How to finally get gud (or at least better)

[DoctorNuu]

Hey guys, average driver here.

Desperately need some tangible advice on what to do specifically in order to get better.
Beyond the usual 'watch Top 10', etc. advice.

Backstory:
On the wheel with TCS off. About 50,000km 'experience'.
DR B, sometimes cracking A if practicing a lot. Tending to low B if getting in unprepared.
This week I put in 2500km at Interlagos. Still unable to beat 1:32.000 in quali and 1:33:500 in race.
(Which is almost 2sec away from the top and about 1sec away from what most A drivers achieve without much practice)
The sad thing is that I had already achieved these times after about 500km of practice.
Sure, I am much more consistent now, but not getting any faster.
Another sad thing is that it takes me an hour to get back to yesterdays performance.
So it really looks like my limit.

So what should I do?
Yes, I watch TOP 10 and other replays. I am aware where I lose time. And yes, I am aware of what I need to do better:
- Trail-brake better/later
- Get on power earlier/better
- Don't unsettle the car (as much as I do)
- Handle the slow corners much better

The real problem is: I just don't know how to implement it.
I cannot understand how some guys just naturally find a faster line.
I always fall back to a kind of safe, casual (and wrong) driving style. When I try to push beyond that it mostly results in errors/spins/running wide.
Only at times am I really locked in and feeling the car. This is when I achieve the times mentioned above. In general I am about another second slower.

What can be done/practiced concretely?
How can I extend my personal limit?

 

[Pigems]

Read this book, it will definitely help.

 

[breeminator]

I've been thinking about this sort of thing quite a bit recently. I'll attempt to convey what I've been seeing in my mind.

The optimal value of each input varies over time. To keep it simple, imagine there is just a single input you have to control, and suppose the optimal values over time are as follows:
5
8
9

Now, you can already see that if you had to hit those values with a throttle pedal, for 1 second each, it would not be easy to accurately do so.

If the value you actually hit is too low, you'll drive safely but slowly. If the value you actually hit is too high, you'll go off the track or spin etc.

So it sounds like you settle into e.g.
3.5 (1.5 too low)
7.5 (0.5 too low)
7.5 (1.5 too low)

i.e. you have some variation, but you set your overall level to a safe place where it's always below the optimal value. If you shift that level upwards, you end up with e.g.
4.5 (0.5 too low)
8.5 (0.5 too high - run wide etc)
8.5 (0.5 too low)

whereas a really top driver might manage something like
4.95 (0.05 too low)
7.98 (0.02 too low)
8.92 (0.08 too low)

All you can really do is:
1. Work on understanding what constitutes optimal inputs, e.g. the book Pigems recommended may well help.
2. Work on finding ways to get as close as you can to consistently delivering those inputs
It seems clear that not everyone is equally good at these things. For example, if you go to any swimming pool, you'll see large numbers of people who are not just less physically fit than elite swimmers, but their movements are massively less elegant. Or when I watch a band at a local pub, I'm always struck by how the guitarists appear to need to put huge effort in to play to a much lower standard than e.g. Steve Vai can while making it appear effortless. I read about some research where they found that some people simply have lower electrical resistance in their brains than other people - signals flow more freely and they can achieve any given level of mental performance with less effort than people with higher electrical resistance, and at full effort their mental performance exceeds that of people with higher resistance. All the same, most people will improve at any given activity with practice, and often they can continue to improve over very long periods of time, hence the "10,000 hours" idea, though that idea, if interpreted in a literal sense, has been pretty robustly debunked, in that there are world class performers who got there with a lot less than 10,000 hours of deliberate practice, and also people who have put in 10,000 hours of deliberate practice without becoming world class.

Also, even activities that might appear to require similar skills can be quite different. I knew someone who was an outstanding swimmer, almost made it onto the Olympic team, but if you watched him play football you'd think him totally lacking in physical coordination. So to some extent you can try to find something you're naturally good at, and might find that more rewarding than investing huge time into something you're less talented at.

