S.E.T.H, if you have a problem with f mode, why don't you just f off ?

Last edited by col, Thu 16 Feb 2012, 14:26.

Quote from MoMo92i :

...the nearest we knows is atm Pacejka model (I would suggest a bigger google research about it)

I suggest you do some more google research yourself.

Look up "brush tyre models" - that's what the current LFS tyre physics is a variation of - not Pacejka, which I understand is getting a bit long in the tooth.

Scawen discusses current LFS tyre approach here

Col

"What's the first thing you look at?"

That depends on whether she's facing towards me, or the other way...

Quote from Becky Rose :

Blue 2 was an idea I had to take what was good about the CTRA and strip out what didnt work - namely the reporting...

The idea was that you could connect every server in LFS to it (at the server owners discretion) and gain features comparable to the CTRA servers and it would operate a CTRA like web page and statistical system - you could compete in numerous championships from STD racing through to clean driver cups, endurance cups and so on to earn badges.

Then each week server owners could apply to be a "Pro" server for a week - one for tin tops, one for open wheelers - and these would be restricted to drivers who achieved the pre-requisite number of badges the previous week. The pro racers would have some cool rewards and because of their restricted nature would likely have the cleanest racing. In return for this "free" traffic to their servers the pro server operators would be required to maintain minimum standards of admining or have to "take a break" from the pro program, and because the pro servers changed each week no one operator would be swamped with a backlog - they apply for a period of one week when they have a team willing to stand by and help out for that week.

That's a great idea. It seems to me that it is also exactly the sort of thing the devs could and should be working on.
Aside from creating a better racing environment, it would also help in pulling the community together, having these server specific web pages as part of the LFS site would be cool. Linking data from the racing system to the main website and the forum... could have user review of servers - like amazon

Quote from Becky Rose :

And in my view achieving 95% accuracy is achievable with a fairly simple rule: In the event of a collision the car which was behind 3 seconds before the contact is at fault in all circumstances when the car in front maintained a normal line.

The collision is worthy of recording when one or more cars involved in the collision complete the section of track (an area much smaller than a split sector - say, 1 corner either side of the incident corner) slower than their usual delta time plus a percentage threshold that would have to be determined with some testing.

If the car infront did not maintain line then determine the relative overlap of the cars at the point the vehicles changed direction to determine fault.

Interesting idea. What about the first few corners, particularly where the track is wide?
Many incidents involving unacceptable driving occur in the first few corners. There are lots of cars side by side. There is no 'normal line'. People get shunted into others. People deviate from their line to avoid being shunted... etc.
Do you ignore starts ? maybe hope that folks who are naughty at the start will be bad elsewhere and eventually get punished for that reason...

I guess it can still work as long as you start at the back and build some sort of 'blame tree' so if someone has been shunted themselves prior to causing an accident, you don't blame them - you pass the blame back.

My main worry would be that with 95% accuracy, you still might get some poor unlucky racer who just gets a string of false positives, then gets banned unfairly - sure for each individual, the chance is low, but the more races and racers there are, the closer the probablility gets to 1 that it will happen.
And it may work out that a few people due to racing style (not unfair, just slightly unusual) set off a lot more false positives due to blind spots in the algorithm...

Quote from NigelY :

If you could obtain percentages for the following could they be used as a pre-filter on the way to moderated higher tier servers.


% races started to finished
% yellow flags
% time spent on racing line (can InSim use the green line)
% lap time
% time spent pressing "restart" after race is under way (after spectating)
% time spent chatting during race (after spectating)
% time spent off tarmac""

I think this is way too simplistic. e.g. If someone is in lots of close battles, they could be a great racer, but spend a lot of time off the racing line. Yellow flags are also flawed IMO...

I think that the only sensible way to use automation is to create and train an advanced AI that uses a large set of data input. You would have to spend a lot of time 'teaching' it. And in the end, it would probably still only be useful as an aide.

If it used data including but not limited to: position, speed, rotation, control inputs, tyre grip state for all cars within a set distance of an incident, over the time from x seconds prior to x seconds after the incident. Combined with detailed info about the track, and some sort of 'understanding' of physics and what a driver can be aware of.... you might be able to use an advanced neural net combined with heuristics to over time build up enough 'knowledge' in the system to get a reasonable percentage of decisions correct....
You would have to present the system with sets of data and tell it who was to blame, who was in the wrong and who was an innocent party....
Maybe after a few thousand inputs or more, you might start getting somewhere ?

