Reader Submission #141 – Looking for Simulation in the Wrong Places

In pursuing the sole goal of perfecting race car dynamics in a semi-static environment, modern simulation developers often lose sight of the immersive sim elements that serve to complete the experience. While teams like Reiza Studios or iRacing will spend months behind the scenes refining and polishing a brand new car to be released into their respective piece of software with an upcoming patch or DLC launch, we very rarely talk about the various race weekend elements that these companies traditionally overlook.

Restricted sets of tires over the course of an entire event, limited backup cars and repair times that carry over from open practice to the race itself, as well as intricate cockpit systems & ignition sequences, are all questionably absent from our simulator experiences despite fans and developers alike lobbying for near 100% accuracy. It’s certainly a bit hypocritical that all major and minor race sim teams boast a hardcore experience, but unlike Flight Simulators, an asterisk is hidden in the fine print; this level of authenticity is confined solely to race car physics, which is only part of the complete package. There’s no debris on the track after an accident that safety trucks can be seen cleaning up, no push starts in iRacing’s World of Outlaws content, furthermore, when these elements are introduced, sim snobs turn their noses up at the title, calling it “simcade” and down-voting you into oblivion on Reddit for suggesting you’re enjoying your time with it.

Today’s Reader Submission from Leo G. believes we’re looking for simulation value in the wrong places, and that the sim community as a whole is a bit hypocritical. They want a hardcore experience, but only when it’s convenient for them, leading to developers awkwardly building pseudo-hardcore titles that technically could all classify as simcade.

Hey PRC team, your recent pre-season testing post has driven me to put into words something I’ve been thinking for a few years now – the sim racing community is full of hypocrites.

Sim racers (and the developers as well) are extremely dedicated to “realism”. We want our sims to be the most realistic experience possible, and creating that should be priority #1 at all times. The problem is that all of the details that can be measured, gathered, and recreated – simulated – don’t mean a thing if the feedback isn’t right. And playing games on the computer is very different to driving the real thing. That doesn’t mean harder or easier, but different. I’ve been sim racing for over twenty years and have been participating in club level motorsport for about the last five. Whilst my sim racing experience definitely helped prepare me for the track, I noticed an interesting development. The more time I spent behind the real in real life, the worse my sim racing became. I believe this is because the feedback that I’m most heavily relying on while in my actual car just doesn’t exist at my computer desk.

But we still chase the perfect sim, absolute realism, as if this is something that can objectively be achieved. As if we’ll reach a point one day where we actually will have a simulation that is 100% accurate and realistic… but what does it matter if the experience itself can’t be recreated?

This brings me back to your pre-season testing post. Your first drive in a new race car and… what was that? Old tires? And the track was covered in debris, dirt, and… glass? I think you’re spinning tales, because I’ve been playing the most realistic sims available for years and I’ve never had to drive on old tires on a filthy track… oh wait, that’s because our “realistic sims” are only telling half the story, and pretending like the other half doesn’t exist.

Have you heard the story of (pretty sure it was) Graham Hill driving four consecutive Grand Prix on the same set of tires? The rubber back then was so hard, the tires got faster as they wore out. Funnily enough, I did a USAC style league race at Michigan on iRacing many years ago in the Lotus 49 – a complete 250 mile event. In testing for this race, I found out that the tires got faster as the race wore on. As in, three seconds a lap faster. On a two mile oval. That’s insane.

It would have been great to be able to run the tires in practice and qualifying to wear them in for the race, just like what would have happened in real life… but that’s not what our sims are about. I found the whole situation to be slightly amusing as it turns out there were many others in the field who hadn’t done their research and kept pitting for new tires during cautions, not realizing that they were only making things worse. This sort of thing would never have happened in the real world.

Another similar story comes from the Blancpain Endurance Series, when Shane Van Gisbergen made his debut at Monza. Van Gisbergen is a multiple time race winner in V8 Supercars, winning the overall championship in 2016. He’s a bit of a gun and is fast in anything he drives. So what does a driver of his caliber get for his first Monza GT3 experience? Old tires. Shane did not get a fresh set of boots at any point throughout the weekend. Teams get five sets of tires for three drivers to share across all of practice, qualifying, and the three hour race. Yeah. If you get invited to race as a co-driver – you don’t just waltz in and own the place. You might not even be allowed to make any setup changes.

