SC4 Devotion Forum Archives

This is a thread that is made for a discussion started in the HSRP Thread and should have been put into its own thread.  All posts from the original discussion will be moved here.

Originally Written by: Cogeo

Andreas, the station looks nice, but I have some points:
- HSR stations (esp big ones) are in many cases terminals, not through stations. So I would love to see this implemented as a terminal station too. A small modification to the BAT is needed (put something like a wall at the one end). You would end up with three BATs/stations, a through station, a terminal-track-at-left station and a terminal-track-at-left station.
- Some mechanism must be devised to keep sims from using HSR for short hops, ie as a commuter train, which would be totally unrealistic. Ideally HSR should be used for intercity travel only. Maybe setting Switch Entry Cost to a quite high value would do this (total commute time wouldn't suffer, because to the much shorter travel time). We could also issue this suggestion for all HSR stations and not just leave it to the players and track layout; otherwise they would be tempted to lay HSR track for intra-city travel, which would be crap.
- Not to mention, I would like to see a small modification of this station(s), for GHSR (I can't see it clearly in the pic, forgive me if it's already).

Originally Written by: Andreas

Well, I didn't make the station, I only updated the stats a bit. ;) It's the old HSR station that was released some two years ago, when the idea of the HSR project came up for the first time. I'll definitely set the Transit Switch Entry Cost to a high value (according to jplumbley's formula, it would be 2.9), and currently, it acts as a transit hub between HSR, el-rail, (heavy) rail and bus. Needless to say, only one or two of those stations should be used in your cities, with the HSR part acting as regional connection, while the rail and el-rail parts should be connected to your local commuter network. I haven't tested the HSR so far, but now that we have a functional beta version, I can try to plop this in my region and see how it performs.

Originally Written by: Cogeo

Quote from: Andreas on March 29, 2008, 07:14:21 AM
I'll definitely set the Transit Switch Entry Cost to a high value (according to jplumbley's formula, it would be 2.9), and currently, it acts as a transit hub between HSR, el-rail, (heavy) rail and bus.

I'm afraid a value of 2.9 would be way too high. Indeed how does this come out? A value high enough so that pedestrians don't use it is a shortcut? It's rather flawed. I think even a much lower value, eg 1.0, would be enough to render the station completely unusable. I was talking about a value like 0.20 or 0.30. Anyways, testing could tell what the optimal value would be.

Originally Written by: Andreas

@cogeo: The value is calculated from a formula that jplumbley came up with along with his and mott's findings about TE lots. I don't know enough about that matter, so I can't say how it will affect the pathfinding engine.

This has become an interesting thread...  Maybe I should weigh in my two cents.
____________________________________________________________________

Lets start with the Traffic Switch Entry Cost:

OK.. from looking at the picture of the lot.  This lot *does* or *did* have a switch from each Traffic Type to Pedestrian Traffic, as well as, most likely a Pedestrian Traffic to Pedestrian Traffic Switch Point.  This clearly would show that the slowest Traffic Type that is using the lot is in fact Pedestrian Traffic.

Now, we can look at this from a few different view points.  Let's start with looking at where the intended Traffic Switch Entries are.

If you look at the picture of the lot, clearly, the North and the South side are both designed to be the terrances for the station.  These are pickup/dropoff areas.  This means if the Traffic Switch Points are setup so that anything the converts to Pedestrian Traffic can exit on the North or the South side, then we have created a spot where the Sim can jump from the North-West Corner of the lot to the South East corner of the lot.  This means that this is the maximum distance a Sim can jump through the lot.  So due to this the calculation must be done based on the diagonal distance of the lot to prevent the Sims from getting "free" commute.

QuoteNote:  IF a Sim enters a TE Lot that it can easily walk past, it is bad for the simulation of the game.  The reason for this is because when a Sim enters the TE Lot he "forgets" his past and must re-calculate his path.  Doing this re-calculation requires for more CPU useage and will distort the Commute Time shown in you Commute Graph and start giving you funny paths.  Only Sims that are using the lot for a purpose of a Traffic Switch such as Ped to Rail, should be using a TE Lot.

