Query: Best starting point for modifying car setups - strategies and process.

Discussion in 'Automobilista 2 - General Discussion' started by SpaceYam, Dec 26, 2024.

  1. SpaceYam

    SpaceYam Member AMS2 Club Member

    Joined:
    Jun 23, 2023
    Messages:
    75
    Likes Received:
    32
    This has always been a weak point for me because I spend ages trying to find my own limit/the limit of the car using default setups because I don't want to modify the setup as a crutch for suboptimal driving.

    I'm trying to change this perspective, and I went to get good at modifying the setups to suit my personal tastes.

    There are a few things I don't really understand how to manage in a sim, because obviously I can't feel what the car is doing like you would IRL. So, things like hitting bump stops, wheels lifting, suspension oscillating, etc., I'm not sure how to determine things like these where I've exceeded their limits.

    At the moment I'm driving the McLaren MP4/6 and training for championship race at Azure Circuit. So far, the main adjustments I've made to the setup was to completely soften front and rear anti-roll bars (it was more an experiment to see how I like it), optimise gear ratios, and bump up front and rear downforce. I also then increased slow bump and damper settings front and rear by 5 clicks each, and again, saw an improvement in driving dynamics that I'm not sure how to put into words. At the moment my best time is about 5 seconds away from the fastest F1 Classic #4 time, and I think about 3 seconds off the fastest MP4/6.

    Something that might help me, is knowing what racing teams are likely to do IRL to optimise the performance of their car around specific circuits, but obviously different cars have different characteristics and something that works for a McLaren GT3 is not going to necessarily be applicable to a V8 Supercar or an F1 car - or even an Audi GT3.

    So, tl;dr version: Does anyone have a simple process that they use to quickly find an optimal setup? And how fast on a default setup do you consider to be reasonable (in relation to say, the top TT times) before you start adjusting away from default? I have seen all the flowcharts etc., that explain what adjusting different settings do and why you might make those changes, but for me I guess it's hard to necessarily find out what is a car limitation and what is a driver limitation.

    Clear as mud, I know, but any help is appreciated :) I just need to get thinking on the right track.
     
  2. shadow82

    shadow82 Active Member

    Joined:
    Dec 4, 2024
    Messages:
    101
    Likes Received:
    39
    First step, most likely the hardest : find what is your driving style. Do you prefer stable rear ? Car that slides ? Do you like pointy front end ? Do you prefer understeer mid corner ? How are your braking skills ? Are you able to keep tire longetivity ? Lots of things to figure out before you can tailor a setup for you specifically. As per explanation that have been provided to me in the past few days on how to approach it in AMS2, default setup should be enough to get you 90% there.

    Then the most impactful settings (outside of any electronics and cooling systems) will be the diff, engine brake and brake power/biais. That what will have the biggest impact on the car in the 3 phases of cornering (brake-coast-exit). Once you get that dialed in to your liking, if the gap is still 5s off the top times, then it is first a priority to work on your driving for that car/track combo before touching anything else.

    Rest will be small increment and also tied to the track requirement itself when you get within 2-3s (numbers depend on car/track combo of course...). Hard to define any set aero balance or mechanical grip profile without having the track needs in mind really.

    Otherwise it might very well be that meta setup from TT could be sufficient as well, it is still hard for me to tell how this should be approached in this game, and better I do not suggest that anyway, so ignore that last phrase, because it is a bit taboo it seems :p.
     
