GAME vs REALITY: Testing the Toyota GT86 on Slovakiaring
Article about comparing the virtual and real car, on the same track, at the limit.
After my comparison test of the Mazda MX-5 and the Porsche 911 Turbo S, I’ve been waiting for a long time to have an opportunity like this. Recently given the chance, I’ve prepared the below extensive material for you which is based on a Toyota GT86, checking both its performance and behaviour on Slovakiaring and comparing it with her virtual counterpart in the Assetto Corsa racing simulator. If you want to know how close the simulation is to reality, you’re in the right place.
I love to analyze driving and performance of both virtual and real cars in order to estimate how fast you can drive during specific conditions. I always try to search for differences between simracing and reality and answer how can you can use both realities to be as fast and consistent of a driver as possible.
photo: Kacper Nowogrodzki / AUTSIDER.pl
Something about "Hachiroku"
Toyota GT86 is a model which was developed in cooperation with Subaru. It contains many references to older Japanese sports cars. As an interesting fact, despite having the “86” moniker, which can be associated mostly with the AE86 (being legendary in Japan and for the fans of “Initial D” manga/anime series and Keiichi Tsuchiya), this model also follows the footsteps of the Toyota 2000GT, another great sports car from Japan which also left a major mark in the entire history of automotive industry.
Trivia:
● The number “86” is not only common with the name of the Corolla AE86, but it’s also a reference to the new engine’s bore & stroke, which equals 86 millimeters ,
● As it was in the 2000GT, Toyota GT86 also has a 2-litre gasoline engine ,
● For Initial D fans – Toyota GT86 is produced & mounted in Gunma prefecture ,
● Application of a flat-4 engine refers not only to Subaru, but also to the Toyota 800 from the 60s, which was one of the first sports cars in Toyota’s history, and was also equipped with a flat engine type .
The GT86 is a well-balanced car (having a distribution of 53:47, which means 53% of the weight is on the front and the remaining 47% on the rear), although with relatively low power, producing 200 HP. But, nevertheless, in case o
f naturally-aspirated 2-litre Boxer engine, it isn’t a necessarily a low figure. I could actually swear to everyone, if you enjoy cars like the Mazda MX-5, I can bet my money on that you’ll also enjoy the GT86. Sporty Toyota may lack in power comparing to other cars available on the market (although with too much horsepower, it would be too competitive with the Supra, a flagship model for Toyota), however, despite having only 200 HP, it’s still bloody fast on the corners and with a good driver, with its handling abilities around the corners the GT86 can definitely still be a “serial supercar killer”.
I’ll be comparing the real GT86 with its virtual version, available in the Assetto Corsa video game (I drove on a PC, but the game is also available on PS4 and X1). The versions of the car slightly differ, as the real one is a post-lift and the virtual one being a pre-lift version, but I can sum up the changes as mostly cosmetic. The track which I drove on was Slovakiaring. This track is available in Assetto Corsa as a mod – an additional content from a person who is independent of game’s developer.
Unfortunately, the virtual track isn’t laser-scanned and I cannot provide precise accuracy within the range of 1cm. However, the circuit as a whole, its profile and reference points are similar to their counterparties in a very much sufficient way, enough so for making a totally honest virtual & real car comparison.
You can check even on this photo comparison, the virtual track unfortunately isn't perfect, but still proper enough :)
Diving into the minor, but still existent and relevant differences between the virtual and real car – in the game we have a pre-lift model, and in the real life I drove a post-lift model. The younger version has little mechanical upgr
ades, for example a tiny bit more powerful engine (from 200 to 205 HP), shorter final gear ratio (from 4.1 to 4.3), improvements on springs, dampers, sway bars and small tweaks in aerodynamics. But still, these aren’t enormous upgrades, not enough to notice any huge differences between both cars, however I did want to have this stated on the record for maximum precision.
Wheel alignment in the real car was adjusted for “road stock”, so in the simulator I’ve done a very similar setup, even if it isn’t the fastest and most comfortable for driving on the circuit. Tyre pressure was around ~1,9 bar (~27 psi) on cold tires. In real life I’ve had a full tank of gasoline, so I’ve done the same adjustment in Assetto Corsa. Furthermore, I’ve had the Michelin Pilot Sport 4 tires on the real car, which are very close to the “Street” tires from Assetto Corsa in terms of their behavior and performance.
How doest the GT86 handle?
GT86 is a very balanced car, as the front-to-rear weight balance equals 53:47, with a little more being put on the front axle. It’s not actually a disadvantage, because during driving we are putting more weight load on the rear axle, due to the weight transfer. Hence, the car’s balance is nearly perfect as a matter of fact. The car remains very precise and predictable in its behavior. Even though, despite the low-powered engine and very good balance it’s not a problem to initiate a drift on the low-speed corners with its road tyres.