Now, for a station that has only one pickup/dropoff on the South Side of the lot can be more beneficial.  It controls where the Pedestrians are allowed to Enter/Exit the Lot.  This means that we can base the calculation for the Traffic Switch Entry Cost off of the length of the South Side of the Lot, in effectiveness lowering the Traffic Switch Entry Cost because there is a shorter maximum distance the Sim can jump through the lot.

You may now start to see big TE Lots can potentially be a BAD thing.

Cogeo, I have based my calculations off of a very good math base.  I assure you I am correct.  I understand the limitations of a high Traffic Switch Entry Cost, probably more than anyone here.  The problem is people do not know what the problem with a low one is, because they may not be able to comprehend what I have been saying.  For the Simulation to be *optimal*, my formula is the one you need to use.

I have been discussing and thinking (*warning* this will get very technical) about certain lots such as Ill Tonkso's large Rail Stations.  From the main formula we have derived that a 7x9 Rail Station should have a Transit Switch Cost of 3.3.  For the MAXIS Vanilla Simulator this is simply too high, the Maximum Commute Time for Mass Transit is 4.  This means there is only a 0.7 Commute left commited to the 2nd Station and the actual Commute, clearly not enough.  If the second Station was a MAXIS Standard Station 2x3, it would have a Traffic Switch Entry Cost of 0.91.. and you wouldnt have any time left for the physical Commute.  This is beyond the limit for the MAXIS Vanilla Traffic Simulator, but Simulator A and Simulator B have been modified and the Maximum Commute Time is different for this.

QuoteMaximum Commute Time (Mass Transit)
Vanilla Simulator = 4
Simulator B = 10
Simulator A = 14

If you were using Simulator B where the Maximum Commute Time for Mass Transit is 10, that would leave:

   10-3.3-.0.91 =  5.79 Commute Value

Speed of Rail = 90 Tiles per Commute Value

Therefore it would leave the maximum trip a Sim cna travel between stations is 521 tiles, or essentially one large city tile in length.  Simulator A will be even longer than that, but unfortunately it wouldnt work for the Vanilla Simulator.

There is still a way for us to lower that Entry Cost though.  As stated above, for many of Ill Tonkso's large Rail Stations the main entrance is the South Side of the lot.  This means we can make it so Sims will only walk out of the South side of the lot and base the calculation off only the length of the South Side.  In this case it would be 7 in the 7x9 lot.  So 7x0.29 = 2.03, much lower than the 3.3.  This means it will work even with the Vanilla Simulator because it will have an initial trip cost of 2.03+0.91 = 2.94 and have 1.06 left over for the physical commute, which is 95 tiles.  In MAXIS scale everything is small so, 95 tiles is pretty good.

QuoteNote:  Remember, each time a Sim reaches a Station it begins a new trip.  So, each station begins a new segment of the trip which has a fresh commute time to start with.

There is one other way to deal with this that will even further reduce the Entry Cost, but this comes at a sacrafice.  The sacrafice we must make is allowing some Pedestrians to use the lot as a shortcut instead of designing for the worst case scenario.  We could design it so that it is designed for 50% of the worst case scenario, or something to that effect.  Basing the calculation exactly the same except by cutting the time in half.  By doing this you will have stopped busses from entering a bus station and then converting back to busses taking a "free" tile, but a Sim walking past the station may choose to enter it and skip a tile.  This is not as evidently seen because it is masked by Sims entering and exiting the bus station properly is masking the Sim walking past and using it when it shouldnt need to.

There needs to be a community concensus about this and we are discussing this and how to apply it into the XTool to automatically set a Traffic Switch Entry Cost for the user.  When XTool is finished we will have properly modded TE Lots without having to think about it.  Myself and a few others have been discussing and testing the possibilities... this includes both Pros and Cons of the situations.  The thing that Creators need to be aware of is that MAXIS did not intend for large stations, that means if you go beyond the MAXIS scale you are hindering your own work and could render it useless.

Originally Written by: Glazert

jplumbley, I found that a very informative essay. I will have to reread it several times to make sure, but I may be beginning to see the light at the end of the tunnel, though that is the wrong transit metaphor.