    • Like Like x 2
  3. Kuku

    Kuku Flying Kiwi AMS2 Club Member

    Joined:
    Aug 30, 2020
    Messages:
    1,630
    Likes Received:
    1,013
    Once you have run default for a while and regardless of how it compares in time to fastest TT , ( because a lot of that is just driver experience , ability and good execution ) and you have a good understanding of the balance and can lap consistently within a few tenths each run , then you might like to start with some basics depending on if you feel the front is to understeering or the rear is too nervous towards oversteer, your flow charts will help you here for sure, as long as you stay disciplined and don’t try to many things at once .
    So it might be just adjusting Roll bars and brake bias , or trimming some downforce if it’s a track with a couple of long straights . Try to retain a similar aero balance if you’re comfy with the balance.

    after a while you can delve into more advanced stuff with dampers and springs, the in game info can help a little here but there’s a lot of good guides around , and even watching some videos from suspension gurus IRL about how to tune dampers can be beneficial if reading stuff doesn’t sink in .

    the same can be said for tweaks to the differential, but a lot can be learned just by trying a small increase to the values and feeling how it impacts the car .
    Using your own ghost in TT can help you tell if there’s a tangible benefit of setup changes , and after that you can then start the process of improvement to faster times , by selecting a ghost of a driver a second faster than you and seeing if you can brake and accelerate at the same points they can .

    if you simply can’t get there then you might consider trying there set up . And see if it’s immediately better for you , if it is study what’s different about it to yours .
    Not everyone uses TT to be the fastest, many people use it to develop a base set up for a upcoming race and some as part of a internal groups TT comp, and these tend to show as quite a lot of very close times , get to know the regular names and if you find one or two hip who have set the car up in a way you like you can keep an eye on them for future.

    Just picking Random setups from anywhere just because it happens to be near the top of the list isn’t a indicator of if it’s a good setup, there’s many many combos and there hasn’t been enough time for people to populate them . test yourself against it with default setup and you will soon see if they are actually a lot faster for some other reason .

    Don’t buy into the Meta set up jargon , I don’t believe it to be significant part of Ams2 . There are just cases where an absolute alien has the extra ability to handle a car that is unreasonably low downforce or extra snappy with oversteer… fast guys are just fast . But they wouldn’t generally run that way in a competitive environment on race day , particularly if that race isn’t your typical 5-10 lap balls out race .

    after you have done all this and feel really good about you pace and setup , you can then have the joy of joining a race in completely different conditions and wonder why you have no grip or your tyres are too cold or too hot etc etc etc … :D
     
    • Like Like x 1
  4. Michael3

    Michael3 Active Member

    Joined:
    Aug 18, 2023
    Messages:
    277
    Likes Received:
    73
    Well it's fairly obvious that the game is a game and the way to approach setups is the same as the way to approach driving. With driving you should do whatever inputs are necessary to achieve whatever goal you want based on nothing else other than the game in front of you.

    Specifically you probably shouldn't waste a great deal of time arguing with yourself or others about how your inputs to the game differ from another game or your concept of physical reality. Here I would suggest that if you find what you think is a bug, report it in a long thread and assume otherwise that Reiza are going to continue to work on and develop the game, including physics regardless.

    Meantime, play the game you have in front of you. If you're on the moon you can jump higher so jump higher like Neil did. Don't sit there in the moon buggy saying "The gravity is unrealistic" - and especially not if you've never been to the moon.

    For the setup it's a set of sliders or values and you set them where they need to be. Without any heed to anything else. If a slider position and a set of wheel inputs makes the car behave in a preferable way to you then that is the slider position and wheel input to use. No matter what you think another game or reality or someone else says about it.

    If you had the time and a consistent enough driver you could, in principle at least, go through every possible set up combination and find the optimum inputs and setup. The problem is you have 5 positions for 1 slider x 10 for another x...and those all multiply to number that's too big for a reasonable exhaustive brute-force search.

    The immediate problem typically is you then start asking yourself, or us, how to pare down that search space. Like a 6 year old playing chess for the first time you're staring at the pieces trying to figure which one to move and what are the consequences of moving a piece, and how you would know.

    At which point, some people who might be using to game to roleplay "racing driver" want to believe all the sliders and steering wheel inputs and pedal inputs match physical reality and they hope knowledge or experience gained or taken from their physical reality can be used to determine what inputs to make.

