Sims Bypassing Potential Path

[infamousjbe 592]

Is there a possible reason why the Sims would bypass the SAM road cut through in favor of the  residential road.

 

I've checked the draw paths to make sure the road is usable, and indeed the paths look sound to me. Just perplexed why they would chose not to use it when it's obviously a shorter route to their destination.

 

Are the speeds on the SAM roads different than the default roads? This is the only reason that seems to make sense as a possible reason. 

 

[fefenc 77]

I don't think so, try to do this street without these curves. It's obvious that the street without curves is the shortest way to the destination.

[infamousjbe 592]

I could see that as a possibility. Though to drive by the SAM by 3 grid squares, just to have to back track 5 grid squares to get to the same point it would take them on the curved road is borderline for a reason. The curves don't seem like it would add so much to the travel distance that the straight street would provide a better route. 

 

Plus, on the SAM street there is no traffic which isn't the case on the street lined with homes. So the traffic alone on the regular road should effect the travel time over the SAM road with no traffic at all.

[Ro-Bo 183]

Roads has higher speed than streets.

 

speed

30 - street
50 - road, avenue
75 - owr
150 - highway

140 - railway
105 - subway, el. rail (glr) --> or is it 115?
225 - monorail

[Indiana Joe 1,750]

Yes, with the roads on the left and right, the overall route is slightly faster.  The curvy gravel street route is only 2 tiles shorter than the other way to begin with.

[APSMS 1,892]

It's also worth noting that the diagonal is a lie, and so the curvy street is around 5-6 tiles longer than the non curvy route (between the roads, anyway). When traffic get high enough to exceed capacity, then it the simulator may start using the other road, but traffic doesn't actually slow down network speeds until congestion appears. This is evident with the NAM intersection behavior, especially around stoplights.

[JJogurt 18]

With NAM installed Sims take the fastest route, not the shortest one. The road offers higher speed limit than the street and is therefore faster.

[infamousjbe 592]

Roads has higher speed than streets.

 

speed

30 - street

50 - road, avenue

75 - owr

150 - highway

140 - railway

105 - subway, el. rail (glr) --> or is it 115?

225 - monorail

Isn't SAM just an overlay texture to the normal street? So wouldn't both have the same speed?  Not the main roads, just the two roads cutting across.

 

 

Yes, with the roads on the left and right, the overall route is slightly faster.  The curvy gravel street route is only 2 tiles shorter than the other way to begin with.

By normal standards, wouldn't even 2 grid spaces shorter should make for a faster, more efficient street. In theory obviously, since they want to snub the SAM for the regular street.

 

 

It's also worth noting that the diagonal is a lie, and so the curvy street is around 5-6 tiles longer than the non curvy route (between the roads, anyway). When traffic get high enough to exceed capacity, then it the simulator may start using the other road, but traffic doesn't actually slow down network speeds until congestion appears. This is evident with the NAM intersection behavior, especially around stoplights.

This kind of sheds some light on the situation.

 

By lie, you mean that there aren't normally diagonal streets by default, so does that make the diagonals become longer in mathematics more than visually? Like for each grid square with a diagonal street essentially becomes 1.5 or 2 grid squares lenght to the travel path?  If so, then that definitely would make the Sims chose the normal over the SAM.

With NAM installed Sims take the fastest route, not the shortest one. The road offers higher speed limit than the street and is therefore faster.

The cross route the Sims are taking are streets which is the same as the SAM. Or at least that's what I believed cause it takes the street to drag SAM roads. 

 

It's not Road to Road to Road from start to finish. It's Road to Street to Road. 

 

Just to further clarify. In the image, the only Roads are the paths N/S, the W/E routes are Streets. 

[Indiana Joe 1,750]

SAM streets are only cosmetic, you are correct that the speeds and capacities are the same.  Diagonal streets don't matter, I'm fairly certain the pathfinder counts tiles.  The reason why the bottom route is faster:

 

The top route, from intersection to intersection, is 25 street tiles.

 

The bottom route is 10 road tiles and 17 street tiles.

 

There are 62.5 tiles in a kilometer.  A vehicle going 30 kph traverses this distance in 2 minutes, and goes 25 tiles in 48 seconds.

 

A vehicle going 50 kph goes 10 tiles in 11.52 seconds.  Going 30 kph on the street, it takes 32.64 seconds.  Add these and you get 44.16 seconds.

[infamousjbe 592]

SAM streets are only cosmetic, you are correct that the speeds and capacities are the same.  Diagonal streets don't matter, I'm fairly certain the pathfinder counts tiles.  The reason why the bottom route is faster:

 

The top route, from intersection to intersection, is 25 street tiles.

 

The bottom route is 10 road tiles and 17 street tiles.

 

There are 62.5 tiles in a kilometer.  A vehicle going 30 kph traverses this distance in 2 minutes, and goes 25 tiles in 48 seconds.

 

A vehicle going 50 kph goes 10 tiles in 11.52 seconds.  Going 30 kph on the street, it takes 32.64 seconds.  Add these and you get 44.16 seconds.

Damnit Joe, my brain doesn't work like that. lol....  I know you make sense in your math, but math is not my strong suite. But I certain appreciate the facts behind the math.   At least I know your not just blowing opinionated smoke.

 

So what it boils down too, is that the curve in the SAM street really does throw it off by just so much. So if I just made the curve of the SAM street cut across the 3 grid squares just below where it is now (just before the S curve, going right to left) probably would be all the difference in the desirability of choosing the SAM over the street. 

