Patch Z Balance Testing & Car Changes

[DeadWolfBones]

Hello, drivers!

As you are aware, Patch Z has introduced some physics changes to the GTR class, including a lower center of gravity for all cars. As such, we need to reassess the balance of the cars for IGTC.

To do so I would like to propose a twofold strategy:

1. I'd like to have at least one but preferably two or three practice races of roughly stint length (so let's say an hour) over the next week or two. We can schedule these whenever is most convenient for the majority of teams, but I think the usual race time (18:00 UTC on Saturday) would be alright. This could be moved to Sunday if teams need more time to prep. Are people ok with a 1hr race this Saturday the 5th?

I'm split on whether we should use KY3R (will be racing there, most teams need practice on it anyway) or a track we've already used like WE1 (teams will already have a decent setup and benchmark times). Any suggestions would be helpful.

2. Meanwhile, I'd like all team leaders to report to me and Benji honest lap times for both qualifying and race trim for the upcoming race as you develop your setups in practice. We also encourage you to fill us in on any changed handling characteristics/tire wear/etc and to generally give us your opinions. In short, we're looking for as much information as we can get. We promise not to share these times with our respective teams--they will be solely used to develop the balancing for the remainder of the season.

***

In addition, due to the physics change we will allow any team who wishes to do so to change their car model for the remainder of the season. Teams who want to make a change should have their manager/co-manager make a post stating the change in the signup thread.

This is a one time change. You will not be allowed to change again this season, barring another physics update.

[srdsprinter]

Thanks Ben!

Good looking plan.

We (S3R) should be able to get some drivers out soon for test races whenever (probably driving some different cars as well).

[DeadWolfBones]

Also, Starblue is working on a program that compares AI times with the new physics. We're not sure how useful it'll be but it's worth checking. :)

[DeadWolfBones]

And I forgot to mention that for the tests, we'd prefer it if you stick to the cars you've been driving all season and are (presumably) more comfortable with.

[Dragonmen]

AMT might try new car(s). :) We will put some drivers on the track I'm sure...

Also if there will be more test races, first one can be done on WE1, but only if it is on this weekend... I'm sure many of the drivers already started their KY3 preparation.

1h race is good! :thumb: No less :tilt:

EDIT:
We'd still prefer to try new cars but we might not do it on this test race if that is general opinion.

[Lotesdelere]

Yep, good idea but:

I'm split on whether we should use KY3 (will be racing there, most teams need practice on it anyway)
Next race will be on KY3R so I guess that's the one you're talking about :)


Meanwhile, I'd like all team leaders to report to me and Benji honest lap times for both qualifying and race trim for the upcoming race as you develop your setups in practice. We also encourage you to fill us in on any changed handling characteristics/tire wear/etc and to generally give us your opinions. In short, we're looking for as much information as we can get. We promise not to share these times with our respective teams--they will be solely used to develop the balancing for the remainder of the season.
Well it's not that we don't trust you but you may be very tempted to tell to your respective teams things like "TeamX or TeamY is faster than us in that sector" even if you're not giving exact times, and we'll have no control over that.

Plus using only one track for the tests is not good IMO.
Why don't we run the tests on the tracks that we have already used during this season ? So we'll have concrete things to compare and nothing to hide.

[scipy]

what is it with this nansy pansy bullshit? where are the good old days when we had 3 cars and if u wanted to win u picked the fastest one? seriously, xrr got a 13 mm lower CoG and the 1.5% backwards moved distribution means that xrr the front tires will last even longer than they do now, meanwhile fzr still has a damn clutch and now it's understeery. just keep the balancing as it is on the master server.

oh, and the thing about "not revealing the pace".. seriously. knock it off. there is NOTHING you can do about someone elses pace, literally, NOTHING. if someone told u they are doing mid 2:12's at start, what would u do? would u go out and buy wider tires and just go faster? ffs.

[srdsprinter]

@ Lotesdelere:

Most of that info can be found on LFSworld currently anyway ;)

I agree maybe re-running some of the previous tracks again would be a great idea.

[DeadWolfBones]

Yep, good idea but:


Next race will be on KY3R so I guess that's the one you're talking about :)

Yep, that's what I meant. Thanks for the catch.

[DeadWolfBones]

what is it with this nansy pansy bullshit? where are the good old days when we had 3 cars and if u wanted to win u picked the fastest one? seriously, xrr got a 13 mm lower CoG and the 1.5% backwards moved distribution means that xrr the front tires will last even longer than they do now, meanwhile fzr still has a damn clutch and now it's understeery. just keep the balancing as it is on the master server.

oh, and the thing about "not revealing the pace".. seriously. knock it off. there is NOTHING you can do about someone elses pace, literally, NOTHING. if someone told u they are doing mid 2:12's at start, what would u do? would u go out and buy wider tires and just go faster? ffs.

