Performance and Magic Data

The log should show "Using performance file ...". You can see this if you press '2', though you may have to scroll back with the numeric keypad.

René Smit, Independent Software Developer.
Download my GPx tools here.

OK guys. I'm conscious that I'm clogging up this thread a bit and I don't want the fantastic info everyone has posted earlier to get lost so I'm going to make two longish posts in succession which will attempt to finally answer mralpha's original question and conclude my tutorial for now.

Maybe my contributions might form the basis of an introductory how to guide in another thread or a sticky at some point - and perhaps I should have thought of that earlier - but to be frank I didn't expect my musings to be so well received as they have been.

That said, apologies to the OP and anyone else who might be a bit miffed with me right now for going over what is (to you), old and well-trodden ground. It wasn't my intention at the outset.

Well, following a bit of sleep I checked again and I DID save part 3, at least an early version of it. But I saved it in a weird place which is why I got confused. Apologies to my cat, especially; although the bandages do make him look kind of heroic so it's not all bad.

OK, let's get focussed. We are about to enter the Zone of Hard Toil. But speaking of cats, you've probably heard that there is more than one way to skin one. Well, what I've written below (and above) is just my way of doing things OK? Some won't like it at all and others will notice unintentional errors, but it's what I have ended up doing in the 2+ years that I've been messing around with GP4. Certainly, I'm not an expert, nor do I claim to be, so anything I say here which clashes with anything in anyone else's thread or post, especially this one, should be disregarded as the mindless ramblings of a fool. In addition I can take no responsibility if listening to my advice causes you to destroy your PC and burn your house down.

Part 3: The MD file.

Each circuit can have it's own unique MD file, but it doesn't have to have one. In fact, when you first download them, quite a few don't (including the 17 originals IIRC). Nevertheless, as a relentless tweaker I consider them an essential component of a circuit and sooner or later I create one for every track I use.

Because each track has different characteristics there can be no such thing as an MD file that can be used for all circuits. However, I have occasionally borrowed an MD file from another circuit to save time when the one included didn't feel right and I wasn't entirely sure what to do to make it better. There's a particular street circuit, I can't remember which one, which is kind of sluggish and annoying to drive with the MD supplied and I recall swapping it out with one from a different street circuit to great effect. Be aware though, that this is not a good strategy to adopt as a first recourse because you can seriously mess up things like pit lane entry and/or many other things. Copy and pasting small sections across from one to the other is usually safer.

Here's the info you need to get started with MDs.

For the purposes of this part of the tutorial, it'll make things easier if you start working with one of the original 17 circuits with a nice long straight, such as Hockenheim, Spa or Monza. Then, once you get the hang of what I'm about to run through with you, you'll find it easier to make any other circuit perform the way you want it too.

I'm guessing you know how to extract or create an MD file for a circuit. If not, there might be a tutorial somewhere or you could just fire up cmagic4_102.exe and give it a whirl. It's intuitive and easy to use. I tend to name the MD files along the lines of 1989Montreal.md3 etc, because I've modified a set of other peoples' mods to allow for a modest 2% increase in performance year on year, covering the years 1978-1993*, which is when my interest in F1 was keenest, ie before Senna was taken from us.

(Yeah, I'm so old I used to have to load up my PC with a bucket of coal each morning, back when the climate change scammers were bashing on about the impending Ice Age and the earliest arcade racing games moved the pixelated scenery around the stationary car! How times have changed...)

*No, in case you were wondering, currently there are no mods for 1978 or '81 as far as I know. Mine are homebrew kludges using amateurishly re-edited cars from '75, '79, '80 and '82. Hopelessly inaccurate and absolute heresy for some no doubt, but not for me, as it was important to have at least one mod containing drivers like James Hunt and Ronnie Petersen, both long gone, who got me into F1 in the first place. It's also easier to learn new tracks in a mod where the cars are less powerful. The old reaction times ain't what they used to be.

But to get back on point. In other words, when I first loaded up the 1990 mod, I could get an MD that was pretty close to what I wanted by using the 1989 MD file for that circuit which I made the previous season. (If that's unclear you'll see what I mean as we go on).

