluid decreases. By accelerating flow under the car, you generate a pressure differential that we call “Downforce”. But there has to be *flow* in the first place.

      Liked by 2 people

      1. The game has varying downforce levels that aren’t progressive by height, it jumps all over the place. In some cars, the downforce at higher ride heights (over 4cm, nearly 5cm… which is too high) is actually over 100% of the max value assigned to the car. In Assetto you want your car at higher ride heights than you would actually believe ideal, or that you’d see on teams’ setup sheets.


  6. Was driving this car a bit and it feels like it has not enough downforce. The downforce of the 962c was completely broken on release and i´m over 7 sec. faster now at Spa thanks to an update (which never happend officially). It´s better to judge the cars few weeks after release and it may occur, that the 787b is as fast as the 962c today. Maybe it´s politics and they want to make this Mazda slower than the Porsche. Maybe they tinkered the aero to match real lap times. The Ferrari 488 GT3 was overpowered on release and not anymore. I would not wonder if Ferrari was telling Kunos to make it faster as part of the licensing-deal.


    1. “which never happened officially”

      From the changelog:

      – Corrected aero downforce and drag values for Porsche 962 Short Tail
      – Corrected drag values for Porsche 962 Long Tail


  7. I have not been here for a while…
    I did know about Austin now working for Mr Bell.
    Oh, and I will not be back after this.
    However, to take a leaf from your book Austin, remember this?

    Oh dear Austin…

    You have been ‘played’.

    Ian has played the long game with you, and you have fallen for it.
    Whatever you say about PCars1/2 will now been seen as, ‘paid for content’.
    PCars people will be lining up, or have been lined up to fill your comment box.

    Your credibility has gone.
    Ian has played you beautifully.
    You have put yourself in a box, a box that Ian opened for you.
    And you ran in.

    Credibility; nice chat with iracing management, via the telephone, and you get $755.60 refunded to you.
    Iracing does not do refunds to banned members, however you got one…

    Perhaps you could explain, exactly how you got the refund, you know to help other ‘banned’ members, because if they refunded you, they have to refund every other banned ex-member.
    You could write a nice piece about the process.

    Still, you got to buy a nice new helmet.

    Credibility; member of your team testing PCars, ‘unpaid worker’ for Ian Bell.
    So how does that work?
    And how does that look, when you consider your credibility.
    I know you have tried to explain why your guy, is now Ian’s guy, however you really should of respectfully declined the offer.
    To preserve your credibility.

    You have been burned by someone who does it for a living. Mr Bell plays with big boys everyday.
    Making money is easy, if you are prepared to use and abuse people.

    Austin, you have been used.

    Game over.

    I posted this sometime ago.
    I was right.
    You have been played.
    Silly boy.
    Still, you have got your silly little car.
    For your silly ovals…
    Credibility gone.
    And you will never, ever get it back.
    Remember, people like Mr Bell don’t keep people around for ever.
    When he has finished using you, you will be dumped.
    He will not return your calls.
    He will not answer your emails.
    So enjoy your car..
    While you still have it..


    1. Another hater speaking obviously

      I wish you deeply from my heart cancer and soon your death. Idiots like you shouldnt be allowed to life or to reproduce.

      Good that at least youre not on this page anymore


      1. Here again we have this “get cancer” and “shouldn’t be allowed to life(sic) or to reproduce” thing.

        From this, I can divine 4 things about this particular anon shit-poster:

        1) English is not their primary language. I hope.
        2) A total lack of imagination with regard to flinging insults.
        3) Machine-like intensity.
        4) No real sense of humor.

        Ergo, the following conclusion is inescapable:

        It’s a Russian Troll-bot.

        Liked by 1 person

      2. This cancer comment guy appeared shortly after Austin announced his love Ian Bell. Proof that PCARS fanboys are the worst of the worst.

        Austin has found a good home.


        1. Is there such a thing as a PCARS fanboy nowadays? If you look on YouTube, it’s just one video after another slamming that poor “sim”. For something that sold OK, it has very few supporters willing to speak up.

