GPxPatch 4.52 [Released]

Posted by SDI 
Re: GPxPatch 4.52 [Released]
Date: August 03, 2019 04:57PM
Posted by: SDI
@Noog: I honestly have no idea about another variable for this behaviour, or if this is indeed something that happens in the code.

@ParisHL: I see no logic behind setting BRMin the negative of BRMax. As far as I know, the difference (BRMax - BRMin) is the braking consistency. So what you're doing is make them brake later sometimes but also making them brake earlier sometimes. Unless you can prove otherwise with experimental data.

René Smit, Independent Software Developer.
Download my GPx tools here.
Re: GPxPatch 4.52 [Released]
Date: August 03, 2019 08:18PM
Posted by: Noog
OK, thanks René. I realized after I posted my q that it wouldn't make sense to be in the (non-driver specific) MD anyway would it?

About BRMin/Max: I still don't see how using two variables (BRMax/Min AND Error Chance) adds much (apart from increasing PF complexity). I'm still yet to play with Error Chance but it strikes me as a very DRIVER-centric variable whereas the absolute braking point is determined by a host of factors, most of which don't have much to do with the driver, eg weight of the car, the weather, the brake balance and whether or not someone is in the way etc. I'm willing to be convinced otherwise but it seems to me that the Error Chance variable rendered BrMin/Max somewhat obsolete the minute you gifted it to us René!
Re: GPxPatch 4.52 [Released]
Date: August 04, 2019 01:03AM
Posted by: SDI
Ok, maybe you're right. After all these variables weren't designed to be driver specific but track specific.

René Smit, Independent Software Developer.
Download my GPx tools here.
Re: GPxPatch 4.52 [Released]
Date: August 04, 2019 03:23AM
Posted by: ParisHL
We use GP4 on an online forum, much like GPGSL. We use the BRMin/Max as driver consistency per lap. So for example, a driver with poor consistency is likely to occasionally lock-up, but then on the next lap might brake far too early etc. So going into the minus like we have been doing seems to make sense.

ErrorChance is similar (some drivers are more prone to mistakes than others) but really affects the amount of "offs" the driver may have during a session. I think if BRMax is set too high, then drivers will obviously brake too late for certain corners and run wide or crash out which isn't because of ErrorChance but more because the BRMax was set too high. By tweaking BRMin/Max too be much smaller, it'll still encourage overtaking and inconsistent lap times with the occasional lock-up without making it super obvious. That way we can differentiate between a driver being inconsistent and being error prone.

So from the bit of testing I've done, BRMin/Max works well for overtaking moves and over one lap such as qualifying where an inconsistent driver is less likely to have a lap time that's representative of their skill level.

We are still trying to tweak the numbers to a realistic level where the number of lock-ups and mistakes is similar to real F1, but I definitely wouldn't be putting those ratings over 500 for now.
Re: GPxPatch 4.52 [Released]
Date: August 04, 2019 10:04AM
Posted by: Noog
Funnily enough, from my limited testing of BRMin/Max I too ended up with 500 BRMax for one or two cars and sometimes one car up to 1250 (when I really didn't like a certain driver's ability and wanted to make him a laughing stock ;) ). I never liked really BRMin at all though and rarely if ever used it as it always seemed the more artificial looking of the two behaviours.

What stopped me using it eventually was the still noticeable effect it had on CC results over time, because back then (and still now) I was working hard on an enhanced Performance Calculator spreadsheet to automatically generate driver-specific PFs for each race of each season and I didn't want to have to accurately feed the effect of the additional variable into my already cumbersome grip calculation formulae to compensate. Basically, even at low levels such as 500 I felt BRMax tended to skew the accuracy of my PFs when considered over several seasons and then one day when I left it out altogether I realized it was actually pretty hard to tell the difference on the track anyway. And then ErrorChance came along, offering another layer of complexity...

I'll probably come back to ErrorChance soon but for now I don't use either. Not sure that makes any difference to anyone else but that was my story.

Good luck finding that sweet-spot. It's probably out there somewhere!