Originally Written by: Memo

I also have a question concerning the traffic switch entry costs.
If I connected a grand rail station like Ill Tonkso's large Rail Station with one single street tile only and left some space between the station and the rest of the streets, would it essentially suffice then to set a traffic switch entry cost of "0", since Sims won't be able to use the station as a short cut??
A little illustration:

s......
s..####
s..####
sss####
s..####
s..####
s......

s = street
. = open space
# = station

Quote from: jplumbley on March 29, 2008, 11:27:49 AM
I have been discussing and thinking (*warning* this will get very technical) about certain lots such as Ill Tonkso's large Rail Stations.  From the main formula we have derived that a 7x9 Rail Station should have a Transit Switch Cost of 3.3.

This is a bit hypothetical, since peds would never be walking that distance anyways, if the station wasn't there.

But even with a Transit Switch Cost of 3.3, peds would still, theoretically, be shortcutting the station.
This all due to the fact that peds don't walk diagonally.

The alternative would be to walk around the station, e.g. from the southeastern corner of the station to the northwestern corner:
  + 7 tiles along the southern edge
  + 1 corner/intersection tile in the southwestern corner
  + 9 tiles along the western edge
  + 1 corner/intersection tile in the northwestern corner
  = 18 tiles

With a pedestrian speed of 3.5 Tiles per Commute Value, this would give a total Commute Value of:
  18 Tiles / 3.5 Tiles per Commute Value = 5.14

Clearly higher than 3.3, but still a lot higher than all other alternatives.
None of the pathfinders would of course choose any of the options above.

Even the slowest alternative, cars on streets in Vanilla gives a considerably lower commute value:
  18 Tiles / 21 Tiles per Commute Value = 0.86

Quote from: memo on March 30, 2008, 02:51:59 AM
If I connected a grand rail station like Ill Tonkso's large Rail Station with one single street tile only and left some space between the station and the rest of the streets, would it essentially suffice then to set a traffic switch entry cost of "0", since Sims won't be able to use the station as a short cut??
A little illustration:

s......
s..####
s..####
sss####
s..####
s..####
s......

s = street
. = open space
# = station

That would of course suffice.
But that's not how users plopping a grand rail station would be doing it. $%Grinno$%

Originally Written by: Cogeo

From the postings above, I think it has now became clear that setting the Transit Switch Cost according to the lot size (so that pedestrians don't use the station as a shortcut) isn't a very good idea. In most cases the resulting values are so high that the station would get unusable. As most ingame and custom stations have a cost of 0 (and roads are always available), sims will just opt for another station and/or form of transportation, often resulting in a longer total commute time and increasing road traffic. For example, very few sims would use a 1x1 station with a Transit Switch Cost of 0.36, the value that works best for such a station would be 0.03-0.05 (most shortcutting eliminated). A value of 0.10 causes a noticable reduction of real usage (excl pedestrians and the through trafic) and eliminates almost all shortcutting for 1x1 stations. I would better accept some little shortcutting instead of setting entry cost to a very high value. It would be interesting to see the test results from the station makers.

And of course the station should also work with the standard pathfinding settings too, and not rely on or require installation of a special pathfinding settings set. Btw which is the default max commute time for SC4/RH. I know about the settings for SC4 Vanilla and the ones suggested by jplumbley/mott (refered to as "Simulator A" and "Simulator B"), but what are the ones for RH? Are they the same as Vanilla's? I think they should be named simply "default" or "standard", otherwise (ie using the term "Vanilla") players may think that the station is for non-RH SC4 only!

Memo, I agree with RippleJet, such a street layout works but it's totally unrealistic for a Hbf. Another alternative would be to limit the transit switches to specific lot sides, eg front only, setting the Transit Switch Point property. I have done so for my Suburban GLR stations (STEX), eg buses can only reach the stations from the front side. Just an idea.

Originally Written by: Warrior

The HSR is dependant on the NAM so all calculations should be based on the NAM as the uesr will have to have the NAM to be able to use HSR which is needed to be able to use that station.


Also this is getting slightly off topic and [almost] all the posts have been made by NAM members so any more posts on the TE settings should go in the HSR private thread, thanks.

-Jonathan.

Originally Written by: Cogeo

Warrior, I meant the default stuff that comes with NAM, the A and B sets are not a standard part of NAM and therefore many players may not have them installed, so stations shouldn't require them. But you're right this is station-making, so it should be posted in another thread.

