Hello.

I just had the idea to include a balancing-mode into the singleplayer of the testpatches.
A quick draft of the system would look as following:
- In normal operation the testpatches work as they always do, no change except the typical features and bugfixes that come in the testpatches as they do now.
- Additionally to the normal bugfixes the testpatches also contain new stats for the cars that can be triggered using an argument to the file (e.g. "LFS.exe -balancing") which will alter the stats of some of the cars, disable the multiplayer and invalidate all hotlaps. In this mode the players who are willing to test can drive the proposed car-stats and see what times are possible with them and propably share them in the forum using the invalid hotlaps.

The pros are easily explained:
- A big group of testers and thus a very low statistical error (there will always be people in the testgroup that know how to handle the XRR very well, f.e.).
- The multiplayer mode is completely untouched. No version mismatches, no forced updates. The same ease of use we all appreciate.
- The hotlaps on LFSW remain untouched even though balancing can be actively improved.
- No disturbance for lesser involved people. You have to modify a shortcut's properties to enable the balancing mode and thus you can only stumble upon it when reading the testpatch thread. After having read that every user should be knowledgeable enough about the balancing mode to know it's restrictions and confusion should be very limited.
- The system could be improved later to allow the public testing of other features, like different stats for the different tyre-types or changes to the graphics-engine that might cause incompatibilities.

The cons aren't too complex either:
- The implementation of the car-stats needs to be altered. Either copies of the current cars with the new stats have to be implemented, complete with encryption, or the format of the cars' stats needs to be enhanced to allow the new stats to be enabled without disclosing them to third party modifying (-> encryption).
- It's a parallel project to the current development of compatible patches. In a one-man-job serial work beats parallel work by far in means of productivity.
- Lots and lots of opinions about the changes and lots and lots of work to filter out the objective bits from those. Someone needs to evaluate all opinions that the community writes into the forum. Though this might be weakened if Victor creates a secondary hotlap chart for the new balancing seperate from LFSW. Then the changes would be clearly visible as a change in laptimes.

I'm not sure wether this is a good idea or not. It has strong positive points, like a better likelyhood of well balanced classes, but also downsides, like additional work. Any opinions?

Vain

I would be suprised if you'd get enough depth in the numbers from this method, however I do not think it would be terribly difficult to balance the cars over time, making small changes in each incompatable patch update.

The problem is interest from the devs, when they want to turn their attention to car balancing they will, but until then we have to wait It's extremely frustrating, which as a fellow STCC driver I am sure you can appreciate (the TBO's are massively mismatched right now).

I'm not sure how well the GTR's are balanced as when i've raced them I do see a selection online and I see varied winners with them, but in TBO there is no such thing, it's FXO to win, FXO for ease of driving, FXO for short races, FXO for long races, and RB4 for RallyX.

I would also like to see a success ballast system for leagues to use, and if this where too happen we would also get a temporary solution to car balancing too .

EDIT: Success ballast could work as a setup option on the info panel (not saved with setup) where an admin can enforce a minimum ballast with a /command. Also a /listballast command to show which drivers have their ballast applied too (so that relogging doesnt sneak a driver through without the ballast).

I don't totally understand the balancing issues in the GTR class, 4Wd is the easy one to drive, why it should be equally fast? The balancing could come in later when more stuff and features are added. The FZR could cook up it's brakes more over time for example. And 4wd could have advantage on slippery surfaces. They could fit some sort of antilag turbo to XRR, or it could have a better gearbox than FZR like said in a previous thread made by Mr. Vain

I think making such system where people could mess around and try to balance things only would take way too much Scawen's time and give back too little.

edit. And how could we actually test the classes properly without multiplayer?

It is frustrating, being in the RB4 (much weaker car) and the XRT drivers must be frustrated too. I personally think that the ballast would be nice, depending on the admin, have say for BL2R RB4 have 100 LB ballast, XRT have 50 and FXO have 0...

Quote from Blackout :

I think making such system where people could mess around and try to balance things only would take way too much Scawen's time and give back too little.

edit. And how could we actually test the classes properly without multiplayer?

1. No user would fiddle around with stats. The devs would include new proposed stats for the cars for testing in the balancing mode. That would happen about once in one or two months.
2. All balancing issues in LFS are directly visible in the WR chart at LFSW. The main problems are FXO vs RB4 + XRT and FZR versus XRR and FXR. These problems can be seen clearly in the WR charts. If we would find values for the cars where this gap is minimized we could assume that the performance in normal racing situation has to be similar and can't be too far off.