 

[DoctorNuu]

Books: Never thought of that, somehow. It's 2020 ;-) but probably more/better information than any shallow video.

Input Value Theory: That's probably how everyone gets consistent. But as I described, I have the suspicion that there is this personal limit (of e.g. 1:32.000 for me) which I can consistently achieve (after practicing 2500km) if I concentrate, but I cannot improve beyond that. Seems like I don't even know the 'optimal values'. Sure, I can watch the replay, but replicating is so hard.

10.000h rule: Well known. But heck, my mileage only amounts to about 500 hours. Felt way more than that. The 'fear' I have is however that after I put in 10k hours, I could still be 'average'. So much time is wasted 'optimizing the values' for a given car+track.

Do you think that top drivers only do that?
I think, for one, they are probably way better and quicker at optimizing. But why and how?
But there must more. Somehow, they must have a better mental model than the average Joe. Intuitively knowing something that I don't. The whole thing is a lot about knowing in advance what the car will do. When you have to react it is too late, even more so in a sim.

 

[Groundfish]

Listen to @Pigems.
I know for me, that book transformed me from c to top 1 percent.
You could lap forever but not improve if you are not able to understand why the car isn’t doing what you want, and what the inputs are to make it faster.
Really can’t say much more, theres no ‘tip’ to make you fast. It’s a lot of factors that need to come together and it’s not an overnight thing.
Read it. Even if you read just a little before bed or something.
Re psychological side, no amount of zen is any good, no amount of hours of gts is substitute for simply understanding exactly what it is you are trying to do.
Only AFTER the UNDERSTANDING part is present can one begin to APPLY that understanding.,.
Much LATER down the road after APPLYING that sound fundamental knowledge does the whole situation shift and become more mental than physical.
Then you can get in the zone.
Thats imo the same situation in learning any skill, really.
Here is a researches measurement techniques
and results.

The precursor to it tho imo is fundamentals, like everything.

 

[The_Tullster]

When you watch the top replays, make sure you have the full HUD displayed and switch to the HUD rather than the play/pause controls. Watch the level and timing of the driver inputs. You'll be surprised by:

• How rarely brakes are fully applied. If ABS has kicked in you're going to need more braking distance
• How often drivers are coasting through turns, no input (or almost none) on brake or throttle
• On the more skittish cars, how often BOTH brake and throttle are partially applied.
• Steering inputs should be applied smoothly, not jerked immediately to the required input.
• In high speed turns you may be surprised by how rarely full lock steering is used. If you've hit the steering stops the tyres are probably scrubbing which is wearing them out and burning off speed (and fuel if fuel consumption is on).
• Pay attention to what revs the driver shifts at. Lots of cars benefit from short shifting, some benefit from revving way beyond what seems to be the limit. Set your own HUD so you can see the tachometer, not just the rev bar.

 

[breeminator]

Input Value Theory: That's probably how everyone gets consistent. But as I described, I have the suspicion that there is this personal limit (of e.g. 1:32.000 for me) which I can consistently achieve (after practicing 2500km) if I concentrate, but I cannot improve beyond that. Seems like I don't even know the 'optimal values'. Sure, I can watch the replay, but replicating is so hard.

What I'm saying is it's the randomness that hurts your times, because to not go over the limit, the more randomness you have, the safer you have to set the average. Imagine a top golfer hitting 50 golf balls onto the green from 100 metres away, and then a poor golfer doing the same thing, the poor golfer will have a much bigger random spread of distances from the hole that the ball ends up. Now imagine that the hole is right at the back of the green, with a river at the back, i.e. you need to reach the hole or come up short, you mustn't hit it too long. Because the top golfer has a smaller random spread, he doesn't need to pull the centre back as far as the poor golfer does. So he can pull the centre of his spread of 50 golf balls back 3m, say, and never go in the water. The poor golfer needs to pull the centre of his spread back 15m, say, to avoid ever going in the water. So it's like that with driving games - the more random variation in your inputs, the safer you have to drive overall to avoid going over the limit.