Oh, and here's another aside. (I'm tired and I ramble much more when I'm sleep deprived). For what it's worth, I don't actually use the real-world circuits. Instead, I re-arrange each mod year to include different circuits to the year before and the year after and bring in many (but by no means all) of the fantastic aftermarket circuits we have available too. I still use the real-world results to generate historically accurate PFs for each race each year, but I swap out tracks I don't care much for, such as Monaco, and replace them with a similar circuit that I do enjoy, like Detroit for example and so over the 16 year mod span I race some circuits up to five times, while others, like Monaco again, only once. I also fix each season to a strict 15 races too because it makes managing my ever-developing PF Calculator spreadsheet easier to manage. But that's just me. I'm not a purist, which can be annoying to some people...

Anyway, the MD lines we are interested in right now are:

desc1 through to desc33, (These are the wing settings for the CC cars, your car and the same again for wet conditions).

desc53 through to desc64, (These don't apply to your car, only the CC cars).

and desc42 (This affects CC cars and the player car).

Let's start from the bottom and work upwards.

Desc 42 - Master/track grip

This is something like a master grip setting for the circuit it is applied to. Changes here affect all cars, including yours. Most of the time this variable won't need changing (especially with the original 17 circuits supplied with the game), but you'll probably find that some of the aftermarket circuits are just a bit too slippery, however much you play with your wings and suspension set up etc, so you might want to increase the value of desc 42 by a hundred or so and keep repeating that until you find a value that feels right.

Conversely, there are also one or two tracks which have too much grip by default (just my opinion). It's fun to drive like this, but if you want to keep it real you'll want to make the circuit just a little tougher to drive by lowering the value of desc 42.

I think of this as balancing the circuit in relation to the others, but as I say, most of the time it'll be just fine as it is. Generally, most circuits will have a desc42 value of between 15000 and 17500. There are one or two exceptions to this, such as the excellent Phoenix circuit that has a default value of 20000, which seems to be necessary for that circuit mainly because of a slightly inefficient CC-line on one corner, ie set the desc=42 value any lower and you'll hit one particular wall too often. Ideally the solution would be to modify the DAT file to adjust the CC line but I'm totally out of my depth when it comes to things like that and always make a mess. (All hail the track builders. We are so fortunate to have had so many bless us with their talents).

Remember that in real life, some circuits have a reputation for offering more grip than others. Some are baby skin smooth with shiny new tarmac while others are bumpy, oily, covered in marbles and worn out, so it's not a question of making a circuit behave the same in terms of grip, it's more about making a particular circuit feel right in relation to the others. I don't make a habit of messing with this value, but from time to time I notice a track is a bit of an odd man out and I bring it into line.

Anyway, that's what you use to resolve all that: desc 42.

Getting your car up to speed.

We're getting there mralpha, just like I promised! But before we dive into the MD some more, we need to backtrack a little. I've never seen too many hints on this entire process, so this is all completely homebrew, but here's what I do when it comes to regulating the speed and my car and the CC cars. It's a bit long-winded, but I can usually refine a newly-acquired track's MD to suit my taste within about 20mins - and by and large, once it's done, it's done for good (with one or two caveats, explained below). Remember however that many of the tracks have great MDs to start with, so it's not like you'll have to do them all!

Now if I were to do this (see my closing comments), first, I work on the power values relating to my car only (which were determined back in the PF). Don't worry about the CC cars just yet.

Grab yourself one of the better cars (that's important) and go out on a fast track (like Hockenheim) with minimal wing (ie 2 or 3). Note down your top speed on the longest straight and make sure the car is not hitting the rev limiter in top gear. There are several ways to check this and most mods have some kind of rev counter built in, but if not the easiest way is simply to listen to the engine note. If it is topping out, go into car setup in the game and stretch your gear ratios until you can hit your fastest speed in top gear and still have a few revs to spare (to allow for slipstreaming). I say stretch because the bottom gear (ie 1st), is probably OK where it is (we'll see later on why this is). As a guide, each increase of 1 in your top gear value represents a fall in revs of around 400rpm at top speed.

At this point, if you want to make your top speed higher or lower (as you requested Mr A), you can go back into the performance file and adjust the power values in there like we discussed earlier (or you can use PRBlanco's Perf. Calculator spreadsheet to do this quicker if you have it), remembering to re-check your gear ratios on the track after you've increased the power in the PF to make sure you're not up against the rev-limiter again.

