Cars and rules for 2026

[Duplode]

Is a fixed -100% coefficient sufficient to keep all the "cheat cars" at bay? I haven't tested them much so I am unsure how fast they really are.

Excellent question. The balanced bonus for PG Indy is somewhere near -30%, so let's round the coefficient to a nice 4/3. A -100% bonus amounts to a multiplier of 2, so the ratio between the driven lap times would be 2/3 for balance. The free/RH Default record with the Indy is around 30 s, while the time of a comparable lap with the Lada Niva Comeback is around 20 s. So yes, the fastest cheat cars might actually be competitive even with a -100% bonus!

(That said, if we were actually going to implement this, we might as well use -150%, or even have two tiers to distinguish cheat cars from the rest.)

[Matei]

Do consider posting replays at ZakStunts just for fun;

Read the end of message here... from which I should have actually translated more, but I can give some of my other messages as example. Anyway:

You don't have to care about scoreboard position or the car bonuses;

I don't care about scoreboard position or the car bonuses, I don't even care about the "races". Do I have to put links to the messages where I already wrote that?

just pick your favourite car out of the 16 and enjoy the ride.

I only have ~4 pairs of skates, but the rides are not very enjoyable because there's cars everywhere.

https://matei.one/osub.html

Clearly, these killing machines have no business being used anywhere there are people around,

If we're referring to favorite games, Stunts is on the second place, here:

https://matei.one/idxscr.html

And it's not actually Stunts. Stunts is not interesting as a game, but has some other things made around it, especially by Cas. Otherwise I wouln't pay any attention to it.

https://wiki.stunts.hu/wiki/Cas

Cas' most significant software contribution to the Stunts community is Bliss Track Editor,

No, his most significant software contribution is R5K_engine, which should be obvious. Besides, there was Track Blaster already and I hate "mouse-driven interfaces" anyway.

[Matei]

Another thing. From the messages here I come up with ideas.

https://forum.stunts.hu/index.php?topic=4482.msg98413#msg98413

Like in Stunts, but not in my game.

Because the framerate is not fixed, but I fixed it to n frames/second and in the new version the replay is now like in Stunts, i.e. the input controls are saved at each frame, instead of the positions. I tested, it works, the replays will be saved in files and I'll post the version here when it's ready. I will keep the replays like they are now in the main version though, but I will use the new type of replays in Skunks, because the replay feature available in previous versions is currently missing there. The same kind of things happen with Wayland, btw. Wayland is a very good protocol for graphics and who knows? Maybe 50...60 years from now it will become usable.

[Frieshansen]

First of all: I joined here in 2021 because I had been observing the competition for quite some time and found it to be just right, and I still do. Apart from a few minor changes, the basic pillars have remained the same, and as far as I'm concerned, that's fine. I particularly like the car podiums as an addition.

When it comes to car selection, I was reminded of RidgeRacer on the PS1. There is a small overview of the most important features for each car:
You cannot view this attachment.

I would love to do something like that for Stunts, but unfortunately I don't have the time. We could perhaps overlay several of these diagrams to see which characteristics a car would need to have to make it a balanced choice. Here's a video showing what it looks like in RidgeRacer, but I'm sure many of you are already familiar with it: https://www.youtube.com/watch?v=qrY2eOG-N5g

[Duplode]

When it comes to car selection, I was reminded of RidgeRacer on the PS1. There is a small overview of the most important features for each car:

This would be very fun to have! I remember having daydreamed with @CTG about a NFS-style display for our cars. Other neighbouring ideas: category and tier lists; Super Trunfo.

[Shoegazing Leo]

Melange + Lola Indy or McLaren Mp4

[HerrNove]

To avoid the PTB exploits mentioned by MiDiaN in the shoutroom (and by me some months ago), we could slightly tweak the bonus in this way:
* 2 points if a driver gets 528 PTB-hours in a single race without the help of the carryover
* 1½ points if a driver gets 528 PTB-hours summing those of the race and the carryover

In both cases, the carryover is reset.

[alanrotoi]

There are no exploits. There is a limit of 2 ptb per race.

[HerrNove]

There are no exploits. There is a limit of 2 ptb per race.

The limit does not guarantee the absence of exploits. In certain (rare) circumstances a driver could get more points by doing less.

[alanrotoi]

Really? We should fix that then.

[dreadnaut]

I think exploits are possible if you look at a single pipsqueak. But when you have a system of competing pipsqueaks, connivance notwithstanding, efforts balance out to avoid the extremes.

And because the "exploit" is publicly visible, it becomes part of the stregic game: "it looks like someone is holding back, ah, let me ruin their plan by doing this other thing!"

I'm interested in digging deeper though, because there are competition 'principles' that I think are important:

- strong pipsqueaks should have to make more difficult choices
- moves that can advantage a strong pipsqueaks should also lift the (not-immediate) followers

If there is an "exploit" that can advantage someone without giving something back to the lower scoreboard, that's something we should look into.

[Duplode]

Indeed, this is the same issue @HerrNove and me discussed here back in February. Here's a quick example. On the one hand, if someone gets 260 hours on ZCT292 and 400 hours on ZCT293, they will get PTB +0.5 on ZCT292, carry 128 hours to ZCT293 and get PTB +2 there (128 + 400 = 528), for a total of 2.5 points. On the other had, if they get 330 hours on ZCT292 and again on ZCT293 (same total hours across the two races, but evenly spread), they will have no carryover and get PTB +1 on each race, for a total of 2 points.

Now, while that is somewhat of a wrinkle in the system, I'm not really convinced of it being enough of a factor in practice to make a more complicated carryover rule worth the trouble. Fundamentally, under the current rules PTB +2 is very hard to achieve, even with carryover. Looking back at the first nine races of the season, a total of 396:01 earned in a single race (what you'd need to get PTB +2 with the maximum carryover of 131:59) was only reached three times (Alan on ZCT282, Argammon on ZCT285, and MiDiaN on ZCT289). That being so, taking advantage of the wrinkle to systematically reach 2.5 points every two races seems far from trivial under the current rules. (That said, if we're ever going to lower the thresholds to make PTB easier to reach, this issue will definitely need to be accounted for.)

(As for HerrNove's proposed solution, it has the merit of being a localised change with little risk of upsetting the balance of the system. However, it might be rather anticlimactic to fight all the way for the final 264 hours and only get 0.5 point in return. Also, I wonder if e.g. 10 hours of carryover should have the same penalty as 100 hours.) 

[Matei]

https://forum.stunts.hu/index.php?topic=4495.msg98527#msg98527

https://en.wikipedia.org/wiki/Freestyle_motocross

a panel of judges assigns each contestant a score based on a 100-point scale, looking for difficult tricks and variations over jumps.

[Argammon]

I should not have a say about the rules for 2026, at least not for those affecting the season's standing.

But I thought I would mention that I agree with @MiDiaN and @HerrNove that it feels strange that it is sometimes beneficial to voluntarily stay just below the 1-bonus-point threshold. However, I also agree with @Duplode that the effect  likely does not have much practical relevance over a whole season.

In any case, how could the system be "fixed"?

It would be an easy fix to either remove carryover altogether or to allow carryover for all hours. As hinted at  by @dreadnaut and @Duplode this would remove the strategic metagame almost entirely. In particular if all hours carry over one could as well just track the total position time over the whole season.

Alternatively, quoting @MiDiaN's idea here, it may make sense to only award the big pot if a pipsqueak obtains 528 hours without the help of carryover. I must say that I am leaning towards this idea. Two bonus points are massive, essentially meaning that a second place plus the bonus is better than a first place together with just 0.5 bonus points.