Have you ever watched recordings of your driving? I find my driving looks much worse to me when I watch a recording, I can see where I'm losing time so easily, whereas while I'm driving my attention is too consumed by the act of driving. For example, different game, but look at 0:26 here:


The van over-rotates a bit to the right. I correct by steering left, but over-correct, and lose quite a bit of speed because I'm then heading too far left. I know I do that quite a lot in such situations, but knowing what I'm doing wrong doesn't let me do it any better. That's simply the speed my brain works at to complete the cycle of detecting the problem, correcting, seeing that the correction has worked, and stopping correcting. A better driver would recognise the problem earlier, and detect that their smaller correction has worked earlier, disrupting the vehicle less, and carrying more speed. Or they might steer more accurately to start with, not exceed the limit, and not over-rotate in the first place. I've got better at driving games with more practice, but I can't see that I'll ever match the best players because most of them probably fairly effortlessly do that sort of thing better with little practice.

Another element of what you talk about is memory. I struggle to hold all the details I learn about driving a car and track combo in my brain. I do better at shorter tracks for this reason, it's easier for me to remember the exact details of how to drive every bit of the track. But do better drivers have better memories, or do they not need to remember it in as much detail because they can just judge it all better without having to memorise lots of tiny details for every corner? I don't know the answer.

 

[sturk0167]

But there must more. Somehow, they must have a better mental model than the average Joe. Intuitively knowing something that I don't. The whole thing is a lot about knowing in advance what the car will do. When you have to react it is too late, even more so in a sim.

Yup. One thing that the really fast guys have is a sense of where their car's gonna be 100-200 meters ahead. For example, when transitioning back to the throttle in turn, they know just how much power the car can take without pushing too wide on the exit. They don't usually guess and hope that they make it. They won't go off the track much, because they know well ahead of time if they're not gonna make it, and make the adjustment.
A lot of people who struggle, don't seem to detect that they're under-rotating until it's too late.

 

[Rosty_14]

I have been seeing a lot of people suggesting this book, heck some people read it for the simple joy of understanding better what they are seeing on F1 and any other type of racing events because of how the book is structured (read several reviews already).

I just purchased the book . It will definitely help me, I don’t have the right foundations or theory so it should be of good help. Thanks!

 

[Granadier]

Z28 Gaming just made a video recently about getting better and it's filled with really good points. Highly recommend watching it.



Most important bit is his first point about learning other car's braking performance to know when they have to start braking into a corner. This plays into the overall ability to predict your opponent's moves which imo is what differentiates between "Fast in quali" and "Successful in race".

 

[O604]

All the above. And more practice.

Practice makes perfect. It improves many areas. But one that I think is underrated is recovery. The more sim experience a driver has the better they get at recovery within the game rules. The best players can turn a spin out crash into a minimal time loss.

I currently watch the Legends Race every Sat. All the drivers are obviously skilled. The difference has been sim experience. You can see the difference in those who practice and those just joining in. There's a lot of wasted time doing donuts while 10 slower drivers pass by.

Fittipaldi has been fun to watch. He started sim racing with this series, learning from his son and is improving every race.

 

[DoctorNuu]

Interim conclusion:
- Went through the wormhole (see video above). Found nothing.
- I know the video and others and the Tullster tips above. (and probably would be further off without them)
- I am reading "the book" despite being a know-it-all-already.
- The book specifically says that bad practice (without clear intent) is just a waste of time. This is exactly what I feel the last 2000km.
- I will watch more of my own replays. Good point. Last time I checked, I was actually almost not trail-braking despite planning and thinking to do so.