Despite the lack of power comparing to another road sports cars, you can have a lot of fun with Hachiroku, trust me! ;) [photo: Kacper Nowogrodzki / AUTSIDER.pl]
Basically, Toyota is equipped with a TorSen differential, which works very well not only on the track, but also in the standard road driving conditions. What’s more, GT86’s low centre of gravity gives a good steerability. I can say that, even though there is a very similar behavior between both virtual and real car, it was way easier to handle the real car and enter with more speed into high-speed corners, whereas the virtual car has had more tendencies to understeer. The braking performance in real car wasn’t as effective as it is in virtual car though.
Suspension & steering system
Suspension works very well on the track and it combines both comfort and precision of driving, such as would be expected of a road car. There isn’t any “boat” effect and you don’t feel excessive tilts during hard-braking or driving in fast-corner sections. In case of hotlap driving, I’ve felt the need for harder suspension settings, especially for the front axle, which probably should improve driving precision even more, without the risk of understeer. Maybe it could even eliminate the GT86’s desire to oversteer during the exit of low-speed corners. But that part of the car’s behaviour does heavily depend on the tires and alignment.
In my opinion, the car’s general behaviour, suspension work and characteristics are in fact very similar in both real and virtual variants of the car. Steering system in the GT86 is very precise, having a short ratio, even for a road car.
If I have to pinpoint something negative, it would be mostly the car’s desire to nose-dive during braking from high speed, which makes trail braking much more difficult. I think that a little bit stiffer suspension would be a good solution (except for sport tires ofc.) for that. But I know this is still a road car and there must be a compromise between handling and comfort.
photo: Kacper Nowogrodzki / AUTSIDER.pl
Braking system
Braking system in the real car which I drove (it’s worth noticing that even if the brakes are “premium versions” in a lot of road cars, in my personal feeling, when compared to a few racing cars, they are still “stock systems” – i.e. you can still feel these were made for a road going car) was stock and relatively effective. As you’d expect from a road car, braking system is balanced for a standard user, providing a little bit more understeer due to safety reasons, which wasn’t helpful for trail braking. Definitely though it does provide a good and safe braking balance.
ABS system works very precise and it supports the driver without disturbing him. Braking points or fields in real life are almost the same as on the virtual track and car in Assetto Corsa. I’m writing about “braking field”, because there are a lot of important factors to consider on which the braking performance depends – such as the whole car’s setup, tires which are used, track conditions, grip and also driver’s skills and technique.
Comparing both cars, the real GT86 doesn’t have as effective of a braking system as the virtual Toyota does, but these aren’t huge differences. Difference between braking points equals around standard car’s length (~4 metres), but it can be mostly a result of an imprecise model of the virtual track and tarmac used. Finally, if I would brake in real life, right exactly on the same spots as I did in Assetto Corsa, I could miss few braking points and go out of the track. But it’s important to check the differences between the real and virtual tracks. As an example I can use the braking to the T1 corner, where you’ll be braking from around 200 km/h’s. In the virtual car, where you see a board with “100 m” information, in the real life in that area there is a “50 m” sign. You can also notice a few differences in the angles, profiles and distances between the corners, straight parts and kerbs, which aren’t correct if we want to be “1:1 precise”. The main reason for that is the mod was made by an independent person, who wasn’t connected in any way with game’s developer (for the PC version of Assetto Corsa you can develop your own mods, such as a car skins, extra apps or even tracks and cars by yourself). Regardless, I’ve tried to search for the most similar reference points and to prepare these virtual conditions as close as I experienced it in real life.
Going to check the track! [photo: Kacper Nowogrodzki / AUTSIDER.pl]
Engine & Transmission
Almost every simulation process features simplified models, which doesn’t include many factors, e.g. power loss, but it’s relatively standard and normal to be so. In reality, for a lot of cars, especially with low-powered, low displacement engines, where the Toyota GT86 belongs, the amount of air temperature or its pressure, engine & fuel efficiency and a few more factors are important. Despite that, the speeds which were achieved and the whole engine characteristic was accurately similar comparing to the virtual car.
Upper boundary of 200 horsepower is reached on 7000 RPM’s and maximum torque of 205 Nm’s is on 6400 RPM’s. When using the higher rev levels of the engine, above the 7000 RPM’s and still before the red line, you can feel that the power is decreasing, according to the characteristics of a 4-cylinder Boxer engine. So, going above 7000 RPM’s on the straights isn’t effective in any way.
It’s actually very similar in Assetto Corsa, where you can easily notice this lack of power and time gap increase on the official “Delta Performance” app. This app compares your actual lap time in real-time with your best one from actual session, and thanks to this app, it’s easy to notice that revving the engine above 7000 RPM’s isn’t very efficient. You can spot the lack of power with your bare eyes, even by watching closely the speedometer, where accelerating will be slower. It’s also noticeable in real life.
The last difference, which isn’t major but I did notice it at the very beginning, is a different final gear ratio in the real car. Toyota GT86 had a 4.1 final gear ratio at first, but post-lift model has since changed that value to the 4.3, which means every gear should be noticeably a bit shorter, which in turn provides better acceleration at the expense of maximum speed on each gear. In places where you would be using the 2nd gear in Assetto Corsa, I don’t think it would be as effective in the real life Toyota post-lift model as it was in virtual reality. And on the topic of gears, the downshifting from 5th to the 4th gear in the real Toyota GT86 was sometimes a little bit problematic, because the gear lever doesn’t shift precisely, even if you are doing everything properly (and I also catched that on the reference lap).