Anyone have any experience in this area ?

Seems like a large and difficult project

Quote from UnknownMaster21 :

"You were too late to respect blue flag"

CTRA: you are banned (yes, banned for being 1 second too late)
cargame.nl: No problem, in race, no one is perfect, at least you tried though

Didn't happen to me once! maybe because I always respected blue flags and other drivers ?
I guess you would only be complaining about this if you kept ignoring blue flags until it was too late ? Lack of respect for others ?

Quote :

EDIT2: and CTRA, that server was full of clowns who thought they are like Chuck Norris. You make a little kiss on race ---> you are a crasher

Sounds like the reasoning of a crasher to me - why else would you be annoyed about a system that effectively stopped crashers?
For the same reason that IRL criminals hate the law?

Last edited by col, Thu 2 Feb 2012, 12:46.

Quote from cargame.nl :

Exactly...
You don't want all this constant 'being in between', as admin, because it-never-ends.

Admin makes a decision. The End!

Quote :

Scawen updated some InSim packets at my request to create auto crash detection and who's fault it might be... Upon this date nobody uses this section of the InSim system.

Wow, that is a really difficult programming challenge. To write AI powerful enough to make a correct and fair decision in all but the simplest of collisions would be a massive achievement. How did it work ?
Or was it just to help the admin rather than to replace?

Quote :

I think it's the philosophy to hand as much as possible to the community and let them (thats us), figure it out.

That's my understanding as well. Many years ago, devs made it clear that they didn't want to get involved in this side of things. I hope they change their minds about this in the future.

On a positive note, if you look at the real world, there are countless clubs in various sports and pastimes where amateurs give their time freely to work as chairmen, treasurers, referees, coaches etc. It is not unusual for folks to take on roles of responsibility and give a significant amount of time unpaid. It's all down to perceived value within the community.

If suitable organisational structure is developed, and there is community support, I think getting people to volunteer as admins is not an insurmountable problem.

Quote from cargame.nl :

I recreated exactly that, but nobody gives a ** ..

Mostly because the explanation is missing and a shiny website, I guess. But well, what does it change.

Well, as you just pointed out, having the 'explanation' and the website changes whether people care about having a points system.

Quote :

I am currently in doubt it would attract more racers.

Everywhere in the world where there is sporting competition that people get excited about, there is some sort of points system where results and standings are posted publicly. Otherwise there would be no competitors... funny that.

I think that the fancy (very professional and high quality) website was a big part of the success of CTRA, and the reporting was the other major thing it had going.
Without a points system with achievements with web presence, and without a good reporting and moderating system, whats left? a server! so what.

If the work load for mods is too high, maybe the reporting system could be tweaked to avoid having most of the 'racing incidents' reported. e.g. if someone reports an incident that is not a clear infringement, then they themselves are penalised - points fine, temp ban etc.? That way you'd get all the wreckers, but the other stuff folks would deal with themselves.

As far as getting people to be mods, if you have a nice pro website and a big user base, then you are more likely to get volunteers to - there will be kudos if it's a core part of the community and there is public recognition of their importance. It works for forums.
People like to feel important, powerful and respected within a community. If the value of the service you are providing is high enough, you will get volunteers. But it would have to be a damn sight more than just another scripted server.
That's another reason why I think it needs the devs behind it - if it's THE official system, then people will want to be mods.

Quote from cargame.nl :

It's more interesting to research why this CTRA system closed then compare differences.. Which you can't because it was in different times. There was less choice in online multiplayer racing. Far less...

Much easier to ban/kick people if the pool is bigger to begin with, you following me?

But any idea why CTRA shutdown?

I know it.

IIRC the folks individuals who created and ran CTRA fell out in a big way, and that soured the whole endeavor. Unfortunately, 'office politics' are always going to be a problem for something like CTRA. Too much work for 1 person, you need a Team. If you have a team, you need to have a leader, because folks don't always agree. Finally, if someone has a different opinion, and you tell them to 'suck it up' because you are in charge, then you better be paying them to take that shit from you .

