PieceOfMind

What are months and years in the game? From what I've read, months sound like they are the same as days. If so, is a year 12 months (12 days)?

What? The technical limitations are going to be based on the real numbers, not the fudged ones. Just because we've now learned there are fudged population numbers doesn't remove those limitations. I'm no fanboy, quite the opposite, but I appreciate there are limitations for an engine which wants to model every sim individually (ignoring the fudge). 25,000 or so is a lot to simulate, even for modern computers.

Looks to me more like a bug with the road network than with the traffic AI. Actually I'm not even sure what people mean when they say AI. The game doesn't have AI, just basic programmed behaviours. It looks as if this sort of thing can happen after a few alterations to the intersection are made, and maybe there is still a phantom road in slightly the wrong spot or something. Does deleting the entire intersection and some of the road around it, and then building it anew help anything? Bear in mind this is all from someone who doesn't have the game, so going on gut instinct.

While probably a placeholder, the $15 pricetag could be even more cunning. An anchor is a powerful tool in marketing.

I don't think that's right. If you watch carefully, it is the moment that 4-way intersection is created that suddenly all the traffic starts going the other route. At that moment, the right turn for that intersection is a deadend. It's almost as if putting in another road forced an update of the entire network and then all of a sudden, agents realised the new route. Put another way, before that 4-way is created, vehicles are still piling onto the dirt road whenever they can. As soon as the 4-way is created, yet no new route is created, no one takes the dirt road anymore. Doesn't seem logical to me... Hmm, youre right, that does look pretty weird. Not much else i can think of that could be causing that then. They did say theyre still working on it, so lets hope the actual fix is even better than this video. Still, its an improvement over the current situation. I agree. I'm just a bit surprised that no one has mentioned any problem with it. I guess it's because the first implementation was so bad that, heck, doing anything to that code probably would have improved it. Or people are so desparate for any sign of the issues being addressed they'll take anything. Anyway, it's better, but without seeing it in action in any more than one specific test case (where even there it was still a bit odd) it's too early to say whether it's good. By the way, if I seem rather pedantic about this, I guess it's because this aspect of the game is the one I must be most interested in. Making an efficient path-finding algorithm that produces beleivable behaviours for agents is something I think must be a great challenge for the devs, and whatever solution they end up at will be interesting to consider.

I don't think that's right. If you watch carefully, it is the moment that 4-way intersection is created that suddenly all the traffic starts going the other route. At that moment, the right turn for that intersection is a deadend. It's almost as if putting in another road forced an update of the entire network and then all of a sudden, agents realised the new route. Put another way, before that 4-way is created, vehicles are still piling onto the dirt road whenever they can. As soon as the 4-way is created, yet no new route is created, no one takes the dirt road anymore. Doesn't seem logical to me...

I made this comment on another forum... Regarding the poor traffic path-finding, my guess is that it was a bug. They probably have coded a factor in for congestion but maybe late in development they broke it while fixing something else, and then there wasn't enough testing after that. The reason I think it was a bug is that Guillaume Pierre seemed genuinely confused at that "abysmal pathfinding" youtube video and said something like "but they should be avoiding congestion". Looking at Pierre's video again though, I'm feeling even less confident it's a good fix. It seems that even after the new road is made, and some cars have headed for it, still many others go onto the already crammed dirt road. I guess in normal cities these sorts of extreme cases will be rare, and easily fixed when they do arise. Still, once the new route is discovered, why isn't it that basically every agent should start going the alternative route? For some weird reason, it's only once he makes the intersection 4-way that suddenly pretty much all the cars start taking the alternate route. Hopefully it's a work in progress and he's addressing that. It looks like the cars that are originally taking the alternative route are only doing so because they won't fit. So maybe he's put in a logical step like: If can't get across intersection due to traffic buildup, turn left or right randomly and seek a new route. But then I can't explain after the 4-way intersection imrpoves it all. Has he elaborated anywhere what the new logic actually is? Or has anyone done much testing of this yet (I hear it is implemented on the test server?) I don't own the game :/

Gosh, your posts are consistently in a higher plane. Always enlightening to read. Good stuff.

Hey guys, I'm new here, but thought I'd share this graph with you. Fairly self explanatory... I just find it easier to visualise this way.