And of course, there is so much more that real racing drivers have to contend with. In the sim, we can blow an engine with two minutes left in qualifying, hit the reset button, and still take pole. We can change every single adjustable component of the car just by loading a different setup. We can destroy our cars with reckless abandon and have a sparkling shiny new one in the blink of an eye. That’s not realistic. That’s not sim. Not even close. In the real world, if you put a car into the fence during your first lap of Friday practice, there’s a real chance you could be done for the weekend. I’m not talking about going all the way to add driver fatalities because that’s absurd, I’m saying it’s insane how we have 99 backup cars and you can hit the track with a new one in five seconds when depending on the damage, resources of the team, and the series schedule, a real team might be out of action for a few weeks.

Another example, the fanfare and attention that Fernando Alonso received for passing up the Monaco Grand Prix in favor of the Indianapolis 500. In a simulator? No big deal, you can run Monaco, Indianapolis, and the Coke 600 in a span of a few hours. And let’s not forget the Nurburgring 24 – they might all be on the same weekend in real life, but this is sim racing – we don’t care for such trivial details here!

So why is no one talking about this sort of thing? If we’re a mob that pride ourselves on our commitment to pure simulation, why are we blatantly ignoring the details that we actually can pursue, features that most certainly could be implemented into our sims that would go a long way to improve their depth, immersion, and experience – things that absolutely would make them more of a true simulation than what we have now.

But no one seems to care. We claim to want the “most realistic simulation possible”, but guarantee if we went down that road, the forums would be filled with cries of “it’s too hard / I don’t like it / why should I have to do it this way”. Come on people – go to the track one day and compare what you see with what we have and ask yourself – is it really good enough? Can we do better? Even watching a race on TV, you pick up on so many details that sim racing conveniently ignores. 

I really wish we could turn the enthusiasm we have for things like “accurate recreation of the effect of different air density on the engine and aerodynamics” into this completely untapped side of racing. There is so much sitting there, waiting to be taken advantage of. Why are we ignoring it?

Hey Leo. Personally I don’t think your specific examples mentioned above are all that compelling, but the overall theme and argument of your submission, however, is.

Developers spend all this time perfecting things such as engine and aero efficiency as it relates to air density, but then the core experience of the physical race weekend itself is still very simplistic and largely unchanged from IndyCar Racing II back in 1996. You have four sessions; three provide infinite vehicle resets with 100% fresh equipment (though iRacing has an option to override this), two of them you can turn laps at your leisure, a third scores qualifying times, and in the race you have one shot at the thing unless you’re offline and presented with a restart button. Vehicle status resets at the beginning of each session, you’re given unlimited backup cars, unlimited spare parts, unlimited sets of tires, and the ability to make radical adjustments you’d never have the time to complete at the track with the click of a mouse button. It’s very… simplistic, and doesn’t really reflect the challenges real drivers are presented with.

So let’s go over what sims could reasonably implement without much trouble.

Modern simulators give you an entire Summit Racing catalog of setup adjustments at your finger tips in the garage area, some of which would be nearly impossible to perform while physically at the race track and the clock ticking. Changing tire pressures is as simple as running an air compressor, but spring and sway bar swaps physically require you to remove and then re-install the parts by hand, cutting into precious practice time – not to mention the complete implausibility of a minor league team coming to the track with an entire collection of gears, or even springs sorted neatly by 25lb increments. My change would be to implement a load-out screen prior to clicking Go or Join Session on the race configuration menu, where you’d have to pre-select the handful of adjustments you wanted to bring to the track, and part of the skill as a sim racer – like a real team – would be knowing ahead of time what parts you needed.