Now about that dual track, what would you suggest? I think memo is right, plus making them this way doesn't require changing the global settings (ie what kind of traffic is allowed on each network type). Also take into account that they may prove quite complicated, as they have to connect and branch to both GHSR and Rail networks (in addition to the dual network itself). I don't know how difficult this can be if implemented as draggable, but even as puzzle pieces it would require a lot of work, because of the many combinations. For example, the "Right Junction" type alone would need the following puzzle pieces (I hope you understand what I mean in each case, descriptions may be confusing so please correct them if needed, my English is poor):

1. Rail/HSR Dual Track - Right Junction
2  Rail/HSR Dual Track - Right Branch to Rail
3. Rail/HSR Dual Track - Right Branch to HSR
4. Rail/HSR Dual Track - Right Turn / Forward Branch to Rail
5. Rail/HSR Dual Track - Right Turn / Forward Branch to HSR
6. Rail/HSR Dual Track - Right Branch to Rail / Forward Branch to HSR
7. Rail/HSR Dual Track - Right Branch to HSR / Forward Branch to Rail
(Hope I didn't forget anything).

Seems a lot of work, though as a design (graphically) these can look almost identical. If this shared track has the look of conventional track (and I think this is the idea, ie HSR terminating at "old" stations), for the above puzzle pieces there would be required 3 textures. One identical to the conventional Right Junction track (1,2,4) one with a forward/rail-to-HSR transition (5,6) and another one with a right/rail-to-HSR transition (3,7), ie both dual track and conventional track sections would be identical.

Comments please.

Originally Written by: Warrior

I think the new simulators will be part of the next NAM. Check the release thread.

Originally Written by: Cogeo

Quote from: Warrior on March 30, 2008, 08:46:21 AM
What if a base set was included in the first beta (providing even that is ready in time) So there are no branches, just orth, diagonal basic crossing (if any) and then transitions?

I think a minimal base set should include the straight one and the branches - these are absolutely essential as both GHSR and conventional track needs to connect to them. You may even omit "basic crossing" (provided that tracks will diverge a few tiles off the station), but branches are absolutely necessary, otherwise how will different types of track connect to? Take as an example the Avenue/GLR puzzle pieces, without the "connectors" they are unusable.

Quote from: Warrior on March 30, 2008, 08:46:21 AM
There should probably be a slight visual difference between normal rail and and the dual network?

Dunno, I guess someone has to post a few pics.

Quote from: Warrior on March 30, 2008, 08:46:21 AM
We could add monorail paths to the whole MAXIS rail network and modify the simulator(which shouldn't be too much of a problem) this way HSR trains can run on rail without having to spend money and time redrawing rail.

This is a radically different approach, and I don't know how much work this can be, or how good it is technically.

Quote from: cogeo on March 30, 2008, 06:56:41 AM
From the postings above, I think it has now became clear that setting the Transit Switch Cost according to the lot size (so that pedestrians don't use the station as a shortcut) isn't a very good idea. In most cases the resulting values are so high that the station would get unusable.

Whether or not it the Transit Switch Cost is based on the speed of Pedestrians, or if it is based on the speed of the slowest alternate Traffic Type in the lot (busses for Streets), the calculation still must be done based on the size of the lot.  A lot that is 1x1 shouldnt have the same Traffic Switch Cost as a 7x9 large rail station.  They are not the same and should not be considered the same from the viewpoint of the Simulator.

QuoteAs most ingame and custom stations have a cost of 0 (and roads are always available), sims will just opt for another station and/or form of transportation, often resulting in a longer total commute time and increasing road traffic.

No TE Lot including the MAXIS Vanilla ones that have a Traffic Switch Cost of "0" have been modded properly.  This is bad for the Simulation for your game and should be fixed.

QuoteFor example, very few sims would use a 1x1 station with a Transit Switch Cost of 0.36, the value that works best for such a station would be 0.03-0.05 (most shortcutting eliminated).