Stuff it, scipy.

[JasonL220]

shall we run +15kg in the fzr in the test races or no handicap?

[DeadWolfBones]

No handicap in the tests.

[Starblue]

Also, Starblue is working on a program that compares AI times with the new physics. We're not sure how useful it'll be but it's worth checking. :)
AIs have already done more than 7000 laps on my PC....they've been driving all night long :)

This is my current plan, feel free to make suggestions:
0) I'm testing Intake Restrictions. Why? Because it's only 1 variable instead of 2, i.e. mass restriction have both kilograms and position to adjust and test.
1) First of all, I run a series of tests to see Intake effects. That is, how performance degrades with various levels of intake restriction.
For every car (FXR, XRR, FZR) and every restriction (from 0 to 19%) I'll run multiple races on a few tracks (KY3R, WE1 and maybe SO4 and AS7).
I've already run 12 races, 10 laps long, for every car, on KY3R with 20 drivers (0-19% intake). The best times graph gives a nice overview of how intake restrictions affect each car's performance. As expected FZR is more affected by restriction than the other cars (e.g. using 19% intake makes FZR about 5% slower, while XRR/FXR are only about 3.5% slower).
I'm now running WE1 to see if changing track affects these values :)
2) However, IGTC races are not 10 laps long and they are not about "fastest laps": cars who use less fuel have better average times and so they have an advantage on long races. For this, I'll check the "Stint effects", how carrying fuel for one stint affects car performance (avg time vs best times).
For every car, I'll run a few 1 hour long races (stint length) on various tracks. During each race, some cars will bring intake restrictions to see if intakes alter the stint effect.
3) It is usually not possible to balance cars so that they are equal on every track ("Track Effect") so I'll do some maths to check the "Season effect" (i.e. track effect over remaining races). That is, summing up Stint effects + Intake effects + Track Effect what would be the restrictions to apply to each car so that all 3 cars have the same chances to grab points in the remaining races?

Obviously all these tests are good for AIs, but they must be checked with some "Human" tests to see which values can be used and which ones can't be applied. For example, Intake effect results might be ok for humans too (e.g. a FZR with 19% intake might run about 5% slower on KY3R, no matter if AI or human driven) while Track Effect might change (e.g. XRR is 3,5% slower than FZR on KY3R if used by AI, but it might be only 1% slower if driven by a human...because AI maybe don't know how to handle XRRs..).

Also, variables related to setups aren't tested...all AI are just driving with a basic restricted setup for now. It would be probably best if we had real race setups, especially for the Stint effect test. If some teams could provide some setups for that, it would be a lot better :)

So, at the end of all the tests, we will not be able to just "blindly" use these values, but still I hope that by doing a proper organized testing we might get a better overall picture of how the variables interact and gain insight on how we could balance them.

Starblue

[R.Kolz]

Doesn´t look like there is going to be a test race tonight, right Ben ?

Z balancing now.This is going to be a very hard task to do.

Even that I feel that the FXR never got involved in any GTR car balancing (http://www.lfsforum.net/showthread.php?p=699620#post699620) and we for that reason as a FXR driver/team might as well stay out of this discussion, please allow me to propose something:

Come with some scedule for 5 races of 1 hour for the next 14 days. Let´s use the tracks of all the IGTC rounds 2008 we raced so far.
Invite the teams as well as some fair and fast drivers who don´t race in the IGTC for these test races. This way you should be able to get some valid data for your balancing of the GTR cars.

I collected some indicating data of the 5 previous rounds 2008 of IGTC for comparism purpose. See attached file.

Regards, R.Kolz

[DeadWolfBones]

Thanks for that data, Rudi.

No race tonight, but there will be one tomorrow (18:00 UTC on the 5th). After that we can discuss where the next one will be held. I was thinking about using WE1, but after looking at your data it (unsurprisingly) shows the largest disparity between the three makes... might not be the best choice.

How does everyone feel about BL1?

[rc10racer]

I would think AS5 would be the best place to test because it has most things that blackwood doesn't e.g. long straights, fast left to right corners plus the most things what gave the fzr the advantage was the power out of the corners.

The xrr has better balance now, fzr has more weight to the front to help take away with the advantage they had, fxr is just about the same from what i can tell.

[DeadWolfBones]

Problem with AS5 is that we haven't run there this season, so none of the teams have a proper baseline setup for it.

[Myros]

I think AS5 is the best choice.

[The Moose]

Cant make the race tonight I'm afraid.

Does anyone have some simple tips on what changes to setup the FZR requires with patch Z pls.

[Gil07]

I think we should go for various tracks over the coming weeks/days, to see how the balance is on them... KY3R for today?