Limited sets of tires have been implemented in Codemasters’ F1 2016 of all games, though this game was obviously deemed simcade by the sim racing community because it’s not a no-nonsense simulator based on the ISI motor engine sporting graphics from ten years ago. I would like to see this implemented across other games, however, as it would encourage sim racers to treat practice sessions in a bit more of a professional manner, turning conservative, safe laps to break in the tires rather than balls out mock qualifying runs that routinely end in destruction for themselves and the cars around them. I’d also like to see general car degradation be cumulative across the entire race weekend, and users be given just one spare backup car – in the long run it’s much simpler than coding in an entire safety system like iRacing have done because it accomplishes the same goal of making sim racers treat each other with a bit more respect, so the developers win on that front.

At the same time, I’d like random mechanical failures to be eradicated – which although it sounds very un-sim-like and against the theme of this post, mechanical failures are down to shitty parts, and sim racing titles don’t have a meta-game inside them where you’re tasked with finding the best parts or engine supplier for your team, so it would be wrong to fuck someone out of participating in a race on what’s essentially a random number generator. But if someone blows a motor naturally by shooting the revs into oblivion, or cuts a tire down by running over debris, that’s on them.

This would also remove bullshit setups out of the equation, as people would be more inclined to create stable setups in an effort to click off controlled laps rather than struggle with hyper-loose cars on the edge of control. If you go into an online league race knowing one wreck over the next two hours across four different sessions could have drastic consequences, you’re not going to even bother testing out a forum setup with 0 wing and crazy stiff sway bars.

These are all features that are pretty easy to implement. But what happens if you go further?

I’ve found it pretty ironic how to this day, iRacing still hasn’t enabled car collisions on pit road. It’s very strange how this is a game that prides itself on simulation value and producing a highly authentic experience that can supposedly be a substitute for a real world motorsports career, but then you can go and warp through cars on pit lane. I’m not sitting here demanding motion captured high definition pit crews that you can send flying for a laugh, I’m talking about basic “let me hit other cars so I learn to respect their physical space on pit road.” Considering this existed in NASCAR Racing 2003 Season over fourteen years ago, there’s no excuse for iRacing not having this in 2017. It should really have been implemented by now.

Yet unfortunately a lot of times I hear things such as “it might be too hard for a significant portion of the userbase to grasp” – which is what also happened Kunos Simulazioni let you flick all the switches in the cockpit back when netKar Pro was their flagship simulator. Sim racers, who jerked themselves off over taking part in a “hardcore” hobby and being somehow better than “Need for Speed and Forza kiddies”, suddenly were like a fish out of water. now that they had to complete an ignition sequence to start their cars.

And this, sadly, is why Leo’s alternative simulation value elements he’s brought up today aren’t implemented, and why developers get a severe case of tunnel vision and only focus on vehicle physics first and foremost. The average sim racer simply isn’t skilled enough to be anything other than frustrated if new simulation value elements are introduced. Sim racing is a genre where users spend an hour downloading rFactor mods, turn five woefully off-pace laps in private testing, and then race to the forums to inexplicably brag how they were unable to control the car and that historic drivers must have been more skilled than modern fighter pilots, despite all historic racing footage clearly indicating a large portion of the field was a combination of drunk/stoned/horny.

The moment you crank up the difficulty level any more than it already is, participation levels are going to fall off a cliff. This obviously results in less people playing the game, and in the sim racing landscape this creates a very dire predicament, as most developers aren’t swimming in basements full of gold coins, but rather struggling to break even and forced to make careful decisions when it comes to licensing and pay models. So a lot of the immersion elements on the drawing board simply aren’t put in at the end of the day because there’s such an intricate line between breaking even and financial peril for these teams.

There’s also the insane levels of cult-like behavior in our community that can fuck over games before they’ve even got a chance to shine. DiRT Rally and Formula One 2016 were both spectacular games, but each of them were almost immediately written off by a community who largely scoffed at additional immersive elements, like having to refine your driving line throughout practice for car upgrade points or hire crew members – which was so incredibly basic I’m almost confused as to how grown men threw hissy fits at what’s basically one menu option you touch every six races. So developers with genuinely good ideas behind closed doors are almost afraid to introduce these elements, as there seems to be a whole bunch of people in our community who aggressively lash out against anything that isn’t a strictly car on track with opponents simulator, and then convince their friends to do the same.