If the lot is 1x1, and is intended to be a Bus Station, the slowest alternate speed of the TE Lot would be Busses which is 35.  This means that the Traffic Switch Cost would be 0.042 instead of 0.42 for Pedestrian Traffic.  This would prevent all Bus, Cars, etc from "jumping" through the lot.  But, it would still allow Sims to walk through the lot at an "increased" speed.  I think this would close to the best situation, but instead of taking the calculation through the diagonal of the lot is should be ( Length + Depth x 0.029¹ = Traffic Switch Cost ), which would mean 0.058 would be the cost for a 1x1 Bus Station.

¹ 0.029 is the cost per tile of the Bus Speed.

QuoteA value of 0.10 causes a noticable reduction of real usage (excl pedestrians and the through trafic) and eliminates almost all shortcutting for 1x1 stations. I would better accept some little shortcutting instead of setting entry cost to a very high value. It would be interesting to see the test results from the station makers.

Me too.

QuoteAnd of course the station should also work with the standard pathfinding settings too, and not rely on or require installation of a special pathfinding settings set. Btw which is the default max commute time for SC4/RH. I know about the settings for SC4 Vanilla and the ones suggested by jplumbley/mott (refered to as "Simulator A" and "Simulator B"), but what are the ones for RH? Are they the same as Vanilla's? I think they should be named simply "default" or "standard", otherwise (ie using the term "Vanilla") players may think that the station is for non-RH SC4 only!

Of course it should work with the Vanilla, it is only the really large stations that people have made that are completely out of scale with the MAXIS Vanilla standards that are the ones that will run into this problem.  As, shown even when they are way out of whack it is still possible to get them under that threshold to a certain point.  People need to remember there is a limit to everything and we cannot go making 20x30 lots intended to be a station.  This is just simply too big compared to the scale of the rest of the game.

The MAXIS Vanilla Maximum Commute Time for Mass Transit was posted in my previous post and I used it to show that Ill Tonkso's 7x9 station would not work unless we used the South Side only approach to modding his lot.  That value is 4. 

Vanilla means SC4/RH made by MAXIS, out of the box.

QuoteMemo, I agree with RippleJet, such a street layout works but it's totally unrealistic for a Hbf. Another alternative would be to limit the transit switches to specific lot sides, eg front only, setting the Transit Switch Point property. I have done so for my Suburban GLR stations (STEX), eg buses can only reach the stations from the front side. Just an idea.

Agreed.  And this is why we must have proper Traffic Switch Entry Costs.

I have been modding some transit stations on my own and did some experiments to see which transit switch entry cost would prevent Sims from walking through the station, with some interesting results.  Four experiments were performed using Mott's park and ride traffic simulator (walking speed 5, only walking can reach destination).  Transit  stations were modded to have no jobs and to only allow pedestrian traffic.  (The cities only allowed pedestrian traffic.)

Experiment 1:  What entry cost would prevent Sims from short cutting through a station of varying length along a road?  In this scenario the road only touched the station along one side.  Stations were varied in size length from 1 to 20.  The results are below.

TSEC: Transit Switch Entry Cost

Based on this data (and some additional testing including walking speed of 1) a curve was fitted to the data with the following equation:


Experiment 2:  What entry cost would prevent Sims from short cutting through a station of varying size at a corner of a road?  In this scenario the road touched the station along two adjacent sides.  Stations were varied in size as listed below.
Surprisingly all stations had the same value that prevented most Sims from using the station as a shortcut 1.3435/Transit Speed.

Experiment 3:  What entry cost would maintain the commute time of Sims traveling across a transit station?  A city was setup with residential on one side and commercial & industrial on the other separated by 9 tiles in one of the following configurations; pedestrian mall tiles, pedestrian mall tiles alternating with a 1X1 transit station (5 pedestrian mall tiles & 4 transit stations), pedestrian mall tiles alternating with a 1X3 transit station (3 pedestrian mall tiles & 2 1X3 transit stations), and pedestrian mall tiles on either side of a 1X7 transit station.
The transit switch entry cost that maintained a constant commute for the 1X1, 1X3 and 1X7 station was found to be the same as found in experiment 2, 1.3435/(Transit Speed).