OK. Happy with that top speed of yours now? Then let's get back into our MD and bring all this to fruition. Part 3.2 follows shortly.



Edited 2 time(s). Last edit at 12/19/2018 08:28PM by Noog.

Desc 53 - The power factor (CC cars only)

Go back out on to the circuit and watch the CC cars. How is the top speed of the better CC cars relative to yours? If it's too high, or low, it's time to get to grips with desc53.

I tend to adjust each skill level, ie ace, pro etc at the same time, ie descs 53,55,57,59 & 61, but there's no good reason I can think of for doing this, and if you only ever play on the ace level (which most people do, I think), you only really need to adjust the ace power value, ie desc 53, and you can ignore the others. As a general guide, if the CC cars are 20kph faster than you try reducing the desc53 value by 10 or so. If they are 10kph slower than you, increase the value by 5 or so, remembering to re-upload the modified MD file into the circuit each time you adjust it.

It seems to be a lot of work at first (and it is) but you'll get a good feel for it after a while and it's rare I have to adjust this value more than once or twice. Better still, I find that when I load another mod, the desc53 setting is usually correct there also, ie I do a quick spin around the track, checking my speed against the CC cars and it's usually spot on.

Once you've got your top speeds calibrated, compare the CC behavior in corners to your car and use the same technique to refine the CC grip setting (desc54). Now I could be wrong, but for some reason, I always seem to have to re-adjust this each time I load a new mod. I can't really explain why I only seem to have to do this with desc 54 and not desc 53, but there it is.

Now, there are some circuits where the CC-line leaves something to be desired. For example the CC cars will slow too much at a particular point on the track or they will race ahead at another point on the track (relative to your car). Unless you can edit CC-lines you're kind of stuck with that. It would seem that even the best of our track builders have sometimes had to live with tiny little CC-errors in their offerings. In fact, there are even one or two instances among the original 17 circuits where the CC-line isn't quite perfect. In those circumstances you have to take a broader view. You won't fix it with the MD file, so instead, concentrate on synchronizing your whole lap times with the CC cars, rather than analysing each corner etc. As an example of what I'm talking about I always do badly in the second sector of Interlagos, relative to the CC-cars, no matter what I do, so I offset it by lowering the desc=53 value by 1, giving me a slight advantage on sectors 1 & 3 where the straighter sections are.

Conversely, the last corner on the '67 Zandvoort always favors me by a several tenths of a second over the CC cars, so I raise desc=53 by a point or two to allow the CC cars to gain it back on the long straights that the circuit has. It's all very much an art rather than a science but for me it's the little quirks like this that make this game so extraordinarily deep and engaging.

A quick aside regarding 1st gear. GP4 has a bug where if the CC cars have a 1st gear ratio of under 22 or so, the CC cars can get stuck in the pitlane (during qualifying and practice sessions), ie the pitcrew push and push but they can't get the car back into the pit garage. The absolute value seems to vary from track to track, but if ever you see this behavior among the CC cars, try raising their first gear ratio by a notch until it goes away. There was a post on the forum someone made which brought this to light and it's well worth knowing once you start playing around with the MD files.

Descs 1-33 - Wings and gears.

OK, now you know your top speed and the gear ratios required to achieve it. So it's time to tell the CC cars to use the same values by altering descs 1-16 - and you can also make your car use them automatically at the start of the session by inserting the values into descs 17-33. Once done, save the MD file.

In fact, I usually add 1 to the top gear of the CC cars relative to mine. This is because of an under-the-surface randomness in GP4 that might make the best CC cars top out from time to time even though yours doesn't (which again we'll touch on later).

Also, I don't bother too much with wet session settings, other than I use my dry settings add one or two to the wings and take one or two off each gear ratio.

OK, quit the game, upload your modified MD file and restart GPxPatch. If all that that worked out, I think you're done.

There's a lot more to MDs than we have covered here and I might do a part 4 to discuss some of those, but that will do for this post I think.

So is it really worth all this effort? It's debatable. What I've written so far should get you where you said you wanted to go, but for me personally, I wouldn't worry too much about hitting a particular speed or achieving an exact real-world lap time. For all its brilliance, GP4 is not and can never be an exact representation of real-life, so my primary aim is always to achieve a nicely-balanced set of cars and tracks that offer a tough challenge every time. I hate to bring up my OS gain like a nerd but it is a bit like when I first started using Linux about ten (actually more like sixteen) years ago. I spent forever trying to make it just like Windows until one day I realized I was missing the point. It would be far easier to changge my own expectations - and from then on it made a lot more sense and we got on much better.