To put in some specific aspects of the question:

1. How exactly do I approach 'imitating' a Top10 time that is 2sec too fast for me?
2. How do I actually "deliberately" practice trail-braking?
3. How do I handle very slow corners? (which are a weak spot and are where most time is lost/gained)
4. How do I handle steering vs throttle on accelerating out? (If your answer includes a rope, you're toast!)
5. How do I quickly find a good BB setting without driving loads of laps with different settings? Always seem to be lost here. (Quali only, as race obv. depends on tire wear. )
   
6. Are there really cases where one does not brake as hard as possible or fades in the brake? (not counting turns where you just tap the brake obviously)
   
7. Is there a place for (slight) pendulum turns, where you steer a tick to the outside to introduce the turn? How do I combine that with braking? (Think I saw this in turn 4 and 6 at Interlagos)
   
8. Despite being a bad practice in general, in which cases is over-scrubbing the front a good thing? (Think I saw Top guys do it)
   
9. How do I get my turn-in consistent for a simple mid-speed corner? Is there a certain preferred pattern how one should rotate the wheel? (I'm always experimenting here)
   
10. Which other patterns/ best practices might I not even be aware of?

 

[RikkiGT-R]

@DoctorNuu - I hear ya. I'm the same really, if a total amateur watched me driving a track I know inside out (The Ring, for example), they'd think I was a genius pro racing driver. My lines look good, the car is totally under control, no mistakes, gear changes are all smooth and so on and so forth.

But when I watch someone genuinely fast driving the course, it beggars belief really. They just seem so much faster, and tbh I don't understand it at all. They take corners at speeds at which I absolutely would crash trying to do the same, it makes no sense to me how they can go much faster around the same corners when I feel like I'm going as fast as the car possibly can

I'm a good driver, but I'm never gonna be up there with the fast drivers.

 

[RacingGrandpa]

Are there really cases where one does not brake as hard as possible or fades in the brake? (not counting turns where you just tap the brake obviously)

Almost every turn actually. I used to go full but part of learning is that you get more brake power and better rotation by not going full on the brakes.
When you start doing it it just feels so wrong, but when you are entering the turn too fast and realize that letting the brake up will be better.
Also, just go round and round on a track you know well and deliberately do trail braking.

 

[Matrix243123]

Probably a pretty basic tip but I find getting familiar with exhaust sounds help reaching the limit. Almost like eventually matching the exhaust sound with each corner. That and when possible, trying to use the same cars over and over to get use to the exhaust levels.

 

[Groundfish]

Interim conclusion:
- Went through the wormhole (see video above). Found nothing.

The mental piece is the final stage when it comes to peak performance in a skill you have built, over the course of time.
You learn you practice, you practice more you learn, and eventually half the battle is learning to simply let go, enter the zone, and flow...Sorry for including that more advanced idea, but thanks for crapping on it.

 

[VBR]

I cannot understand how some guys just naturally find a faster line.

I think you've answered your own question there. Well, that's my opinion anyway: some people are just naturally better at racing than others. I've always been a second or two off the aliens, but it's never bothered me, as long as I can find people on my own pace to race I'm happy. I bought Going Faster: Mastering The Art of Race Driving for the GT Academy 2010 Time Trial. I learned all the theory, I studied all the top drivers times, & I improved a little but still ended up a second off the aliens lap times...

 

[breeminator]

Again, I don't know the answer to this, but I'm going to throw it in here, as it's relevant to the OP's question.

Why are some people better at some driving games than others? For example, the winner of the world's fastest gamer competition is a fair bit off the pace of the fastest GT Sport players.

The fact that such differences exist mean that whatever makes someone the best at one game, it can't be the same thing that makes someone else the best at a different game.

 

[kilesa4568]

How can I extend my personal limit?

By constantly pushing on it. You have to make it a familiar territory so that you know how to deal with it. Alpha_Cypher said that to me, I think...



Check out the throttle and brake control along with the finger tip steering to see how he avoids scrubbing speed. Finessing the car around the track to keep it happy.

 

[breeminator]

Another thing to add to this - some of the current top players will have been playing driving games like GT from a very young age, and the vast majority of people will never develop the same ability if they started later in life. Again, comparing with swimming, where we have had many decades of children taking swimming seriously from a very young age, it's extremely rare to see someone in an Olympic final if they didn't become an excellent swimmer in childhood.