[Indiana Joe 1,750]

It's just ratios; this is pre-algebra!  

 

You would have to make the path of the SAM street at least 2 tiles shorter.

[Ganaram Inukshuk 906]

Things get more complicated when you take into account that as a network gets more used, its speed begins to decrease. The speeds you see are "idealised" speeds; I don't know what specifically, but it's either when the network has no use whatsoever or when it has very little use, but they could be even faster when they're empty. (I need to remind myself...)

[zimmie 521]

I didn't even realize that you could create curves like that with the SAM streets.

[Kuewr665 40]

They are part of the Rural Roads feature included in the Network Addon Mod.

[infamousjbe 592]

Things get more complicated when you take into account that as a network gets more used, its speed begins to decrease. The speeds you see are "idealised" speeds; I don't know what specifically, but it's either when the network has no use whatsoever or when it has very little use, but they could be even faster when they're empty. (I need to remind myself...

This is another reason why I was so perplexed by the situation. I figured even if the SAM path had a couple more tiles length, the traffic alone on the normal street lined with homes would create traffic to make that route undesirable. Guess it's not congested enough to make the SAM more efficient. There is absolutely no traffic on the SAM street which.

 

I didn't even realize that you could create curves like that with the SAM streets.

 

As Kuewr665 said, look for the rural roads in the NAM. 

 

Be careful with the small 90 degree SAM dirt road. It has a broken path that you need to patch up or through traffic won't work. Only pedestrians will be able to use it, without the patch.

 

I would post the patch here if I knew how to upload files to a forum post. However, you can try and search for it by it's name or part of it.

zzz-Rural_Roads_SAM_Patch_12282010

 

Make sure to place it in the root Doc/SC4/Plugins folder to override the NAM file that's broken. 

[Kuewr665 40]

I can't find it. I also checked my Plugins folder.

[infamousjbe 592]

I can't find it. I also checked my Plugins folder.

It won't be readily available on the STEX or in the NAM install. Had to grab it from a post in a thread over on SC4Devotion. 

 

Tracked it down for ya

https://www.sc4devotion.com/forums/index.php?topic=1617.msg359036;topicseen#msg359036

 

Here's the thread I created for it with an image of the broken path, just so you know exactly what I was speaking of.

[Kuewr665 40]

Ah, that's from 2010. It's likely I installed it when it was released, and I think the mod was updated since.

Huh, so it wasn't. Guess I'll update it. Thanks!

[zimmie 521]

Awesome. Thanks for the info, greatly appreciated. It's something that I've wanted to be able to do for a long time and didn't even know it was already possible.

[Tarkus 6,932]

Things get more complicated when you take into account that as a network gets more used, its speed begins to decrease. The speeds you see are "idealised" speeds; I don't know what specifically, but it's either when the network has no use whatsoever or when it has very little use, but they could be even faster when they're empty. (I need to remind myself...)

 

They are indeed faster than the listed catalog speed in the simulator exemplar.  The listed speed is actually the speed at 100% of catalog capacity.  The Congestion vs. Speed property determines this.  These values are essentially points along the overall trajectory, and if you're at 10% of capacity, the actual speed factor being used will be somewhere in between 1.3 and 1.15.

 

% of Capacity          Speed Factor

    0%                        1.3     * catalog speed

  45%                        1.15   * catalog speed

100%                        1.0     * catalog speed

150%                        0.875 * catalog speed

200%                        0.65   * catalog speed

250%                        0.3     * catalog speed

 

So, if we were to take the Road network into consideration, this is how the breakdown would actually look, if you're using the Medium simulator values (where Road has a catalog capacity of 4000):

 

Volume                    Speed

       0                        65km/h (40.3mph)

 1800                        57.5km/h (35.63mph)

 4000                        50km/h (31mph)

 6000                        43.63km/h (27.125mph)

 8000                        32.5km/h (20.15mph)

10000                       15km/h (9.3mph)

 

The default Maxis settings for this property use a maximum speed factor of 1.0, covering the entire range from 0-100% of capacity, so there's no real advantage to avoiding moderate congestion.  Beyond the way tuned-down pathfinding heuristic value, this is part of why the Maxis settings cause the tendency for vehicles to glom onto Streets, when they're the shortest path.

 

Using a value >1.0 on the low end of the percent of capacity scale is also what enables the phenomenon known as "spreading" on multi-tile networks with crossover paths--e.g. the wider RHW and NWM networks.  If one were to build an RHW-8S, for instance, using a simulator where 1.0 is the highest multiplier on the Congestion vs. Speed property, what happens is that traffic fills up only the row of tiles directly next to any entrance ramps, and vehicles stay strictly to the right lane (in RHD), not using 6 lanes of the highway, until those right lanes jam up, and then the traffic starts to use the full highway.  With "spreading", the traffic sees switching to a less crowded lane as quicker, and the whole highway gets even usage, rather than the right lane traffic jam.

 

It's also worth noting that it is not possible to specify a speed multiplier in the Congestion vs. Speed property less than 0.3, without completely breaking the game's Congestion Data View and causing other bizarre things.  The old "Simulator A" settings actually specified values less than 0.3, causing congestion to severely underreport. 

 

-Tarkus

 

[A Nonny Moose 8,261]

Taking the diminution of the speed due to traffic conditions, this could make a very nice calculus problem.  God, I am glad I don't teach that anymore.  This is the kind of thing that makes simulations fun.