Finally, remember that all your tweaks are subject to something else lurking in the background and shrouded in mystery: GP4's own built in randomness. You can spend hours playing around with MDs, PFs and all the rest and get everything just perfect, but then the next time you race the same circuit, it seems like you need to tweak it all again. It's just GP4 behind the scenes making things more interesting and it's best to accept that too.

Hope that helped.

Thanks for listening. I'm out...



Edited 2 time(s). Last edit at 12/19/2018 08:40PM by Noog.

[Applause]

Very nice write up. The recent progress in GPxPatch to ignore these values for the player has really opened up our ability to calibrate Player vs. AI performance, which goes a long way to making GP4 a great experience.

=====================================================


Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel

Noog Wrote:
-------------------------------------------------------

> I find that
> exciting - and as you'll see in my next
> installment, I'm not really the type of person
> that insists on historical accuracy in all cases.
>
> As for your point about choosing your PF manually
> before the Q&R I'm not sure what you mean. I've
> always found GPXPatch to work pretty flawlessly
> and I think I'm right in saying that GPxPatch
> creates the gpxset.ini, so it must be using it
> correctly, no?


I feel like you Noog, I like the surprise too, to dont know whats going happen. Thats why i change in some races the PF. But dont know of it helps.

What is the deference between using the PF with GPxSet or importing the PF with "TeamEditor"without GPxSet)?

I'm not the best person to ask, but if I were to guess it would be that Team Editor actually patches GP4.exe directly with the data in the PF before you run the game, whereas GPxSet patches the memory with the PF after GP4.exe is loaded. In other words I don't think GPxSet actually writes the PF data to GP4.exe, whereas Team Editor, I think, does. You can see this if you run GP4.exe directly to start the game rather than using GPxPatch to call it.

Personally, I don't use Team Editor anymore. I did use it a lot early on, but ultimately, there's nothing it can do that you can't do another way and there are things it can't do at all. Also, if you don't use CSM and you have a lot of separate GP4 folders like I do, it's easy to patch the wrong one with Team Editor by mistake.

Excellent! I do not know how to thank you. A super tutorial. As I was saying, I have recently started making substantial changes to GP4. Let's see if I can...
I do not use CSM too. I have several "GP4" folders, but I created one with 70 circuits. That's why I needed to have a more performing AI and not very slow adversaries in some circuits and unattainable in others.

Thanks a lot.

Fabio

I use for every mod a gp4 folder too. For me is the CSM more for testing (mods and tracks) and stand-alone converting .
Some times it should me easier to use one PF for all season, but like I said before, the cars retirement are almost the same in every race. The original GP4(2001) wasn´t like that.

@ landex - I never really played the original GP4 much so I don't know.

I liked GP1 on my Amiga 500, then I absolutely adored GP2 on my Amiga 1200 where I had all the tweaking tools and a hard disc to store them on. I even managed to create a couple of decent circuits with beautiful young ladies and other pieces of fine art lying around (yes I really did!), but GP3 and GP4 never really interested me when they came out. I played both a few times, but not for long because I missed all the GP2 tools the wonderful community of the day had built up and my (by then) PCs of the era never really ran 3 or 4 very well in any case. Plus, the beautiful young ladies and fast cars had become real and it was a very busy time for me in general back then. So I'm really quite a newbie when it comes to GP4 - only getting into it in late 2016, after a thirteen year break!

Race-specific PFs (with a single car-specific failure probability value) is definitely the best combination I think. The single PF is great to have for a single non-championship blast around a track but across a season the results are too uniform and predictable for my liking.

I said earlier I don't use CSM and/or Team Editor. Actually I do, but only when I load a mod for the very first time. I use CSM to install the mod to a patched but otherwise unmodified GP4 folder, then I use TE to take a quick look around and I take a few screenshots for future reference. After that I tend to create another default GP4 folder and pull what I want to keep from the mod into it before starting to re-mix myself a homebrew version. Yes wonderful hard-working modders, I confess. I really do pull apart your lovingly-crafted mods almost immediately. It's not meant as a disrespectful act and I know you put loads of time into them but I just can't help myself and it has forced me to learn a lot in a fairly short period of time, which I needed to get up to speed.