So in the end, it isn't going to last without funding from somewhere.
Which is why I feel it can't work without support of some kind from the devs.
Maybe they develop the system and the community moderates it? That would get around the problem of breakups due to policy disagreements...
Alternatively (but much less likely) a third party develops and runs the system and is funded by the Devs.
Or Devs provide hosting, integration into LFS official website and 'official' status, giving a 3rd party a better chance of making enough money through Advertising etc. to fund the project ?
I'm no business expert, but there must be a way for something like this to work?

Quote from tristancliffe :

Where's the option "I used CTRA, hated it, and raced only on non-CTRA servers"?

sorry Tristan, you're right, I should have had a more negative option.
Is it possible to add another in there ?

Quote from nigely :

ctra worked for me because it was an informal but disciplined way to race. No requirement to join a team, no commitments to race at a certain time on a certain day.

+1

Quote from MoMo92i :

What are you planning exactly? Something even bigger than CTRA? Gimme more info I really miss CTRA since I was quite new in LFS when it stopped.

Didn't mean to get your hopes up. I have no plans to develop something like this. I think it would only work if it was official with the devs involved - collapse of the original CTRA proved that.

Quote from boothy :

Ah CTRA, was that the system where you had to drive a lot to get points to move up from the entry level servers, to Race 2/SS2 which were mostly empty, to SS3 which was totally empty or Race 3 which you needed even more points just to drive an NGT?

The best racing I ever had was on the second tier CTRA servers. I remember it being difficult to get on the server sometimes due to popularity.
Third tier was usually empty, partly because CTRA was still young so not enough people had enough points, and partly because a lot of folk felt that the best racing was in the turbos

Quote from Flame CZE :

EQ Worry has also tried something bigger but no luck whatsoever, sadly.

I was thinking more of something the devs might be involved to some degree - assuming there was enough interest.
I'm not interested particularly in 'teams'. CTRA wasn't about teams getting together either for a big inter-team project either.
I'm just thinking about the fact that there's not much structure to LFS as a game. Just go on-line and try to find a sever.

I don't play online these days, but I understand that its hard to find good racing sometimes.

I reckon if there was an official future update that included a system similar to CTRA, then the official update would bring lots of racers back online, and the structured system with achievements and tiered racing would hold a lot of peoples interest, so activity wouldn't drop as fast.

Better for everyone 'cept cruisers

As I mentioned in another thread recently, I think that CTRA made such a huge difference to the quality of the online racing experience in LFS that many folks lost interest (to some degree) when it closed.
For me, of all the things that could be added to LFS that might make it better as a gaming experience #1 is a new CTRA type system ideally created or officially supported by the devs.
So I figured a poll would be interesting, to see what other folks think about this.
I've split the selection up between folks who used the original CTRA, and those who didn't.

If you don't know what CTRA is, please do some research to get an idea how it worked before voting

cheers

Col

Last edited by col, Mon 30 Jan 2012, 20:45.

Quote from Dandy Dust :

@CheerioDM:
actually you are right with what you discribe:
Even in our team it was obvious many times, that newcomers and such drivers (like me) have never made it to the top three (lets say) got frustrated and left LFS for ever.

This is a very good point. In a true simulation of motor racing with pickup racing and no structured hierarchy, only the more talented or very hard working drivers will have much chance of winning.

Newbies and folks with full time jobs or other similar commitments are going to be at a disadvantage compared with layabout students and schoolchildren .

I haven't played for a long time because I currently don't have a PC. I used to be pretty fast - no alien - but I usually had a chance. Last time I made a 'comeback', I was vote banned from a few servers "too fast, don't recognise the name, must be a cheat" seemed to be the reasoning, or it might have just been vindictive - "he's winning, so lets ban him". That's the flip-side of the frustration you describe.

The only time this problem of ability levels was sorted was when the CTRA was in action. IMO the Devs really will have to work on this side of things a bit harder. Maybe it's something Victor could work on as an expansion of the website side of things ?. CTRA really did turn a great sim into a great gaming experience for people at all levels.

LFS needs a CTRA style tiered competition system to give everyone a better experience, and it has to be official for it to work long term. We all know that a lot of people have stopped playing LFS. I think that for many it's not because of any weakness in the sim, or lack of content, but because CTRA closed, and it's never been the same since that day !