For example, I’ve heard people from the Assetto Corsa kingdom knock iRacing for being a giant ePeen dickwaving contest due to iRacing’s ELO rankying system, when in reality the concept of both safety rating and iRacing is objectively one of the service’s best features and it’s why the game has been so successful. These old men who seemingly hate progression, fun, and anything that isn’t strictly car on track are partially what’s holding sim racing back, as developers will then fear adding new elements in fear of backlash from the vocal minority, because sim racing in itself is just a bigger collection of vocal minorities spread over five or six main message boards. Unlike Call of Duty, where one angry YouTube video is countered by 15,000 sales, sim racing devs don’t have that luxury. The vocal minorities on the forums are your customers.

Would I like sim racing developers to move past splitting hairs over car physics, and into other realms of realism? Sure. Safety trucks on the track after accidents would be cool, limited repair times and a finite box of spare parts or backup cars I think have the potential to be welcome additions, but unfortunately developers have basically been backed into a corner by the community. The average sim racer isn’t talented enough to cope with anything more hardcore than what we already have, and any genuinely creative elements are met with immense hostility from grown men for being too entertaining. So this has created a landscape like what we have now, where developers obsess over transmission behavior and doppler effect refinements rather than fleshing out the metaphorical world around these cars.

Advertisements

Reader Submission #140 – Calamity Forces HSO to Re-Schedule Indianapolis 500

An incredible package of open wheel cars warranted an equally captivating online championship, but for the Historic Sim Racing Organization, things on the competition end of the spectrum just aren’t going according to plan.

A few months back, we here at PRC introduced our readers to the stellar CART 88 mod for the original rFactor, a download bundling every single vehicle variant and driver combination that took the green flag during the 1988 American open wheel season into a light-weight download for the legendary rFactor simulation. While many were blown away by the sheer challenge of pushing these cars to the limit, and borderline-autistic attention to detail that replicated engine improvements and chassis swaps across each event, CART 88 as an rFactor mod was only part of the complete sales pitch; HSO would conduct their own full-length, full-distance online championship throughout the 2017 calendar year, allowing hardcore sim racers to step back in time and subject themselves to the same challenges their favorite drivers faced depending on their car of choice – whether that meant limping an under-powered backmarker entry to the finish line, or repeating Danny Sullivan’s dominance.

Yet after only two and a half events, HSO’s CART 88 championship is being remembered not for the intense battles, crafty race strategies, and stellar displays of sim racing prowess, but instead for ineptitude on the part of its entrants. Today’s Reader Submission from an anonymous sim racer competing in the series is here to tell us that while the mod is amazing, the series it was built for, isn’t.

Hello PRC. As I’m sure most of you know, the long process of practicing, qualifying and eventually racing the Indianapolis 500 is well underway. To coincide with this, the Historic Sim Racing organization are holding their own virtual rendition of the event. Using their exceptional, in-house developed CART 88 mod, HSO are running a full season using the mod, with the Indy 500 event being the jewel in the crown. However, this race – the most notable on the entire HSO schedule – has been absolutely plagued by issues. It has been memorable for all the wrong reasons, as the Indy 500 was cancelled (and promptly re-scheduled, to the credit of HSO) without even reaching a tenth of the proposed distance, as ignorance from competitors reigned supreme.

The race was due to start at 20:30 CET, but was delayed for over half an hour due to connection issues experienced by the head administrator. Inconvenient, but very understandable that we’d be required to wait for the primary official, because that’s who is going to guarantee us a fair race. However, this would be the most minor of problems all night, as in my opinion it descended into anarchy.

HSO’s Indy race called on a total of four manual formation laps. The first one was to be completed in rows of two, the next two run in single file, and the field assembling in the iconic rows of three for the final lap. One may ask how competitors were to know when to get underway considering rFactor can only support one formation lap, and the answer lies in hiring a manual pace car driver to do the dirty work for the three additional laps, and then typing “green” as a way to indicate a live green flag.

This is where the first start faltered. After four formation laps, the pace car driver miscounted the laps. He didn’t type green into the chat, and was still on the course. Confused, an admin further back down the field gave the go-ahead. This was disastrous, as the pace car was still on the racing surface. Despite the driver’s best efforts to get out of the way, the field was flustered and a major turn one pile up ensued, with the race restarting.