@ fabio - Thanks again. Happy to help. I think you've inspired me to write some more tutorials later on too. I used to produce manuals and presentations professionally at one time and I also wrote about 'serious matters' constantly on a popular website I once had some influence over, but I gave up both in 2007 and I think I only wrote one very short comment on one website between then and when I made my first post here recently. I was surprised to find I can still enjoy doing it after all this time.



Edited 1 time(s). Last edit at 12/20/2018 10:58PM by Noog.

Hey landex - I just remembered something I'd forgotten that I built into my PF Calculator recently and I was wondering what you think of it.

Instead of using a constant per-car failure rate for the whole season, what I did was I put in a facility to apply a curve. It's an idea I first came across in MS Project years ago.

Each car's failure rate is averaged over a season, but then I apply one of three curves to the data (I don't think curves was the word MS used but I don't care, I call them curves): Front-loaded, Bell, End-loaded and Flat.

A front-loaded curve will increase a car's failure rate early on in the season and tail it off towards the end (ie simulating a team that slowly resolves a series of recurrent problems)

A bell curve will make the majority of the failures occur around mid-season (Maybe the car falls apart to the point that they have to start replacing the expensive bits they hoped would last)

End-loaded is the front-loaded in reverse (ie a team starts running out of cash as the season goes on; reliability and morale nosedives!)

Flat is flat. (Rather boring but probably the most common).

When I use my spreadsheet and can analyze all the results in front of me, I do see these patterns in the real-world results occurring sometimes.

So far I've made it just an arbitrary thing that I can apply to create unique failure rates for each PF but I might try to find a way to get my spreadsheet to identify the trends in the data and do it automatically.

What do you think? You still wouldn't know which races you'd be failing in, but you'd know the general direction your reliability would be heading as the season went on.



Edited 1 time(s). Last edit at 12/20/2018 11:33PM by Noog.

It's me that I have to thank you. Even this last topic is very interesting. A truly unique study. Perhaps in GP4 no one had even thought about it

Just to throw another idea into this conversion - the AI car setup obviously has an effect on their performance, but I am not convinced the default setups are optimal. That's something else that might be tweaked to provide a better, more consistent experience across different tracks.

=====================================================


Intel NUC 8i3, 8GB RAM, MS Sidewinder Wheel

Noog schrieb:
-------------------------------------------------------
>
> I said earlier I don't use CSM and/or Team Editor.
> Actually I do, but only when I load a mod for the
> very first time. I use CSM to install the mod to a
> patched but otherwise unmodified GP4 folder, then
> I use TE to take a quick look around and I take a
> few screenshots for future reference. After that I
> tend to create another default GP4 folder and pull
> what I want to keep from the mod into it before
> starting to re-mix myself a homebrew version. Yes
> wonderful hard-working modders, I confess. I
> really do pull apart your lovingly-crafted mods
> almost immediately. It's not meant as a
> disrespectful act and I know you put loads of time
> into them but I just can't help myself and it has
> forced me to learn a lot in a fairly short period
> of time, which I needed to get up to speed.








@Tom: yeah, unless you have a trackpack specifically compiled to work with a mod, you almost always have to adjust the AI setup. Fuel usage as well btw, forgetting that screwed me more than once already.

---

Some mods
F1 1996 | F1 2002 | F1 2007 | F1 2011 | F1 2013 | F1 2015 | F1 2018

ROFL!

Think of it this way though Kedy89. Before 'puters came along I used to pull the back off everything in the real world, rather than the virtual ones. When I was little my parents had to keep moving around because if they stayed in one town too long I'd turn the entire area into a pile of rubble. They used to call me the Abominable Termite from Hades and as I grew stronger they were forced to keep me chained up for months at a time until they discovered that I could be subdued for several hours with a hammer blow to my left temple.