Experiment 4:  What entry cost balance the number of Sims traveling across two equal distance paths?  A city was setup with residential on one side and commercial & industrial on the other separated by 9 tiles with one path of 9 pedestrian mall tiles and the other in one of the following configurations;  pedestrian mall tiles alternating with a 1X1 transit station (5 pedestrian mall tiles & 4 transit stations), pedestrian mall tiles alternating with a 1X3 transit station (3 pedestrian mall tiles & 2 1X3 transit stations), and pedestrian mall tiles on either side of a 1X7 transit station.
The transit switch entry cost that kept all the Sims on the pedestrian mall tiles was 1.3445/(Transit Speed).  At 1.323/(Transit Speed) half of the Sims used each route.  With 1.303/(Transit Speed) all the Sims traveled the route through the transit stations.


Another interesting point is that 1.3435 is 95% of the diagonal across a 1X1 tile.  I do not know if this means anything or not.

Is it possible that the transit switch entry cost is applied for each tile across a station?

If this information is of interest to anyone, or if there are questions please let me know.

Christopher

This indeed is very interesting results.

I did briefly see your post last week... Ive just been running around like a chicken with my head cut off and doing a few other things.

This really does seem like it has been a thurough test and it is very interesting that you came up with essentially the same number in all your tests.  That is almost mind boggling.  It really does make sense that the Traffic Switch Cost would be applied to every tile.

So....  From what I am understanding.  Since most Transit Stations allow Traffic to enter from all 4 sides the best Experiment to work from is Experiment 2.  Here you said that 1.3435/Traffic Speed prevented a change in the commute.  This would mean that for a Simulator with a Pedestrian Speed of 5 you would get a value of 0.2687 to prevent Sims from "short cutting" through a station.  Is this correct?

If so... that is definately surprising.  But, yet makes sense.

I wonder at what point does the Simulator choose Driving before Walking and if different size TE Lots has an impact on this.  A Test for this could be set so that there is a TE Lot 10 tiles from the residential and work 150 tiles from the residential on a Road with Mott's Simulator.

QuoteThis would mean that for a Simulator with a Pedestrian Speed of 5 you would get a value of 0.2687 to prevent Sims from "short cutting" through a station.  Is this correct?

This is the gist of what I found when a TE lot was placed on a corner.  The value 0.2687 seamed to fit better with the rest of the experiments.  At this value 2 Sims actually cut across the station.  Much like experiment 4 a value of 0.13445/Transit Speed (0.2689 for a transit speed of 5) totally eliminates all short cutting.  As this value is gradually decreased more Sims cut across.  If roads encircled a lot or formed a "U" around the lot with the opening North then to prevent Sims from cutting across the lot on the North would require a larger value. 

QuoteI wonder at what point does the Simulator choose Driving before Walking and if different size TE Lots has an impact on this.  A Test for this could be set so that there is a TE Lot 10 tiles from the residential and work 150 tiles from the residential on a Road with Mott's Simulator.

I am not sure I totally understand the question.  Would this be similar to experiment 3 with one pathway across ped malls and TE Lots, and a longer road that loops around connecting the residential and work (commercial & industrial)?

If I understand the simulator correctly choosing driving versus walking is based on a total trip cost which is a balance between the preferred travel strategy (which determines trip starting cost), transit speed, distance, and transit station entry costs.  When driving and buses were allowed the slow bus speed on roads of Mott's simulator (transit speed of 5, same as walking) made it so all Sims drove.  As the bus speed on roads was increased more people took the bus.  An alternative way to encourage bus traffic was to increase the trip starting cost that Mott had for cars from 1 to 1.95.

Uhhh, I just like to watch the cars go back and forth...

CLR1SC4D...

...that's a mouthful, can we call you CLR for short?

Those two comments have to be the best informed and most detailed first two posts any member has ever made here.  A major hats off! to you- if you can catch Jason's (jplumbley) attention like that, you've got quite a future ahead of you here.

Glad to have you aboard.


David

Quote from: CLR1SC4D on May 10, 2008, 01:40:04 PM
I am not sure I totally understand the question.  Would this be similar to experiment 3 with one pathway across ped malls and TE Lots, and a longer road that loops around connecting the residential and work (commercial & industrial)?