From that point on, the race was unable to restart smoothly. Several drivers who had partaken in the first start now had connection problems and could not see any cars, as their games did not sync with the server. There were several attempts at formation laps, many of which resulted in additional crashes due to some drivers not being able to see anyone. This was an unwise decision as the people having connection issues could have their problems solved by quitting the server and rejoining.

A short break was taken after the race failed to launch in a clean fashion, with the event now running an hour behind schedule.

Unfortunately, things got even worse. There were more pile-ups on the formation laps, leading to more restarts. In some instances this wasn’t the fault of HSO or any admins;  some drivers just had incompatible internet connections yet still tried to race when they should have thought twice signing up for an online league knowing their internet can’t handle it. As a result, the last restart sealed the fate of the race. A caution was called during the start procedure, adding an extra lap before the start as more sim racers embarrassingly crashed on the formation lap. The race did eventually get underway, but the group failed to complete a green flag lap before the yellow flew again. One driver, who had caused full race restarts in the past due to poor internet, put his foot to the floor way earlier than anyone else. He plowed through stationary cars, taking out no less than ten drivers while somehow managing to remain undamaged thanks to shoddy netcode. This ended up being the final straw for HSO admins, who cancelled the event shortly thereafter.

For HSO to operate properly in races of high notability and of a large entry list, they really need good restrictions on the quality of their competitors’ connections, and begrudgingly bar anyone with connection issues. If you can’t see anyone or are lagging like crazy, you should have the decency to park your car, simple. I’m not placing the blame totally on HSO here, the competitors should have the sense to pull out if their connection is like this. However, the admins should regulate how good their entrants’ connections are. I hope there are plans to do so.

This is not a hit-piece on HSO. I compete in the league on a regular basis and wholeheartedly enjoy it. Their in-house mods are the stuff of dreams and most of their races (which are on a smaller scale in terms of entrants) are meticulously organized and go down without issue. However, I would like them to learn some lessons from this farcical occasion. They are a non-profit organization at the end of the day, however their recent gigantic success in the area of mod development has catapulted them into the spotlight, and they need to be giving off a better showing than this.

I agree some effort should be made to impose internet connection requirements not just in HSO, but as many sim racing leagues and clubs as possible. Yes, back in 2005, high speed internet was a bit of a luxury, and at the time it would slightly inappropriate to demand the average sim racer to fork over a pretty penny just to continue what were essentially “gaming nights with the boys,” but times have changed. It’s been twelve years since the original rFactor came out, and cell phones can now download high resolution pornography at light speed while sim racing as a genre has grown exponentially, so there’s really no excuse for dropping loads on pricey wheels, pedals monitors, and button boxes while simultaneously lagging all over the fucking place because your ass suddenly can’t afford anything more than Wal-Mart WiFi.

Call me a major asshole, but you’ve had twelve years and a few console generations to put some money aside so you can upgrade your internet. At some point you have to stop pandering to these people.

However, I will say that the old adage of “sim racers turning five off-pace laps in a historic car before running to the forums and bragging they’re unable to drive it” is a large portion of why the CART 88 series isn’t going the way many have wanted it to. Though we rightfully gave HSO heavy coverage for a phenomenal mod, myself and Dustin ended up backing out after just two races because it was becoming very apparent that only a fraction of the grid could handle these cars for multiple laps in a row. Not only was building a setup for these beasts extremely finicky, they were very difficult to master on warm tires – we’re talking Grand Prix Legends in fast forward. By comparison, most of the guys who signed up for the league weren’t capable of driving the cars in a safe fashion; they either grew up watching the cars, or had an interest in historic online racing. That combination works well in other HSO series that use big, bulky classic GT cars with the weight and general performance characteristics of a modern family sedan, but certainly not here with these speeds.

There is a silver lining to all of this: I’ve been told that HSO will re-schedule the race and do everything in their power to force their league members to prove themselves (and their internet connection) in a different class before moving up to higher-powered cars such as 80’s CART monstrosities, it just sucks these things weren’t figured out prior to the season starting.

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.