Nowadays I just keep one room in my house for 'projects', but I haven't been able open the door in months and I'm thinking maybe I should find another entrance. I might dig a tunnel so I can come up in the middle. They're always great fun. Once I dug a tunnel and came across a underground chamber full of freemasons, worshiping Baphomet and eating babies legs. That was a scary thing, but as my father said years ago after I had secretly turbo-charged his seventeen year old Ford Pickup, "Life is full of surprises". (I never told him I couldn't re-fit the brake pads until after the accident, but I don't think he heard me. His body was spread over several hundred feet and I never did find his ears).

Seriously though, I'm starting to think that the actual car setups (both player and CC), while they do have an effect, are very much subordinate to the more dramatic effects that the PF & MD offer. It's like I use the PF and MD to get things in the right ballpark and then I use in-game wings etc just to knock a few tenths off my time.

In relation to the fuel consumption, I'm currently expanding my PF spreadsheet to include an additional area where some of the values in the MD can be suggested automatically based on a set of user-defined circuit characteristics. Right now it's pretty useless, but it's early days.

Noog, you PF Calculator sounds really impressive. But you have to make PF for each race too, or do you run calculation along the season with one PF?

You make me thing:
The way you do, you should make a profile for each team: Money, etc.... And then, along the season, then wins, 2nd ... the program should calculate and give the price money for improving the car and make the PF automatic. But this should be too "real" and lots of work instead of racing.
Question: Increases the probability of failure in the race if we hit someone?

You played GP2 on the Amiga 1200? I thought that GP2 wasnt on the Amiga! I played on the PC (MS-DOS) but the game was to heavy for the hardware of 1996. I didnt like it much, because there wasnt wet races like F1GP. F1GP on the amiga500 was my 1st one and it was really great, just little bit slow, but if we decrease the detail level it was lovely.

Just one thing: I miss a lot of pretty girls in the mods, thats why I put them in, like you did in GP2



Edited 1 time(s). Last edit at 12/21/2018 04:10PM by landex.

It might sound impressive but it's getting harder and harder to work on. I keep forgetting how it works and it takes me twenty minutes to get my head around it all again. The worst part is, I keep thinking it would be better to re-design the whole thing from scratch with an emphasis on user-friendliness and with the mechanical aspects kept out of sight: meaning that the user makes a few selections from a series of drop-downs and his or her PFs clunk down into the tray (metaphorically speaking), like buying a Coke from a vending machine. It's of little importance if it never becomes an asset to the community, rather than just an elaborate toy for me to play with, which is all it is right now.

A profile for each team involving real things like money etc? I guess you could, but very often we can't be sure why a given team did better or worse at different points in the season. It's just an observable trend in the data. I was just coming up with off the top of my head scenarios to account for it, but I'm really looking to engineer a kind of compromise between failure likelihood being totally randomized across the season, ie a single PF failure value and the multiple PF failure value situation where it's a bit too deterministic. The middle ground would be to add a mild layer of uncertainty, ie a curve.

Thus, if it's early in the season and your reliability is poor but you know your car is going to get more reliable later on, each early success would 'taste' better than it might otherwise. Equally, if you knew your car would turn into a complete dog later in the season you'd be more inclined to give it all you've got in the first few races and build up a cushion in your campaign. Remember though that in more recent years F1 cars have become much more reliable anyway so you might only fail once or twice across the whole season - in which case a curve is irrelevant. That said, I think games should tease our minds a little with a pinch or two of spice wherever possible.

Speaking of spice: yeah those girls man. I wonder if we were to park up next to one they might behave like those 'ladies of the night' in Grand Theft Auto; know what I mean...? Phnaargh!

You ever raced the wonderful Bathurst in GP4? Take a look at what seems to be a cell-phone tower to the right of the first straight. I spent several sessions as I went round the track, trying to catch a glimpse of the whole structure, but you NEVER do. If it was intended that way it's truly masterful teasing by the author don't you think? I also like circuits with statues and other attractive bits of architecture lying around. In fact my first experiment with editing a track was inserting a .3ds of a rickety old water tower into the '67 Zandvoort circuit which is still there to this day.

GP2 not on the Amiga? Oh. Maybe I'd finally consented to dropping down to a PC by then. To be honest my memories are pretty scrambled around that period and I really don't remember for sure. Like you, I definitely played F1GP on an Amiga though. I loved those Amigas back in the day - and just like GP2 with GP2Edit.exe (as I think it was called), you could tweak forever and never run out of things to do!