ParisHL Wrote:
-------------------------------------------------------
> ErrorChance is similar (some drivers are more
> prone to mistakes than others) but really affects
> the amount of "offs" the driver may have during a
> session. I think if BRMax is set too high, then
> drivers will obviously brake too late for certain
> corners and run wide or crash out which isn't
> because of ErrorChance but more because the BRMax
> was set too high. By tweaking BRMin/Max too be
> much smaller, it'll still encourage overtaking and
> inconsistent lap times with the occasional lock-up
> without making it super obvious. That way we can
> differentiate between a driver being inconsistent
> and being error prone.
>
> So from the bit of testing I've done, BRMin/Max
> works well for overtaking moves and over one lap
> such as qualifying where an inconsistent driver is
> less likely to have a lap time that's
> representative of their skill level.
>
> We are still trying to tweak the numbers to a
> realistic level where the number of lock-ups and
> mistakes is similar to real F1, but I definitely
> wouldn't be putting those ratings over 500 for
> now.
Re: GPxPatch 4.52 [Released]
Date: August 07, 2019 09:03AM
Posted by: kedy89
@René

do you have any idea why the font gets messed up in 4K resolution for the digital 2010 mode? The other 3 modes seem to be fine
[i.imgur.com]



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

Some other stuff ::: Some more other stuff
Re: GPxPatch 4.52 [Released]
Date: August 07, 2019 09:20AM
Posted by: munnzy
what resolution are you running
Re: GPxPatch 4.52 [Released]
Date: August 07, 2019 01:13PM
Posted by: TGF1Driver2
Rene, I might have found a bug concerning the autosave function yesterday. When driving at Magny-Cours 2007 (70 laps) I wanted to repeat the last lap. The file autosave.raj is the standing after the last lap, the file autosave1.raj is the standing before the last lap, so I took this. After driving the last lap again I wanted to drive the last lap once more, but for some reason autosave1.raj was now the standing after 70 laps as well. That's quite annoying but interesting as well. I tried it with other tracks in other slots as well and it's the same. Can you have a look at this issue please? Thanks.
Re: GPxPatch 4.52 [Released]
Date: August 07, 2019 08:45PM
Posted by: SDI
@kedy89: I'm not sure, but in the past it helped to lower the font quality by adding a Quality=0.75 (for example) in fonts.ini under the appropriate section for the font.

@TGF1Driver2: This is how I intended it to work. When you do the last lap again, then at the end of it the autosave.raj (which was the old last lap) is moved to autosave1.raj, then the autosave.raj is created (which is the new last lap).
If you're keen on keeping the old autosave1.raj you should quicksave right after loading it.

René Smit, Independent Software Developer.
Download my GPx tools here.
Re: GPxPatch 4.52 [Released]
Date: August 08, 2019 08:07AM
Posted by: kedy89
SDI schrieb:
-------------------------------------------------------
> @kedy89: I'm not sure, but in the past it helped
> to lower the font quality by adding a Quality=0.75
> (for example) in fonts.ini under the appropriate
> section for the font.


Thanks a lot, that seems to have fixed it for most part. Only still noticed it when a special character was involved. I'll keep playing around with the value, see if I can get rid of it completely.



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

Some other stuff ::: Some more other stuff
Re: GPxPatch 4.52 [Released]
Date: August 08, 2019 08:45PM
Posted by: TGF1Driver2
SDI schrieb:
-------------------------------------------------------
> @kedy89: I'm not sure, but in the past it helped
> to lower the font quality by adding a Quality=0.75
> (for example) in fonts.ini under the appropriate
> section for the font.
>
> @TGF1Driver2: This is how I intended it to work.
> When you do the last lap again, then at the end of
> it the autosave.raj (which was the old last lap)
> is moved to autosave1.raj, then the autosave.raj
> is created (which is the new last lap).
> If you're keen on keeping the old autosave1.raj
> you should quicksave right after loading it.

Well yeah I should have used quicksave, but I didn't know it before that it will override the lap before the last lap. Well, I drove the whole race again today and was 41.3 seconds faster than the day before yesterday! ;)(Y)(Y);-)
Re: GPxPatch 4.52 [Released]
Date: August 28, 2019 12:44PM
Posted by: Soutsen
@SDI since you're able to extract data from MDfile and add it to the perf file, could you make add sectors from desc102 to desc124 to it so we can simulate the pit-stop strategies (instead of random) for the AI cars? Also ideally it could be fun if you'll add the ability to use all the MD values into your perf files, cos it would be cool to simulate the turbo engines for Pedro's mods or to simulate the extra grip for the ground-effected cars, or to simulate the different fuel/tyre usage for the people who known for saving a fuel/tyres (especially tyres these days like Perez). I know it can be complicated with all of these nubers added, but those, who don't wanna use it will stay with the simplified perf files.