Col

Quote from troy :

...There are some key components missing to catapult it in new spheres though, dynamic lightning (day/night), dynamic ambient temperature, wear on mechanical parts, more server slots.

I probably forgot something important but these would be my "must have" additions to stop the 24h of hotlapping, get some of the key endurance racing aspects in the sim and make it stand out of the strong competition there is these days in the simracing market.

You missed my No1 missing feature (and I'd think it's one Scawen want's and intends): Live Track! ie. dynamic changes to track surface and grip caused by e.g. rubber laid down, marbles, dirt, temp changes, puddles, drying lines (no point in having 'weather' until there is a good dynamic track surface model), ruts forming in rallyx.
IMO this is by far the biggest missing component in the sim and will be a major challenge - particularly in terms of netcode. I think/hope it's the next 'big' development project for Scawen after next gen tyre physics has been completed. I'm sure he'll do lots of smaller bits and pieces in between.

Col

Quote from Hoshimodo :

Would you care to explain this theory in detail please

I think he means, if the devs cave in to community pressure and release if before it's ready, it won't be as good and won't have the same lasting appeal.
So he thinks they should ignore all the whiners and haters and take however long it takes to get it right. Then it will be so good that we'll still be here 10 years after that waiting for the next update.

I agree with him.

Quote from [Audi TT] :

There is progress, but I want to see the grid tires 0.6B and compare it with the picture in 1 post.

There is no similar 'grid tires' for 0.6B.

These wire frame graphics are from a non-realtime model Scawen has developed to help him produce data for testing and developing the realtime tyre model.
I think?? he only started this approach after the tyre phisics in 0.6B were already done. So there can be no comparison, because there is no equivalent non-realtime model for the 0.6B physics.

Last edited by col, Sun 8 Jan 2012, 19:48.

OK, here's my statement of development:

Scawen is working on a new version of LFS. When it's done it will be released. Until then, some will whine and complain. Some will defend the devs. Some will throw their toys out of their prams. Some will wait patiently. Some will leave. Some will join. Some will troll. Some will feed the troll.

Last edited by col, Sun 8 Jan 2012, 02:29.

Quote from undertaker00 :

baw was great

what about biggie ?

no one reach that level

different eras, different aliens...

Macest, MATH...

Quote from cargame.nl :

This is starting to sound like these guys which say;

it's 21 Jun 2009 end of the world now
it's 15 Jul 2010 really end of the world now
11 Mar 2011 really endend of the world now
25 dec 2012 really endendend of the world NOW

And one day they will be right!
And maybe one day we will be too... unless they are first!?
..........

The end of the beta is nigh!
Repent ye haters
REPENT ye or burn (badly modeled rubber) for eternity in rFactor
Repent ye and be saved by the power of the physics update
For Scawen has spoken and the word is "soon we hope"
And so it shall be that "soon" the forum will be cleansed of whiners
Cleansed in a mighty flood of download mirrors
And we shall see the light
And it will be green.

I wonder if we didn't get a Christmas prezzy because Scavier are aiming to do something on LFS's 10th birthday?
We certainly need to do something for the 10th anniversary, even if the dev's don't.
Does anyone know when the first demo was made publicly available?

Quote from rageshgr :

That is a whole new discussion. Actually yours is a question which has been in mind too. My question is, how come the tyre physics limitations did not surface during the development of the 5-6 real cars, but it surfaced during the development of Sirocco. Probably it was in a slow road car like Sirocco, with ESP perhaps that the limitations of the current tire physics was more pronounced and exposed. Probably they started using some new testing techniques which were never used before, which exposed the limitations of the current tyre physics. The development of Sirocco, is waiting for new tyre physics and that is the main reason why it is getting late.

The main problem I think is that in the interest of realism (and maybe to allow for realistic ESP?), Scawen wanted to limit the setup options of the Sirocco to realistic parameters. With a setup like the real car, the sim didn't behave correctly.
All the existing cars in LFS have many setup variables and wide setting ranges in each, so it was possible to make the cars more realistic/manageable by using less realistic setup options. The intent in the future is to restrict setup options to more realistic values.

Col