 

[JEverettL]

@DoctorNuu - I hear ya. I'm the same really, if a total amateur watched me driving a track I know inside out (The Ring, for example), they'd think I was a genius pro racing driver. My lines look good, the car is totally under control, no mistakes, gear changes are all smooth and so on and so forth.

But when I watch someone genuinely fast driving the course, it beggars belief really. They just seem so much faster, and tbh I don't understand it at all. They take corners at speeds at which I absolutely would crash trying to do the same, it makes no sense to me how they can go much faster around the same corners when I feel like I'm going as fast as the car possibly can

I'm a good driver, but I'm never gonna be up there with the fast drivers.

This is me 100%! I’ll watch my replays and everything is smooth and solid. I’ll then watch a lap guide from Mistah_MCA for example and it looks like I’m doing everything flawlessly, hitting the same braking points and nailing apexes, but compare lap times and I’m like 5 seconds off. SMH.

 

[breeminator]

This is me 100%! I’ll watch my replays and everything is smooth and solid. I’ll then watch a lap guide from Mistah_MCA for example and it looks like I’m doing everything flawlessly, hitting the same braking points and nailing apexes, but compare lap times and I’m like 5 seconds off. SMH.

Pretty much the only way that can happen is if you're driving the same lines without being on the limit of grip, i.e. your speed at the same point on the track is lower. If you're braking in the same place, then you're braking for longer than he is, and taking the corner at a lower speed. If you were to run the videos side by side and step through frame by frame, it only takes a single extra frame in the video of braking to ruin your lap time. This is the one thing that is quite hard to see as an error when watching other people's driving, because you can't feel what they're feeling through the wheel to know if they're on the grip limit or not. The sound provides some guide, you should be hearing mild screeching from the tyres rather than no screeching or loud screeching.

 

[Groundfish]

I think we need to define gud.
Because if ‘gud’ means equalling top ten qual times I’d say like good luck unless you invest a ridiculous amount of time.
Those laps aren’t racepace.
Also it’s quite ironic to me that imo for gt cars the absolute 2 best things to do...no one does...TCS default and cones and driving line on.
Before you laugh let me just say that IMO that’s because it’s gonna teach you to hit apexes and it’s gonna teach you throttle control on exit. You can do very well if you think a bit and get it to where it’s on default but not cutting power, in fact it’s helping you find the earliest time and maximum throttle you can put in. It’s also gonna help getting you steering correctly...
I just think it’s horrid everyone immediately turns off these tools. It’s to me stupid not to take advantage of the tools the people who built the game gave you.
So what is ‘gud?’ I’d say start there, then I can figure out if I meet the standards of gud we want to discuss.
Re the laps looking the same but slower by 4-5. If your car feels like it’s on rails and it’s not balanced into a state of neutral steer or slightly over under your times will be much slower. That’s discussed in the book also and again holds true in this game.
Also everyone thinks the latest possible by a meter braking results in a lot of time gained but other things are way more important.
But first what is gud? You gotta start there...
For me if that means getting top ten time, I think that’s silly. Those are one off laps almost always. Re starting late I started myself at late 40’s in driving games when this game came out, or rather a few months from release... I’m just not sure I meet the definition of gud as used here.

Define ‘gud’

 

[Pigems]

I think we need to define gud.
Because if ‘gud’ means equalling top ten qual times I’d say like good luck unless you invest a ridiculous amount of time.
Those laps aren’t racepace.
So what is ‘gud?’ I’d say start there, then I can figure out if I meet the standards of gud we want to discuss.
Re the laps looking the same but slower by 4-5. If your car feels like it’s on rails and it’s not balanced into a state of neutral steer or slightly over under your times will be much slower. That’s discussed in the book also and again holds true in this game.
Also everyone thinks the latest possible by a meter braking results in a lot of time gained but other things are way more important.
But first what is gud? You gotta start there...
For me if that means getting top ten time, I think that’s silly. Those are one off laps almost always. Re starting late I started myself at late 40’s in driving games when this game came out, or rather a few months from release... I’m just not sure I meet the definition of gud as used here.

