Great to see a reminder of how good STCC servers actually were too. It was at the end phase when I joined S2, but I did a few races on it as a newbie, it was actually the first server I entered after purchasing S2.

I just say, cargame.nl in nowadays is a lot better than CTRA was. I raced on Bump'n Jump server and sometimes on GTi server. CTRA was too demanding, like they thought they could own whole europe, while in cargame.nl, you are allowed to relax and drive for fun without any complications, which suits to me as best. Of course I respect the rules, at least I TRY so, while in CTRA it was impossible even to show the respect...

"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




Just my 2 cents of my experiences...


EDIT: NO need to say why rules are made of... I obeyed rules as much as possible


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

haha, hahaaahahahha... fools

(Sorry, makes me annoyed when I think about it, now where is my beer... ? --->)

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

You are quite clearly talking out of your arse, sorry.

CTRA blue flag etiquette was that the faster people had to make a pass, not that the slower people had to move out of the way.

The only exception to this was for mid-race joiners, who were expected to not interfere with the ongoing race.

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?

IIRC, mid race joiners (I did that a lot) got a chat message saying something like 'Make way for Bean0' and a waved blue flag message in the screen. I'm pretty sure nobody actually ever got banned for not obeying a blue flag rule and I never actually saw anyone dive off the road.

When I do get online every so often, it just depresses me that a populated server means an admin who has put down however may £ it is to rent a server, sat back and just used Airio as a crutch to run the server. You simply can't rely on technology to run a server for you.

CTRA was what....2005-2007 or 8? I joined in 05 and I distinctly remember joining CTRA and just doing my best Captain Slow impression.

As for the original question. No, it wouldn't help as the community's moved on from structured and pick up racing now to a more 'relaxed' style of racing. The demand may be there, but outside of a small niche, who would actually use a CTRA/STCC type of series?

Oh and Becky, can we still download the STCC anywhere?

CTRA had to be from 2007. I still saw stcc in late 2006 early 2007 i think.

Quote from DieKolkrabe :

Oh and Becky, can we still download the STCC anywhere?

As far as I know, the races are still on Youtube.

I had an appreciation for CTRA because it helped to weed out a lot of the idiots that you sometimes get on public servers.

However, it was far from perfect. I have to use myself as an example because it's all I really have to go on. I'm a constant mid-pack finisher. I sometimes make it into the top 1/3 of the field, but 99% of the time I'm towards the front of the middle. If you understand me. But I am also a very considerate driver. I brake early at T1 to avoid hitting the inevitable chaos (and usually get rear-ended for my troubles). I'd rather lose a place and avoid hitting someone than keep my own position and run them off the road. And it's really hard for an automated system to reward that kind of driving. The flag detection is always a bit useless so that can't be used as a signal of quality driving.

Unfortunately there's no way for an automated script to go "hey, he actually braked to make sure the idiot beside him didn't crash, let's give him a point reward". And that means people like myself were relegated to grind like a Korean MMO just to gain access to the cars we actually want to drive. That's not fair.

I wanted to race the TBO classes but at 1 or 2 points per race that would take literally thousands of races to get there. So until there's a way to accurately tell when someone is driving well (irrelevant of their finishing position), every system like this is going to be flawed.

Quote from DieKolkrabe :

When I do get online every so often, it just depresses me that a populated server means an admin who has put down however may £ it is to rent a server, sat back and just used Airio as a crutch to run the server. You simply can't rely on technology to run a server for you.

True...

So?

What exactly do you expect more then? Choose your words carefully

Quote from Torben :

Why was that auto-spectate feature invented? I think it came from the attitude that started to emerge on the servers.

Scenario a) Driver A is spinning. He thinks: "Darn, spinning, darn wall coming closer, darn, impact. Keep standing still until following drivers can pass the crash site safely. Hm.. I have serious damage, but I can make it back to the pits. .. Ok, wait until the track is free,....., alright, now I can re-enter the track." Driver A crawls back on track, stays off the racing line and crawls back to pits for repairs.

Scenario b) Driver A is spinning. He thinks: "Darn, spinning, darn wall coming closer, darn, impact. Hm, well I have serious damage, but the car still somewhat moves, lets get back to racing." Driver A crawls back on track and *boom*, a car approaching slams into him.


Which scenario do you think lead to the invention of auto-spectate? And which scenario is more likely to happen on public servers..

I honestly think it's our own (the racers) fault that the servers are what they are. No admin has the nerves to keep arguing with everyone about who's fault it was, who spun, who didn't slow down despite the yellow flag etc..
It seemed nobody cares about general racing/race track rules anymore. So mechanisms had to be invented that made dealing with this easier..