If I understand the simulator correctly choosing driving versus walking is based on a total trip cost which is a balance between the preferred travel strategy (which determines trip starting cost), transit speed, distance, and transit station entry costs.  When driving and buses were allowed the slow bus speed on roads of Mott's simulator (transit speed of 5, same as walking) made it so all Sims drove.  As the bus speed on roads was increased more people took the bus.  An alternative way to encourage bus traffic was to increase the trip starting cost that Mott had for cars from 1 to 1.95.

Well, if we were to make both MT and Car "equal" (by setting the Preffered Trip Costs to 0) and make a test where the time it takes to reach the destination is equal (destination for MT is a station).  Then we can test differing lengths of Stations where the Traffic Switch Entry Cost is the same and we can see if the different lengths of stations will add to the total commute, even with the same Traffic Switch Entry Cost.  Doing this will confirm whether or not the Traffic Switch Entry Cost is being added to each tile in the TE Lot, or if it is just a single cost.

Sorry about the long time between responses.  Other than the normal reasons like being busy I have been further testing the commute engine to better understand it, and isolate parameters so that I am dealing with only one at a time.

@dedgren: Thank you.  CLR, Chris, or Christopher is fine.

While working with the commute engine exemplar to refine information on "Transit Switch Entry Cost" I have come across some interesting info and discrepancies between what has been posted and what the tests are showing.  If I am missing something, or the test information can be repeated/confirmed it would be good to know.  (If this would be better posted in new thread or in the commute engine tweaking thread started by Mott let me know and I can post there or it can be moved.)

All the information was gathered with experiments where each Sim had the same network travel length by setting up a city in some form as below (except for point 4):

R   C
R   C
B==B==B
R   I
R   I

Where:
"R" = Residential Zone
"C" = Commercial Zone
"I" = Industrial Zone
"B" = Bus station
"=" = Road
In some scenarios the bus stations were replaced with roads or changed in size.  Most scenarios the bus stations did not provide jobs.  Also some experiments had a road looping around the top.

As the bus station in the middle was increased in length, with all other parameters remaining the same, the Commute Time increased proportionally to the increase in number of bus station lot tiles and the "Transit Switch Entry Cost".

Below are the summarized results in no specific order.  Information related to transit switch entry cost has been bolded.

1.) Commute Time displayed can be calculated by:


with the following criteria:

• Sims start at the center of the tile beside the residential.
• Sims enter a lot at the center of the tile beside the lot.  If the commuter is entering for a job the "Transit Switch Entry Cost" is not counted.
• Sims leave a transit enabled lot at the center of the lot tile next to the transit network at the speed of the transit network it is going to.
• When calculating the number of Transit Station Tiles Transversed only those tiles crossed in the station plus the entry tile are counted.  If the width of a 1X10 station is crossed then one tile is counted, if it is crossed in the length then ten tiles are counted.

To get the appropriate "Transit Switch Entry Cost" to maintain commute time across a transit station can be calculated by:
To maintain consistency with the previous information on preventing Sims from cutting across a station on a corner a "transit Switch Entry Cost" could be calculated by:
(96% of the diagonal across a 1X1 tile)

This Commute Time calculation might also roughly work with a scenario that jplumbley posted in Mott's thread on commute engine tweaking.  A Sim traveling a average network speed of 60 over 225 tiles would have a commute time of 90

2.) The "Trip Length to Minutes Display Multiplier" seams to have no effect on the commute time graph.  I have changed this by factors of 10 from 0.001 to 1000 with no changes.  This has occurred even when starting a new city in a new region.  Is this possibly a pre RH property or do I need to be looking for the effect elsewhere.

3.) "Trip Starting Cost by travel type" does not affect the commute time by adding to it directly.  It is the amount of extra time a Sim is willing to take on his preferred travel method (mass transit vs. car) before switching to the faster method.

4.) For non-rail travel (walking, bus, car, freight truck, monorail) adding to congestion also adds to air pollution.  I had a city of pedestrians only that I allowed to add to congestion and I obtained air pollution on my roads.  Likewise be switching off monorail congestion I was able to eliminate the pollution it was causing.  (Different city used for testing than the rest of the experiments.)