I'm not speaking of a 'solution' - every car should have it's advantage on some track or in some situation. I'm speaking about a way to merely improve the balancing. I don't want to see any cars that are always and in every aspect superior.

And before anyone wants to tell me I'm just being jealous at faster drivers: My favorite GTR is the FZR because of it's RR layout. But it is so far superior over the other cars that there is no challange in driving against those unless I find the WR holder of that car. So actually there is no GTR class. Because there is no competition (I just took a look at LFSW - I didn't find a track with similar laptimes unless you search for sub one minutes tracks).

Vain

If we want to slow down the FZR, shrink it's fuel tank to 80 litres or so.

How does it get slower with that?

Less fuel in the car means less chemical energy in the car means less power at the wheels. Obviously. Duh. Or perhaps the space that was taken up by tank (mostly empty most of the time) is replaced with lead, or really dense fruit cake.

Quote from tristancliffe :

or really dense fruit cake.

this is not going to help draw drivers away from the FZR, but the if the wheels on the XRR where wagon wheels...

Please don't make balancing-suggestions in this thread.

If you wonder why I just said that please look at the topic.

Vain

P.S.: Stop making loads of questions and comments that I have to answer. My sadness rating is already higher than it should be!

mmmm...cake

Doesn't the devs allready have their test team which should do these kind of things.

Quote from Blackout :

Doesn't the devs allready have their test team which should do these kind of things.

They do. With obvious results.

Vain

Quote :

Doesn't the devs allready have their test team which should do these kind of things.

Look at the TBO world records and tell me that's true... It's even worse when racing as the dominant car, the FXO, is easier to drive with sub-wr driver skills which even further exagerates the differences.

Ahh, the INVINCIBLE FXO SYNDROME again.

Giving responsibility of the car stats evolution to a trusted person outside of Scavier would solve them the extra workload of monitoring. There's probably no need to have more than one person in that position of direction, since all they'd have to do is monitor the test players' feedback, and change the stats accordingly.
Such a third person would also mean much more frequent adaptations to feeback.

The comparative performance of cars are a lot more appreciable in realtime multiplayer than in reported offline laptime charts. Keeping it offline is good for the reasons you mentionned, but it does make for a significantly slower dev. time in comparison.

It's worth trying.


Why is a disparate racing class better than an even one? The FXR and XRR are meant to compete with the FZR, not be automaticaly left in the dust.

Regarding means of balancing: the more variables, the closer we'll get to ideal parity (even laptimes in any conditions). But braking endurance would need new modeling code in addition to the balancing work.
Having the FZR carry less fuel wouldn't work out for the best on at least a few tracks where pitting in takes a lot of time, e.g. the oval (unless it was meant in the sense that engine consumption - and consequently power - was reduced (i.e. starved) to keep the same mileage for a smaller tank). It would increase disparity on only certain tracks and require even more complex balancing. The solution needs to be as simple as possible, not complicated.

Ballast would mean performance adjustment not only on a per-car basis, but account for differences in each player as well.. obviously it serves as a negative handicap only to level gaps in performance that should be fixed in the first place, but it could help the regular non-test players in the mean time.

The cars shouldn't be balanced in a way which they all drive the same laptimes under the same conditions. I'd much rather see them balanced in different areas. For example endurance racing, one of them could be quick but have a slightly smaller tank giving that advantage to another car in long runs.

When S2 Alpha was first released I used to always use the UFR over the XFR. It gave me a slight disadvantage but I still won a lot of public races because I drove it well using its braking and cornering ability to keep ahead of the XFRs. Was brilliant fun IMO. That's the sort of balancing I'd rather see, not where the cars are on par with each other under every condition.

Keiran

Yes, I don't mean exactly the same laptimes, only in comparison to the present 1-2+ second gaps.

I'm hugely in favour of anything which gets these damn cars balanced. Find it fairly disappointing that the fxo has *always* been faster. Almost seems as if Scavier has an idea about what the cars specs should be to be balanced given perfect physics, and is unwilling to compromise on this for less than perfect physics...

I'm inclined to agree, it does feel as if car balancing is not a current priority - and yet it would make a huge difference to the playability of the game - *especially* in leagues. I think it's importance has been undervalued.