So what about the Trackmania colision system? Do we need this in LFS?

Seriously, LFS is a simulation, if you see a Yellow flag warning message, you just need to be prepared to slow down or stop your car, like IRL...

Quote from Bean0 :

As far as I know, the races are still on Youtube.

http://www.youtube.com/playlis ... 844CE9F4&feature=plcp

I'm going to see some of those videos seeing everybody say they were great

Quote from cargame.nl :

True...

So?

What exactly do you expect more then? Choose your words carefully

It's not unreasonable to expect a clean server that doesn't spam your screen by default is it?

Quote from FBUT Bishop :

So what about the Trackmania colision system? Do we need this in LFS?

Huh? Trackmania?

Did they changed something? In the first version its impossible to collide with each other because other players act like ghost cars

Quote from DieKolkrabe :

It's not unreasonable to expect a clean server that doesn't spam your screen by default is it?

=> -

If something really spams then its the LFS chat in general on busy servers.

.

Quote from cargame.nl :

Huh? Trackmania?

Did they changed something? In the first version its impossible to collide with each other because other players act like ghost cars

.

That's what i say, we don't need that system in LFS, an neither those collisions warning systems...

P.S: Damn, my english goes bad since CTRA is down... I don't speak in english anymore because i almost stop playing lfs since the dead of CTRA

Quote from cargame.nl :

Huh? Trackmania?

Did they changed something? In the first version its impossible to collide with each other because other players act like ghost cars



=> -

If something really spams then its the LFS chat in general on busy servers.

.

Don't know about that. I mean, I when someone enters into a server, I do not need to know about their LFS Experience, Rank, that their skin wasn't found on LFS world, when they last raced, how fast they were down a straight. That stuff is kinda spammy.

Quote from FBUT Bishop :

So what about the Trackmania colision system? Do we need this in LFS?

Seriously, LFS is a simulation, if you see a Yellow flag warning message, you just need to be prepared to slow down or stop your car, like IRL...

Yes, absolutely. I completely agree. (EDIT: Agree with the prepare when you see yellow flag, not the trackmania collision system, that is out of question not suitable for any simulation)
But people also need to realize that if they spin, be it their own fault or not, they have to stand still and wait for traffic to get by before turning around/reentering the track. Not doing so will most definately cause a collition, whether upcoming drivers slow down or not.

EDIT2: In the one example from earlier: Example; http://www.youtube.com/watch?v=fmRSpbvujs0 . Hello, spectate anyone? ..
If Piotr didn't reverse back on the ideal line the major crash would never have happend. All following cars could've gone by no problem. And in this situation preparing for stopping probably wouldn't have been enough when the yellow warning message appears.
A real life example: http://www.youtube.com/watch?f ... &v=3BD3xhX0TcE#t=894s
Here Ricardo Patrese gets spun around and has to sit and wait, let couple of cars pass, before backing up. That's the way to do it.
Yes you loose positions, tough luck, thats racing.

http://www.youtube.com/watch?v=MIZJfI2QjWY

That's what you mean?

Edit: http://www.youtube.com/watch?v ... NG5a4&feature=related Or that?

Quote from englishlord :

Don't know about that. I mean, I when someone enters into a server, I do not need to know about their LFS Experience, Rank, that their skin wasn't found on LFS world, when they last raced, how fast they were down a straight. That stuff is kinda spammy.

Why don't you remove all the messages in the AIRIO options?

Quote from Dajmin :

I had an appreciation for CTRA because it helped to weed out a lot of the idiots that you sometimes get on public servers.

However, it was far from perfect. I have to use myself as an example because it's all I really have to go on. I'm a constant mid-pack finisher. I sometimes make it into the top 1/3 of the field, but 99% of the time I'm towards the front of the middle. If you understand me. But I am also a very considerate driver. I brake early at T1 to avoid hitting the inevitable chaos (and usually get rear-ended for my troubles). I'd rather lose a place and avoid hitting someone than keep my own position and run them off the road. And it's really hard for an automated system to reward that kind of driving. The flag detection is always a bit useless so that can't be used as a signal of quality driving.

Unfortunately there's no way for an automated script to go "hey, he actually braked to make sure the idiot beside him didn't crash, let's give him a point reward". And that means people like myself were relegated to grind like a Korean MMO just to gain access to the cars we actually want to drive. That's not fair.