I would also like to notice that the wheel sizes on both cars are almost similar (215/45 R17), which also affects the final ratio on each gear – to simplify it, tyre size could affect the max. speed on each gear and acceleration, just like changing the gear ratios, but int this case - it doesn’t. To sum up – both cars behaviour and characteristics are nearly the same, especially in terms of engine’s performance.
Summary
Toyota GT86 is a very interesting sports car for someone who wants to have something sporty with fancy & modern design and not necessarily a high-power engine, for example because he or she wouldn’t feel comfortable with it or doesn’t need more power. This model is not only a good addendum for the lineup of sports cars in general (not only for Toyota), but also among cars from another manufacturers with similar performance and naturally-aspirated engines.
It’s difficult to compare this car with other sports cars or even hot-hatches, which are very popular nowadays. A lot of them have engines with similar displacement, but they are equipped with either a turbocharger or a supercharger – and comparing cars like this in the field of their power or acceleration is unreliable. There is a reason why in racing we have a separation between cars in terms of “classes”, depending on their engine’s displacement or if these are equipped with turbo/supercharger or not. A 2-litre Turbo engine won’t be in the same class as 2-litre N/A engine (which Toyota GT86 has), because it wouldn’t be fair.
For the best possible comparison of my testing, I have prepared a video of both a virtual drive and a real on-board, which shows how close both cars are between themselves in the field of performance and behavior. As we are talking about and considering mostly hotlaps and racing, it’s important to discuss the most important thing in circuit driving – lap times.
Check the video with on-board comparison!
Real car: 2:43,21 (measured by the RaceChrono app)
Virtual car: 2:43,313
The difference between lap times comes mainly from the tire model used in the simulator, imperfection of the virtual track (it’s not laser scanned, as mentioned earlier) and also in a major way – myself, as a human factor. I wanted to drive as clean, smooth and fast as possible, not only for the purpose of this material, but also for myself. However I also had in the back of my mind that I need to give the car back in one piece (especially when I had brand-new Supra A90 to test, about which I’ll write about soon, just follow my Tribe 😉 ).
I think a difference of 0,1 seconds on a 6-kilometre track (second longest in Europe, the first one being Spa) is still very, very close. Unfortunately I didn’t have the opportunity to plug-in any kind of telemetry or data logging system, because I didn’t have one with myself and from a stock car I didn’t have enough data. If there will come a nice opportunity to test something “fast and race-prepared”, I’ll do my best to show many things from data logging perspective, and then detailed mathematics will provide a much more detailed and further explanation.
I want to give special thanks to:
● Michał Horodeński from Complex Motorsport / Toyota Team Classic and also for the whole team from Toyota Media Cup, for their support, cooperation and trust put in my person and skills,
● Kacper Nowogrodzki from AUTSIDER.pl, for the great photos which are major part in this material
● My good friend – Krzysiek Święcicki (www.facebook.com/SwiecickiMotorsport/) who saved me at the very last moment by supporting me with his video camera
● My another good friend – Konrad Oszczepalski, who supports me in the text correction and proofreading in English
I write my articles from passion, as a racing driver & instructor and especially, as a true petrolhead. From my articles I didn’t have (at least not yet) any direct income, and I don’t want even to say how much it takes to prepare the whole honest material like this one, it’s sharing and advertising (especially paid-ads from my pocket). So, if you have enjoyed this text, I will be very grateful for your feedback, likes & comments and sharing it further. Let’s show other people how you can prepare for real racing with saving a lot of money!
Yup, that's me again! :) [photo: Kacper Nowogrodzki / AUTSIDER.pl]
For other articles, comparisons and fun material you can follow my Tribe and also my Facebook profile (facebook.com/lempartracing)
Also, check my previous material about “PlayStation Generation on a Race Track” and “Mazda MX-5 – Game vs Reality comparison”!
Text correction: Adrian Drozdek
English Text Correction: Konrad Oszczepalski
Real photos: Kacper Nowogrodzki / AUTSIDER.pl
Virtual photos: Damian Lempart / Assetto Corsa
Video: Damian Lempart
Join In
Comments (2)
i think you get the reasoning for your shifting wrong. you want to rev above 7000rpm, yes sure you accelerate less because the power decreases, but when you shift at 7350-7400rpm where redline starts, but before the limiter, your revs naturally drop when engaging the next higher gear, but they don't drop to as low of a point as they would if you shifted earlier. due to that, you'll be on higher point in the powerband when engaging the next higher gear, so your overall acceleration will be higher. this goes for the vast majority of naturally aspirated cars, especially cars that are designed to be a little sportier than the average commuter. turbocharged cars might differ in this regard, because you also lose boostpressure when shifting and have to deal with other problems than you'd have to with a naturally aspirated one.
I agree with you, but it's individual for every engine and it's own characteristics. This which you have described works well with high-RPM engines, such as N/A Type-R engine (K-series) or BMW S-series engines, but not for GT86's engine....
Read more