    The biggest 2 issues with that approach being (a) You'll soon get distracted and caught up in this dead end of whether the game is doing what it should do based on physical reality. You see that thing where Reiza do an update, videos appear saying the game is better. Some people rejoice, they play the game. A week or 2 later someone else says something about the game and physical reality, and now they fret and worry. They lose all the enjoyment the game brought them because of some perceived flaw. The way the back end of the car moves or whatever. They sit waiting for the next update which will take months, possibly even a year or more. At which point lots of videos appear saying how improved the game is and they rejoice, play it for a week or so until another video or post appears saying something negative about the handling and...well, this is a dead end. Play the game we have in front of us. Every other game has the same thing - until it isn't being updated.

    Or (b) That you'll miss potentially more optimal settings and inputs because you discount them based on what you believe, or someone tells you, should or shouldn't happen in physical reality.

    Or (c) The big temptation when you're somewhere where people are roleplaying is to keep popping up saying "Hey, that's not a real sword dude" or pulling out a mobile phone "Ride on horseback to tell the king? Why not just whatsapp him?" that's not going to go down well. They know the game isn't real but you're playing it too. The bigger pretence - that you're actually Sir Lancelot or that your buddy online is - makes no sense either.

    Of course, there is also the possibility you might have the advantage that if the game does mirror physical reality to a small or large extent that your experience or knowledge bears fruit. It gives you a place to start. But, if that were the case then you wouldn't be asking us how to do a setup.

    My suggestion would be that anyone who has no idea how to set up a car in real life would be better off not letting themselves get bogged down fretting about real life. Approach it for what it is : a game. Experiment and draw your own conclusions in the game. The real advantage the game gives us is several modes where you have the impossible : a consistent environment, the car performs the same, the track surface, the tyres etc - and it'll give you (at least in one mode) a ghost car that you can see how your car behaves differently compared with it (assuming you can drive consistently enough)

    But the thing to note is : it's complicated, or at least nuanced. Firstly you need to know what inputs to make to get a fast lap time. More specifically you need to be able to drive in a consistently fast way at the limit, lap after lap after lap. Or have someone who can that you can give setups to.

    Secondly you need to know or figure out how changes to the setup affect the wheel inputs and pedal inputs and vice versa. i.e it's a cyclic process. Your inputs determine which setup is best, but the setup changes might affect the inputs you need to make - and round and round like that.

    It's the princess and the pea. You've been given the task of testing mattresses and, if you can't feel the pea under the mattress then you can't do the job. The positive side of this is : If you can't tell then it really isn't having a big impact is it? You can just get a good night's sleep on any mattress.

    If one lap to the next you're all over the place, doing different steering and pedal inputs then you'll never be able to tell if setup changes are better or not. Nor what effect they had. Similarly if you drive smoothly enough but without really noticing the behaviour of the car as you drive then you won't see setup changes.

    I'm going to argue that your setup is going to have a big impact on how well you can both drive consistently and see and feel the car moving. VR is going to give an edge over flat screens. No doubt DD wheel, a load cell brake etc. But equally driving games existed before these things so they are not essential. They'll just make the job easier.

    Other potential ways of deciding if the setup change worked. Like if your lap time was lower are not reliable if your driving is not consistent. Undoubtedly players, myself included, have moved a few sliders in the setup screen, done a few more laps, got a faster lap and that faster lap is uploaded on the TT with the setup. Sitting there waiting for the person who thinks that setup is some key puzzle piece. But it's just another 6 year old sat across from you at the chessboard who moved a piece without really knowing why.

    As I said the other day in a post, if you can't tell the difference then it really doesn't matter. This is the main reason why people will tell you not to fret too much about setups because, for the vast majority of people playing a game there's more variability in their lap times from the way they drive than there is from setup changes. They would all spend better time doing a couple of thousand hours of driving laps than wiggling sliders around in the set up screens if their goal is to get faster.