Reader Submission #138 – From One Extreme to the Next

After the surprisingly competent Porsche Cup release earlier this year, which saw a new iteration of iRacing’s long-standing tire model project impress even the harshest critics who had been slowly scaling back their activity within the simulator, iRacing’s latest update has now brought these specific tire model revisions to a slew of other vehicles as well. Dubbed NTM 6.5, the latest iRacing rubber redux has coated the forums in a film of optimism that is long overdue for the aging service, though today’s Reader Submission indicates not everybody is completely sold on the new changes as of yet. While certain cars are indeed much better to drive than they were previously, users are pointing out that tires have in some cases been hastily copy/pasted from one car to the next, in some instances leading to amateur rFactor mod troubleshooting scenarios which place slicks with inaccurate properties on the wrong car.

Wishing to remain anonymous, partially in fear of a backlash from the community for daring to talk to PRC, today’s Reader Submission has more.

Yeah, no introduction here, I just wanted to shoot you guys a short message about iRacing’s new tires. Very soft and spongy, if I have to cut things down for length.

It appears iRacing basically took the tires from the new Porsche GT3 Cup and slapped them on the HPD. While they solved the problem of skating on ice and being able to save death slides, they are obviously just a fudge of the tires in an attempt to try and make them act more normal-ish. It seems one of the main things they did was just make the sidewalls two inches taller. This added height makes the car wobble more, and feel very spongy in the first fifteen degrees of rotation in either direction of the steering wheel. It feels like running on tires with 12 PSI, where the shoulders are just rolling over. When loading an HPD setup from last season, the only thing wrong is the ride heights all the way around – they’re too high, and it doesn’t pass the built-in tech inspection.

As you may know, anyone decent in the Acura HPD runs their front spring rages from 160 to 210 n/mm, and then whatever their fronts are, the rears are a click or two softer. What’s funny, is that the spring range for the HPD is 130 through 550 n/mm, so effectively, the only viable spring rates you can use for the car are in the first 25% of their available range. And here is what happens now if you run within that range – the tire protrudes through the fender.

So suffice to say, I believe iRacing fucked up and put GT tires on a prototype, and it feels awful and completely unrealistic with huge sidewalls that make the car feel like you’re running on a sport utility vehicle suspension. I wasn’t even hitting the track with the front splitter all the way down;, obviously at Sebring that would be detrimental, but it would be even worse if you did. I was about four clicks up on each side from the lowest ride height. Back to the drawing board, I guess.

I’m certainly not trying to turn this place into an iRacers Anonymous meeting by any means, but what you’re experiencing is why people like myself, Dustin, and some of the commenters which frequent PRC often talk about the tendency for the staff to totally wing it, only for the low-level punters to sit there and sip on the Kool-Aid as if this is a perfectly normal part of software development with this much money and research behind the project. I could certainly sit here and throw out a bunch of insults directed at that specific neck of the woods, but instead I want to bring up a very specific example:

A few days ago, Billy Strange of Inside Sim Racing uploaded a lengthy preview video of the new build, and in it he discussed the new GT1 tires. Now, I like Billy, I like the concept of the video above – introduction skit included – so this isn’t a knock on him per se, but just an example of how the community operates and sort of looks the other way when it comes to genuine issues. About halfway through the clip, Strange begins evaluating the Corvette C6R GT1, and states that the car now feels “underneath you”, and previous builds had this weird off-throttle oversteer, or something to that effect. The portion of the video you’re looking for begins at 6:20.

My question is, why is this only being discussed now? When the last build was released, why wasn’t Strange coming out and saying “yeah guys, it’s not very good, the car doesn’t feel planted, it does some funny things, I’d be on the fence about buying new content.” Those are findings a lot of people would like to be made aware of. Instead, the fallacy of iRacers praising the new update and shit-canning the last build, when three months ago they praised that update and spoke negatively of the second-to-last build, continues, regardless of how contradictory they may be.

The point I’m getting at, is you are simply ahead of the curve here. You are correct that iRacing have a tendency to wing certain updates, and confirmation will come in about three to four months, when iRacers both on the forums and in YouTube videos will begin bashing this current build of the game. This is unfortunately part of iRacing’s development process; the in-house testers simply aren’t very good at what they do, and it’s impossible for iRacing to recruit aliens to test the software because it’ll then compromise the playing field (seriously, who would report exploits only they’ve found with $10,000 on the line), so the reality is that iRacing members are for the most part paying beta testers.