Increasing the probability of failure (as in breaking something) when we hit something is determined somewhere else, but I can't remember where and it's not something I've ever experimented with. By default my spreadsheet, like PRBlanco's original, doesn't discriminate as to why someone didn't finish; only that they did or didn't, so the formula for the single PF is something like this:

(# of failures / # of races) x (All cars will fail value [24578 IIRC] / race length percentage)

Yes, the cars are more reliable than in the classic era 60s to middle 90s. But i dont know why, i like that. And the mods released are really great. Only the car failure dont seams to fit in some mods.
. Zandvoort 67 its just great. I put some ads for 79, I thing. And I little modify the pit entrance.
Those ladies in GTA needed a horn to... I never saw a F1 car with a horn but it should be fun instead of blue flags...
I dont have Bathurst, where can get it?

There is a failure(engine) who is not generate the beginning the race, can appear while you racing, I test it in GP3 and GP4 and this great and it should be always like: If you drive whit high revs, p.ex. always in 1st and 2nd gear, the temperture goes high and the engine blows. Just great! Thats the way it should be. I dont know if are more failure who can appear, instead of generate them at the begining the race.
In some mods the AI cars retires because it gets out of fuel. Even if i have fuel just for less 1 lap i can manage it by slowing to get in the pit. I lose some places but its better than retiring.

here we go
[www.mediafire.com]

Quote
landex
Yes, the cars are more reliable than in the classic era 60s to middle 90s. But i dont know why, i like that. And the mods released are really great. Only the car failure dont seams to fit in some mods.

I like it too.

When you say the failure doesn't fit, do you mean that the total number of failures for a driver is very different to what it was in real-world, or do you mean the reasons for breaking down differ to those in the real world?

If it's the former, remember that PFs can have failure determined by team, but these can be overruled by using individual settings for each car. My early experience with car specific values (which I learned about only recently somewhere on the forum) is very good. Since I can't find the post again, I'll outline it here. Basically you can put all these things in any PF after the 'teams' section.

[FailureProbability]
Team #00=10000,10000
Team #01=25000,27777
Team #02=13333,32142
Team #03=16666,26923
Team #04=16666,32142
Team #05=28571,11538
Team #06=28571,19230
Team #07=33333,28571
Team #08=23076,23333
Team #09=46153,16666
Team #10=30000,23333

[QualifyPower]
Team #00=734,735
Team #01=715,715
Team #02=723,724
Team #03=728,728
Team #04=723,721
Team #05=695,694
Team #06=738,740
Team #07=708,706
Team #08=715,712
Team #09=732,732
Team #10=719,716

[RacePower]
Team #00=725,724
Team #01=700,700
Team #02=716,714
Team #03=720,720
Team #04=706,707
Team #05=685,686
Team #06=720,719
Team #07=690,692
Team #08=697,700
Team #09=712,712
Team #10=697,699

(FP seems to work well but I'm currently unsure about the utility of QP and RP above. Like some of the mods do I'm starting to think about using a single QP for all cars and another for the RP).

Quote
landex
Zandvoort 67 its just great. I put some ads for 79, I thing. And I little modify the pit entrance.

Do you find the CC cars too slow at the last corner at Zandvoort? If you do, try changing desc51 (cc aggressiveness) in the MD file from 60024 (the default in mine I think) to 0. It doesn't resolve the problem completely but I found it useful. I was messing around with it recently trying understand desc51 better. That's how I came to make the Christmas picture...

Bathurst can be found here

You are in for a treat. It's definitely one of my favorite circuits. It's pretty intense, like the Kleine Nordschleife; another exceptional circuit.

Kleine Nordschleife can be found here

I usually overheat the engine deliberately if I lose my transmission. Once your engine blows you get your transmission back! It seems GP4 doesn't let you have two failures at once. It's cheating for sure, but I've made a custom engine blowing up sound and I still enjoy hearing it!

If you have problems with too little fuel, see desc70 (player fuel consumption) & 71 (cc fuel consumption) in the MD. When I adjust it I change it in increments of 1000. Higher number = less fuel. (Usually the CC figure is 1000-2000 less then the player figure). Remember to leave a bit of fuel spare too: if you have an off into some deep gravel you'll be wasting fuel getting out of it.



Edited 1 time(s). Last edit at 12/22/2018 08:04PM by Noog.