    Smooth fast driving. Not driving under the limit consistently slowly. So you can't do 10 laps of a track in 1:31 and think "Right, I'm consistent" - you want to be consistently wringing everything you can out of the car as it is setup now, lap after lap. It's smooth fast driving at the limit that should give you information about what the car does at the limit and then allow you to see how that behavior subtly changes when you tweak the setup.

    I can't pretend I'm there.

    But Yes, tl;dr the approach to anything in a game should always be the game. It's a computer game - the faster player wiggled a wheel and pressed some pedals and you can do the same and achieve the same results.
     
    • Like Like x 1
  5. Gabriel "Pai" Legnini

    Gabriel "Pai" Legnini Well-Known Member AMS2 Club Member

    Joined:
    May 27, 2020
    Messages:
    1,049
    Likes Received:
    552
    In short: no, there is not a clear setup process to follow that is easy to teach and replicate. To compound matters, the combo you are currently tackling is both difficult and unique. Not only needs a very complete skillset on setup building to get the most out of it, but also will produce an end result that won't be of much use at anywhere else.

    The good thing is that AMS2 cars are very responsive to setup changes, and the process of doing one is both interesting and rewarding. But what complicates matters is that several default setups, simply put, suck. They have many crucial values way off from where they should be, so the car is out of whack. You can drive around the issues, but the issues are still there and are a hindrance to the driving experience. There is a big difference between having to tweak a useful baseline to suit both your particular needs and the track you are currently driving at, and having to overhaul it for having a good starting point. And quite often when it's the first time you drive a car in the game, what you need to do is the latter, not the former.

    Having said this, and not wanting to complicate matters a lot in a single post, advices for this process are similar to what others said: try to chain several laps in a row on default setup, no less than 5, and try to be as consistent as possible with your driving as possible. That will give you an idea of where you are standing with the car, what is good, what is bad. Try to ask yourself as many questions as possible regarding what the car is doing at every part of the track. Try to divide each corner in your mind in as many phases as possible. You probably won't be able to separate it in more than two or three at the beginning, but trying to differentiate between initial braking application, going to threshold braking, the initial release of it as you turn the wheel, etc., you should be able to slow things down on your mind and divide any corner in no less than 6 phases, and even more as your brain learns to process the information. This may seem excessive, but both your driving and the setup development will benefit a lot from this, they are two things that get better as more and better information is inputted to them. Getting to understand what the car is exactly doing in answer to what you do with your hands and feet is more important than knowing what to tweak on the setup screen, that comes slowly afterwards as you delve in the subject.

    Finally, just to mention something about the setup screen, there are values that are more geared to suit your hardware and how you use it, like brake pressure and steering lock. But regarding the variables that are pertinent to the virtual car itself, the #1 most important thing to set first is spring rates. If those are wrong, the car will never, ever behave or feel good. A good starting point is to follow weight distribution, and have them stiffer where the most weight is. You may be able to match the weight bias with the rates, but that's a rabbit-hole I don't recommend following, as most of times it won't work due to other variables at play like aero for example. Unfortunately, if the setup was geared to find a balance with wrong spring rates, correcting them means scrapping the whole thing and having to tweak most of the rest of the parameters to get it to work, which is what happens on several AMS2 cars.
     
    • Like Like x 1
    • Agree Agree x 1
  6. SpaceYam

    SpaceYam Member AMS2 Club Member

    Joined:
    Jun 23, 2023
    Messages:
    75
    Likes Received:
    32
    Hey all thanks heaps for the information. There's a lot in here for me to use. I do typically try to run several laps together and agree that consistency is key. I also never really considered modifying the diff and engine braking settings (I've only ever played with them once or twice) as I always figured suspension was much more important. Most of the time on a 'moderate' length track (e.g., 1:20 to 1:45 lap times) I can get within 3-5 seconds on a default setup reasonably quickly (I always tweak gear ratios to suit the track, so to me that is still part of the 'default' setup) so I think I have a fairly good baseline.

    Once I recover from night shift today I will try and get some more practice in :) Monaco is quite difficult so this practice with setups will hopefully help build up some consistency
     

Share This Page