And to get these people to notice changes in the first place, you’ve got to make adjustments that go from one extreme, all the way to the other. Not everyone is the sim racing equivalent of Niki Lauda, blessed with sensitive ass cheeks and championship-winning driving skills. To get proper feedback on the new tires, they’ve got to make the change from one rendition to the next so drastic, even the punters (I like that term, thanks) with 1327 iRating feel obligated to hit up the forums and say “yes, I feel there has been an improvement. in these specific areas.”

Shitty? Yes, but that’s part of the deal when you sign up for iRacing. Some builds you’re racing, other builds you’re an elaborate beta tester. Now you see why some of us have taken the money and left for greener pastures.

Reader Submission #137 – iRacing and the Formula Renault 3.5 *UPDATED*

There’s a bit of a story developing on the iRacing forums today, and it’s one that you most likely won’t see covered on any of the major sim racing outlets, as it makes the team behind iRacing look extremely incompetent and lazy; willing to throw around thousands of dollars in licensing fees without anything actually coming of them. Today’s Reader Submission comes from a user by the name of Shannon and Nim are Ruining iRacing, though despite the obviously biased name, the entry is anything but – iRacing have teased the Formula Renault 3.5 for many months as a sister car for a vehicle already on the service – the Formula Renault 2.0 single-seater – but many have wondered why this car has not materialized in the same manner as the open-wheel ride pictured above.

We’ve now found out why, and it should make any iRacing member willing to spend hundreds of dollars on the service in the hopes of receiving a bunch of new content, question where their money is being put towards. In some cases, it’s merely being squandered.

Picture the scene. You’re running one of the biggest racing simulator services on the planet, and you acquire the rights to the Formula Renault 2.0 and 3.5. Excellent, a real coup that’ll bring in European racers in droves, which is good because that’s the market you’re weakest in. Finally, a proper ladder for the European road series. So you contact teams, and in true iRacing fashion, you find that you can’t get the data required to replicate the car.

Or do you even contact them in the first place?

In recent years, it has become quite commonplace for iRacing to announce something, only to never deliver. Long Beach is the oft-cited classic example, but we also have the Honda BTCC cars, an array of British motor racing circuits, the Silverstone update, various oval re-scans, and the list goes on. The tracks, perhaps, understandably got bumped, after all the Nordschleife and Le Mans were a lot of work, but what of the cars? Why don’t we have the Honda’s, the other three Aston Martin’s announced, the other BMW’s, and the Formula Renault 3.5? Even the Ferrari is seeming somewhat uncertain at this point.

Today, we have the answer, and it comes, like most information about iRacing, in the form of a post on the members-only message board. Way back, when people first began questioning the failure of the Formula Renault 3.5 to appear, a chap by the name of Diederik Kinds offered to introduce iRacing to a team he worked with, after iRacing had said they were unable to get any of the 3.5 teams to work with them. Now, that’s quite some stroke of luck. Sadly, the thread I’m sourcing has since been deleted for very obvious reasons, but thankfully I have taken enough screenshots to ensure everyone at PRC can understand the overall premise; those waiting for the FR 3.5 have come to know Diederik as the best possible route to getting this car in the game.

So the exchange starts with Tony Gardner himself:

Although we have no shortage of car production projects, it is our intention to build the 3.5 car. We are finally making some progress thanks to member Diederik Kinds who hooked us up with a good contact with a team this week who seems willing to work us. That was what we needed. Good job Diederik!

Diederik last posted on the topic in October of 2016, and we all thought that progress was being made, cars being lined up to scan, that sort of thing. Now it seems sensible that iRacing would be quite keen to work on this, after all they’ve spent money on licenses, and a license unused is money down the drain. None of us like to throw money away, and when running a business that is especially true, considering the number of loyal iRacing customers who will buy almost anything they produce, even if they never plan on driving it. The car would obviously have sold well, as evidenced by the enormous popularity of the Formula Renault 2.0.