I wanted to race the TBO classes but at 1 or 2 points per race that would take literally thousands of races to get there. So until there's a way to accurately tell when someone is driving well (irrelevant of their finishing position), every system like this is going to be flawed.

Dajmin, you seem to be the same type of racer as me, not very quick but wanting clean mid-pack racing. I managed to get up the rankings to where I wanted to be with a bit of effort. Are automated systems and the people running them looking at the problem from the wrong end? Instead of looking for the wreckers should they be trying to find the decent racers in the pack. I'm sure when CTRA was running Sam and Becky said their aim was to improve racing standards to find new talent, it wasn't about penalising people. Shorty after the demise of CTRA I was posting in a thread with Sam and Becky (and other people) about the merits/possibility of automated lower rank tiers and moving on to higher ranks where admin should be a lot less onerous. Sam wasn't optimistic and had no interest in any form of automation.

Just off to look for that post.

Quote from FBUT Bishop :

url1
That's what you mean?

Edit: url2 Or that?

url1: holy sheite!
url2: yep, just like that

Quote from kart-36 :

CTRA was awesome and I would love to see something like that come back.

+1. Loved b'n'j...

""No automatic system can work perfectly, but does anyone think a system that monitors certain parameters that a driver must stay within could work? 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 posted this 2 and a half years ago in different times with people floundering around after the CTRA rug had been pulled from under their/our feet and rumours of a comeback in the air

Could this work ? Anyone interested?

I'll get me coat now and bugger off back to NetKar (10 racers online) and await the physics update and hope things racing-wise improve at the same time.

One last thing, Becky did say there was a correlation between yellow flag ratio and clean racers.

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 NigelY :

One last thing, Becky did say there was a correlation between yellow flag ratio and clean racers.

There is a definite relationship but there are a number of factors to consider.

When we introduced the statistic to the CTRA a lot of drivers became very scared of it, and that we may use the statistic to make judicial decisions - and that the statistic itself would be fundamentally flawed. Actually we just thought it was cool - we knew there would be concerns of that nature so we where very clear in establishing the precedent for not using the yellow flag ratio in our judicial reviews.

However what we discovered was that once the system had collected a bit of data there was an absolute correlation between a drivers yellow flag ratio and the drivers who caused issues.

At that time we used LFS' default yellow flag detection - and LFS is often a bit keen to get the yellow flag out. I did consider writing my own code, even a simple 3 second delay on counting it as an actual yellow might have been enough. Also there was a bug on Blackwood Reverse which triggered a yellow flag each lap (on discovery we actually disabled yellow flag tracking on that circuit).

So on the raw stats as LFS generated them about a 5% ratio was considered an excellent yellow flag ratio - which would be an alarming number in real racing! Most people getting reported had 25% or more. Which even by LFS standards is quite attrocious when you think about it, especially as we only tracked a maximum of 1 yellow flag per sector.

If memory serves the majority of drivers fell in the 8-12% range, including myself. I even at one point stopped racing because I was so recognised on the servers that I couldnt focus on racing and concentrate and ended up parked up and chatting to people, and this blew my yellow flag ratio!

I think what most people fear from an automated decision is that it may, on ocassion, generate an incorrect result. But what is often overlooked is that one event has no real bearing. An automated system which is only 95% accurate is actually more than sufficient if action is taken not against a single incident - but against a pattern of incidents.

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.

This rule will not provide a 100% accurate measurement of fault, however it should provide a "reasonably" accurate one, which whilst utterly incapable of taking action against a single incident would be sufficient to determine a pattern of behavior over time.

Provided a system is capable of reasonable accuracy, even in the absense of absolute accuracy, and that the data generated as a result is used in such a way as to respond to patterns of behavior and not absolutely applied to a single incident, then an automated system can work and a CTRA like system could return.

The rule I give above is a theory and is untested, but I believe it should work - albeit perhaps with some tweaking with live data and a prototype.

And before anyone asks, no I do not have a prototype ready! My personal situation is that as much as I would love to bring back a CTRA like project I do not think I could do it from a hobbyist position anymore.

col, it is simplistic (probably way too much). Struggling to explain, much like when I originally posted it .

This is about non racers learning how to race. I don't know about you, but when I started sim racing learning the racing line and getting up to a consistent pace was bloody hard. Dreams of overtaking another car cleanly didn't come into the equation (probably shouldn't have started with GLP). Out there in the real world they wouldn't let you race on a circuit until they thought you were good enough (I realise the implications are more serious). If new racers on a beginner server knew that they were being monitored, not necessarily for finishing in the top 3, but for consistency wouldn't some of those ratios make more sense?