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]
Re: GPxPatch 4.52 [Released]
Date: August 30, 2019 02:42PM
Posted by: Soutsen
Also, have anyone made enhanced perf files, using all the latest features of the GPxPatch tool, such as different braking range values and specific driver errors? I'd like you to share them here

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 04:30AM
Posted by: Soutsen
It might be me doing something not right, but I can't find how can I solve the issue. This is my perf file for the latest GPxPatch (4.52):
deleted to save you from scrolling through it


I load it using a "power and grip" setting ticked up. But as you can see on the screen, cars are lined up using a [RacePower] values (and their power during the race itself is also programmed accordingly to [RacePower] value) with the two Ferraris and the Prost at the back, while the [QualifyPower] seem to be completely missed out the calculations :-/

Also, what the [File]version=213 does? Is it some sort of a mandatory line?

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]



Edited 1 time(s). Last edit at 09/02/2019 05:47PM by Soutsen.
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 08:52AM
Posted by: Noog
I believe there is a lower limit for power, below which things behave oddly. I don't know what it is, but try making the lowest power level 530 or higher.

Also, just for the purposes of experimentation, you can ditch the "Performance=1,1,1" lines altogether if you like. It causes problems with TeamEditor3.5 to leave them out but I would get rid of those lines for now, then make sure no-one has a QP or RP power rating lower than 530 and see what happens. Even a power difference of 10 between cars (with grip being equal like you have it) should show clearly on the track.
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 10:26AM
Posted by: Soutsen
Yeah, the low values can bug out the game sometimes, I remember it from trying to form a custom starting grid before the custom grid tool came out.
Anyway, it don't seem to be an issue this time. The game still uses ONLY the race values to form the grid.
Could you use the following perf file on your side and check the results of the skipped qualification session before the non-champ race (or before the short race - I believe the grid formation works the same way). From these values Ferraris should start the race from the lead, but I have them at the back:
to see the spoiler: right click on it - open in a new tab

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]



Edited 3 time(s). Last edit at 09/02/2019 10:29AM by Soutsen.
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 03:03PM
Posted by: ParisHL
Is there a way to have two teams like Red Bull and Ferrari where one team is faster in corners and the other faster on straights?

I've tried giving teams more bhp for higher top speed but should I just bump driver rating up for faster speeds in the corners?

I want to try and emulate a team struggling in Monza but doing very well in Monaco/Hungary etc. And vice versa
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 04:45PM
Posted by: Soutsen
Yes, basically more grip (driving ability) makes them go through the corners more effectively, while bhp is for the top speed mainly. Still, you probably need to make a track-specific perf files to make it even more effecient. Also the track configuration (magic data) and CC-line plays a role at how good they are going through the corners. The main value for it is the 51 line of the Magic Data ("CC braking aggresiveness" or something like that). It is varied with the different values at the different tracks though. Usually I prefer AI cars to brake a little earlier to avoid tangling up at the chicanes, but afaik from now on you can use the different values for each driver to make them brake earlier (lower values) or later (higher values). I never tested it out yet, but I suspect though that the randomisation of this values adds a little bit more of tangling up or bumping each other due to different braking values. 42 line in MD file (track grip) is also important at this area. I think this post is the most useful to understand how the MD works, though it can be not really intuitive. Also the values for "CC aggressiveness" seem a bit odd to me at this post, since it can surely go above 45000 without game braking into any severe bugs or the odd behaviour.

Almost forgot - GP4 car physics also plays a role in that, but I have like 0,001 experience in tweaking the physics file yet.

@SDI or anyone - could you look at the QualyPower issue I have? Just let me know if you have the same problems using these values as I do.

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]



Edited 5 time(s). Last edit at 09/02/2019 04:53PM by Soutsen.
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 05:22PM
Posted by: Noog
You're right Soutsen.

Just tested your PF with two different circuits; made several changes and tried again.

QualifyPower is not being used by the game to determine qualifying position. It's using RacePower - no doubt about it.

@ParisHL - Using a separate Performance File for each race is way to do what you want. Remember that nothing in the Magic Data file is car or team specific so only the PF will do.
Re: GPxPatch 4.52 [Released]
Date: September 02, 2019 05:45PM
Posted by: Soutsen
Yay, we've noticed a bug! Version 4.53 is going to happen, I hope
Thank you for offering your help too

ccc
___________________________________________________________________________
For a list of EVERY download for GP4, look here:[docs.google.com]



Edited 1 time(s). Last edit at 09/02/2019 05:53PM by Soutsen.
Sorry, only registered users may post in this forum.

Click here to login

Maintainer: mortal, stephan | Design: stephan, Lo2k | Moderatoren: mortal, Vader, stephan | Downloads: Lo2k | Supported by: Atlassian Experts Berlin | Forum Rules | Policy