Fast-forward to March of 2017, in a lengthy thread where someone asked if there was any news on the Formula Renault, eventually Steve Myers (the executive vice president and executive producer) eventually chimed in:

Sorry, nothing has changed here. I don’t own a 3.5 car and I haven’t found anyone that does that is willing to work with us.

He then quickly threw a distraction in there, saying fourteen cars would receive tire improvements. Initially the thread focused on the tire model changes, but eventually people started pointing out that there was a guy who said he’d work with them, and wondered what had happened with that. One chap wondered if iRacing just weren’t interested in making the car, given that the Formula Renault series isn’t anywhere near Formula One in terms of secrecy, so there wouldn’t be an obligation for teams to hide data from iRacing and refuse to work with them. Steve Myers did not take that observation very well.

If there is one thing I really don’t appreciate it is someone implying I am lying. I think anyone that knows me would vouch for the fact that almost to a fault I tell people how it is. If you can’t look at the roster of cars that we have built and figured out that we build what we get data for than that is your problem. We will happily build the car if someone can put me in touch with a team that actually will work with us and not disappear when they find out all the data we need to build it.

Now let’s go back again to how the deal was made. iRacing obviously spent money on licenses for the FR3.5 (and the Honda’s, the BMW’s and the Aston Martin’s). In many of those cases, they claim they couldn’t get a car to work with. Let’s think about that for a moment, just imagine how easy that is to take advantage of: “Yes Mr. Myers, feel free to pay me thousands for this license, now I’ll allow you to use our car in your game” is a pretty easy thing to say during a licensing deal if you know none of the teams running your car are going to let them do it. Money for nothing, right? And yet what business would actually operate that way?

So with all that out of the way, we finally receive the bombshell which exposes exactly how iRacing works. From Diederik Kinds himself, the chap with the contacts to help get iRacing the car if they really wanted it:

Just chew on that for a moment. It exposes a staggering lack of professionalism in iRacing’s approach. It leads to some big questions. Do iRacing really want the car? Do they rub team owners up the wrong way? Is there a communication problem? One has to wonder if, when one is bankrolled by a billionaire, one starts to get a bit lackadaisical about spending money on licenses that don’t get used. Perhaps when hiring friends, one doesn’t acquire the best negotiators. Is this what happened with the Honda’s, the Aston Martin’s and the BMW’s?

All of this money, all of these resources, all of these licenses, they send one Email, and then lie to the customers about teams being unwilling to work with them. iRacing haven’t even tried to get the Formula Renault 3.5 into the service, they basically sent an introductory Email and left it at that, but yet they’re paying for the license, and they’re doing so with the money iRacers have provided them with. You’re giving them money to acquire cars they don’t even put the effort into creating in the first place.

So just think, how many other times has this happened?

What am I supposed to do, act surprised? Everyone knows iRacing is a small team full of hobbyists and left-over Papyrus staff members that have been given a literal mountain of money to run their operation with, so it’s certainly not all that much of a revelation to learn a lot of bone-headed decisions are made behind the scenes, and the company on occasion will intentionally mislead or outright lie to their customers – like telling users a server outage is due to DDoS attacks, when it’s actually just the result of iRacers constantly mashing F5.

The most we can do is just continue to make people aware of it, and hopefully sim racers will either make smarter decisions in the future and not give a company like this their money, or simply continue to ask so many questions on the official iRacing forums, the staff can’t possibly delete every last thread on the topic without looking like they’re censoring criticism.

Which, of course, we know for a fact they do.

However, all is not lost. Because of the discussion surrounding the existence of the Formula Renault 3.5, as well as some of Diederik’s exact suggestions, Steve Myers has confirmed that they will now be looking into contacting another Formula Renault 3.5 team in an effort to finish the car for the European iRacing audience.

Unfortunately, this now raises more questions than answers. There have been several pieces of content announced by iRacing, only to vanish into thin air, and it took months of customers begging for information, passive aggressive answers from Steve Myers bitching about a team “disappearing” on them, and the guy who organized it all to start asking basic questions, just for one future car to get back on track. Will this process need to be repeated every single a time an upcoming piece of content vaporizes?