[BigTime]

Problem with AS5 is that we haven't run there this season, so none of the teams have a proper baseline setup for it.


True but AS4 is basically the same track. I doubt I'd even make a setup adjustment...

[rc10racer]

True but AS4 is basically the same track. I doubt I'd even make a setup adjustment...

Maybe a little play with the downforce then your sorted.

[DeadWolfBones]

Ok, so no one (except WBT, thanks Myros) showed up for this.

Can I please get some ideas as to when to hold a test race that's convenient for everyone?

[srdsprinter]

Ah, I thought we were going at 3... I never get the UTC times.

[DeadWolfBones]

I can stick around till 3 if anyone else will show.

[Scott_Michaels]

If you aren't going to make it a little more obvious (ie, giving the race its own thread) people aren't going to notice!

[DeadWolfBones]

I assumed people would look in the balancing thread, since it's explicitly about the balancing test races.

:shrug:

[Scott_Michaels]

Some people wont read all posts with a fine tooth comb. I know I just try to get the jist of things and look out for important info (as I'm not familiar enough with LFS to offer any insight).

[Starblue]

0) I'm testing Intake Restrictions. Why? Because it's only 1 variable instead of 2, i.e. mass restriction have both kilograms and position to adjust and test.

Notice: I have been told that people do not like Intake Restrictions...so I'm now testing Mass ballast effect. As for the position of mass ballast, I've assumed:
41% front for FZR
50% front for XRR
53% front for FXR
Adding ballast on these positions doesn't alter car weight distribution with 100% fuel and driver on (maybe it would have been better to check on 50% fuel actually).

Starblue

[Scott_Michaels]

Why don't people like intake restrictions? Surely it would affect your setup less and make the car easier to drive, thus conserving tyres, and also fuel?

[BigTime]

Ballast is fine and all but there is nothing wrong with intake restrictions IMO. One thing about ballast is, you can move from front to rear. I think it would actually be harder to prepare for a round with the new system.

[Rudy van Buren]

Problem is XRR is already faster in corners, giv fzr extra weight and the diference will get even bigger, im for air restriction :thumb:

[JasonL220]

i think intake would be better, as adding wieght means that the setup will be have to have quite a few change but intake would need very few changes

[Lotesdelere]

Why don't people like intake restrictions?
Because we want to drive full GTR's with a lot of power. Not 75% or 85% of a GTR.

We could also use the XFR and the UFR, but that's not what we have signed for. We have joined a serie with 500 bhp cars, which can do 280 kph and brake in 100 meters. Because it's going to be hard and not because it's going to be easy ;)

[DeadWolfBones]

The amount of balancing we'd need would likely only necessitate a few % change, but yeah, that was also my issue with the idea of intake restriction. Well put, Lotes.

[rcpilot]

I found a bit of ballast to play with nice in the LOTA GTC series. (Season performance ballasting) Get to have a little say in the weight distribution of the car on a track by track basis, it improved the handling of the FZR a lot back then but I had about 120 kg to play with. (And the FZR was way too far back in its weight distribution in the first place even for a good MR setup, and is still a little too far back from my experience.) Could actually gain a few tenths back and partially negate the damage dealt by having the ballast.

[Kardum31]

This is just a thought as I don't have much experience with LFS myself, but if we'll be sticking with ballast through next season, maybe it would be worth considering starting with a base of something like 20 kg of ballast for every car (adding more where needed to faster cars) and giving the field another setting to toy with.

[Scott_Michaels]

So a performance based system like the DTM? Sounds like it could work, although its slightly complicated but easily manageable.

[Kardum31]

I was thinking more along the lines of everyone getting 20 kg, and then adding an additional 15 (or whatever management decide) for the FZR.

[DeadWolfBones]

I was thinking more along the lines of everyone getting 20 kg, and then adding an additional 15 (or whatever management decide) for the FZR.

It's an interesting idea, but not one I really want to consider for a mid-season change. I'll file it away for later, though. ;]

[DeadWolfBones]

After reviewing the available data (such as it is) and consulting "expert" opinions, it's clear to the admins that we don't have the necessary information to accurately and fairly impose any external balancing.

We feel that the master server balancing allows for close competition between the two faster cars, and while it may favor one car or the other in some way, it's undoubtedly by a very small margin. The FXR remains roughly where it was with the patch Y balancing, which we also feel to be appropriate.

Therefore, for the remainder of the season we will use the master server/default LFS balancing. All teams are welcome to change their car model at this point, but the car you race in round six will be the car you race for the remaining two rounds.

Looking ahead, we will use the results of the final three races, and similar results from other endurance leagues to determine how to handle balancing for the 2009 season.

Thanks for your patience over this period of uncertainty, and thanks to the marshals, teams, drivers, and friends of IGTC who helped in our testing efforts.