          I do like the 2 DRM/Gr 5 cars it has, and night+storm races on the Nordschleife are actually really fun (reminds me of growing up on Lookout Mtn, TN and racing in the rain with friends – incredibly dangerous and stupid, but it’s where I learned to drive).

          Maybe PC2 can have some actual QA this time. It’s reassuring that Austin is over there hopefully representing our interests. PCARS coulda/shoulda been the PC version of Forza/GT, but with good physics/FFB.

          I admire that vision, but like they say in American Football: Execution Matters.


  8. So after reading the article and commentary here, I will also post my thoughts but hopefully broken down in easy to follow paragraphs:

    Why are people defending the cars released when they are used as an excuse to only be Le Mans version? Even being just Le Mans, that’s a big loss of downforce and the correlation of numbers isn’t ….. correlating (if only this could have been a funny pun). Assetto lands the first Porsche content after the EA deal and they can’t EVER get a non-Le Mans spec version of a car?

    What’s the problem keeping Aris from releasing a car that has downforce and actually drives like the real car should? The Lotus 98T was a joke and supremely fast. What’s the excuse there, “F1 doesn’t run Le Mans”? :p

    It’s as if since the 98T it’s like we only get Le Mans crap, and even then with senseless stuff. Look at other cars and their acd file how the downforce and ride height values make no sense. The argument isn’t them only giving the LM-spec excuse, it’s the car not being adequate. THE CAR NOT BEING ADEQUATE. And I wonder if anyone will look into the GT3 Porsche and its information because it doesn’t correlate with real life and performance loss in one area is being made up with gains in another. HOW ON EARTH is that publicized as an accurate depiction?

    Lastly, why are people so keen on giving excuses for developers instead of holding them accountable for their work, and above all pushing for better quality?! What happened, when did things shift so insanely like this? It doesn’t matter if it is Assetto Corsa, rFactor series, pCARS, R3E, iRacing, or any other title that actively advertises itself in a way it doesn’t deliver.


    1. Additionally:

      If the car is being disguised as Le Mans, why is there absolutely no mention of it in the game? There’s too much consumer disservice going on, and above all, the neutering of a car that is supposed to be incredible to drive.


      1. The 787B won 1 race in its entire career, which was Le Mans. It never even came close at any other race. So (a) yes they chose to do the winning car, which is the Le Mans car, and (b) of course it has shitty aero compared to the cars that were winning races.

        Liked by 1 person

        1. It doesn’t have that poor aero with such little downforce. Their Ferrari F1 cars are way understated in that aspect as well. It is the developer impacting the car, not recreating the car.


    2. “that’s a big loss of downforce”

      For cars made after 1989, that’s really not the case. The difference between the Sprint and LM versions of the 787B would be a few percent. It’s not like the pre-1990 cars such as the C9 where the Sprint version had over twice as much downforce.


  9. “the range between the two cars (one Le Mans spec and one supposedly not) is 0.4CL at 200kmh, which equates to roughly 154kg of downforce.”

    The Mazda has 468kg of downforce 185kg of drag, the C9 has 446kg of downforce, 138kg of drag in the screenshots. Which is 24kg of difference. Dunno why you chose to read the “TOT CL” to determine how much downforce the cars have when it’s printed right there.


  10. There is no way a C9 has less drag than the 787B, a much smaller and less downforcy car, in any configuration. Aris is pulling numbers our of his Ass.


      1. I base this on real life data for the C9 vs a simple relation of power vs acheved speeds from the Mazda, You dont need to be Adrian Newey to figure that one out.


    1. ok associat0r
      Too bad we only get simple cars to simulate in rf2. That way you can never go wrong, or can you? I guess people just aren’t bothered with checking how accurate those cars are in rf2. But we have associat0r telling us, so is enough.


  11. They didn’t buy you a car, so i guess you’re entitled to slander Kunos. Now go suck “mr. Bell” as you so lovingly name him now, and lick where you spat before.


Ratio of vowels to consonants will be monitored. Post at your own discretion.

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s