Define ‘gud’

Just as an example of time invested, I put in over 2300 hours in 2019 alone over 3 accounts, and over 200,000km since I got the game in Oct of 2018, and I’m still well off true Alien pace.

 

[Groundfish]

Just as an example of time invested, I put in over 2300 hours in 2019 alone over 3 accounts, and over 200,000km since I got the game in Oct of 2018, and I’m still well off true Alien pace.

Right my most used accounts only a bit over I think 45 days of driving time too.
You’re faster than me by a bit, but maybe neither of us are gud lmao.

 

[breeminator]

I think we need to define gud.

I think he means considerably better than our level. I can understand his thinking - I also have often thought the aliens have worked something out that the rest of us haven't. But whatever that information is, if it exists it never gets posted in threads like this.

I did once see a YouTube video where the person reckoned you have no chance of competing at the top level in any driving game without inside knowledge from the devs about how the physics work. They reckoned all top players in all games have got this knowledge either directly or indirectly from a dev. But they were talking about the last tenth or two, the difference between being 1st or 5th, say, in top split, not the difference between being 1.7 seconds off the pace and 0.5 seconds off the pace.

 

[Groundfish]

I think he means considerably better than our level

Well, some combos I can beat guys in the top ten and have. I’ve been top ten...I’m not sure I have ever raced you...
But right if gud only means top .000000001 percent, then this thread is mental masturbation.
@Pigems has beaten guys who went to world tours. Is he gud?

 

[breeminator]

Well, some combos I can beat guys in the top ten and have. I’ve been top ten...
But right if gud only means top .000000001 percent, then this thread is mental masturbation.
@Pigems has beaten guys who went to world tours. Is he gud?

Pigems is normally only a little faster than me for daily race qualifying times.

The OP mentioned 1:32.0 for Interlagos daily race qualifying. EMEA #1 is 1:30.3. So I assumed he wasn't looking to get from 1.7 to 1.6 seconds off the pace, but to close a much bigger chunk of the gap than that. Bear in mind he sees his current top 2% standard as "average", so I assumed "gud" must be quite a bit faster than "average".

I understand where the mindset comes from. If you're in Kie's live stream chats, then if you're not above 70k, you're seen as someone who must be utterly clueless, and they just can't comprehend how anyone can drive so slowly .

 

[Groundfish]

Pigems is normally only a little faster than me for daily race qualifying times.

The OP mentioned 1:32.0 for Interlagos daily race qualifying. EMEA #1 is 1:30.3. So I assumed he wasn't looking to get from 1.7 to 1.6 seconds off the pace, but to close a much bigger chunk of the gap than that. Bear in mind he sees his current top 2% standard as "average", so I assumed "gud" must be quite a bit faster than "average".

He said his best qual was 32 with most laps a second slower.
I’m here to tell ya that’s diff from what I can do by a lot.
If I got anal and wanted to waste time to get optimal from my 31.4 I’d be at I think 30.5.
I ran a few for fun late yesterday and did 32 flat in the fricken db9 at default tcs.
That wasn’t try hard mode.
So is that ‘gud’
If you define gud by qual time that’s dumb.
To me that’s whoever is motivated enough to max it out by repeated efforts and putting time in.
So ok I’m not gud, let the mental masturbation turn to a giant collective circle!

 

[ASH32]

Videos are the best way to get better if you are seconds off the pace, by that I mean post a video of your best lap and people who are about .5 to a second off the top times will tell you what you are doing wrong, this should be more in reach than trying to chase those alien laps where they do things that most will never be able to. Of course you still have to do it.

If I had a better wheel, more secure stand, better pedals, comfy racing seat, newer TV, PS4 pro, etc, so you know the best possible equipment (whatever that is), it would probably only find me a few tenths at the very most so often I'd still be at least half a second slower than the top times, they just drive in a way that us mortals just won't be able to and if I were to try to emulate them I know I'd just be slower than I was before driving my way.

So in short don't chase the impossible, chase the people around my level, I fully believe it's possible anyone can get to this level, afterwards good luck.