5.) "Commute trip max time" and "Max Mass Transit Strategy Trip Length" both seamed to be for one way of the commute.  When I set up a scenario with walking/driving speed of 1 and a "Commute trip max time" of 4 the Sims would travel a total of 4 tiles one way to reach a job.  Residential zoned further out would not develop.  This was consistent for other max trip times/lengths and network travel speeds.  (Network speed of 15 and a Commute trip max time of 17 allowed a Sim to travel 255 tiles across a map one way.)

Christopher

Im gonna go backwards here and start with #5.

5.)   What doesnt entirely make sense is that the Sims will walk upto 11 (tested) tiles in the default Simulator.  Any homes beyond that are abandoned.  This varifies the calculation mott did where you divide the "Commute trip max time" property by 2 and multiply by traffic speed.

6 max commute /2 = 3 * 3.5 speed = 10.5 tiles

So this test contradicts your findings...

4.)  Yes, the pollution is something that was known in the past and not realizing it when I packaged mine and mott's Similators for inclusion with the NAM... I set the monorail to be affected by congestion.  MAXIS dropped the ball on that one!!

3.)  I dont agree.  The values for "Trip Starting Cost by travel type for Mass Transit" are 0,1.95,0,0,0,0,0,0,0.....  The 1.95 value is in the position for Car Travel type, which would make sense only if it was adding this as a negative effect to the Car Travel type to promote the Sim to use MT.  Vice Versa the Car Preffered has 0.1 in the Pedestrian Travel type and again would only make sense if it was adding this as a negative effect to walking (MT) and promoting the Sim to use his Car.

2.)  The "Trip Length to Minutes Display Multiplier" has a minor effect, actually only a "cosmetic" one in reality.  The "Commute Time Graph" will always look identical in shape, what changes is the value on the Time.  When the value of the Multiplier is set to 1, it should display the "raw" average Commute time of your Sims in the display graph.  If it is equal to 2 then and they are all walking at a speed of 3.5 then your average walking distance is 7 tiles.

Now, if you change this value what you are changing is the value of minutes that each Commute Time Point is worth.  So, it is is set at 17, each 1 in Commute time is equal to 17 mintues of travel on the graph.  At a travel speed of 60 for Avenues that means the Sims go 60 tiles in 17 minutes.


1.)  I need some more time to read what you have said, so i will respond to this later.  Plus GF is getting angry with the clicking from my fingers hitting the keyboard.

Some additional information found on the points you addressed

5.)  I will go back to the default simulator and make more tests to see if the differences can be clarified.  I was also interested in testing how one-way roads may affect this when the morning and evening commute are different lengths.

Edit: Testing with the default simulator (walking speed = 3.5, "Commute Trip Max Time" = 6) I had Sims walking 21 tiles from the residential to the commercial.  Why this is different from previous tests with the default simulator is a mystery.  When testing with one-way roads, driving speed 5, "Commute Trip Max Time" = 6 Sims would travel at max 31 tiles for the morning commute in approximately 150 time from the residential to the commercial.  The one-way road then continued in a loop greater then 3 times (> 100 Tiles) the distance between the residential and commercial for the morning commute which the Sims traveled for their evening commute with out problems.
<tt>
Diagram of one-way road testing.  The one-way road was laid in a counter clockwise direction.
===========
=         =
=         =
=         =
=         =
=         =
=         =
=         =
=  R   C  =
===========
   R   I

Where:
"R" = Residential Zone
"C" = Commercial Zone
"I" = Industrial Zone
"B" = Bus station
"=" = One-Way Road
</tt>

3.)  When I was testing this I had a city that started with car traffic only and the "Trip Starting Cost by travel type for Mass Transit" set to 0,0,0,0....  As I increased the "Trip Starting Cost by travel type for Mass Transit" there was no increase in Commute Time values until the Sims switched to travelling by bus, at which time the cummute time was that for the mass transit trip.  The time that the Sims switched to mass transit was when the commute time for car plus the "Trip Starting Cost by travel type for Mass Transit" was equal to the mass transit trip time, but the "Trip Starting Cost by travel type for Mass Transit" never appeared to increase the Commute Time Value.

2.)   It is the value on the Time that I was looking at and there was no difference as I changed the "Trip Length to Minutes Display Multiplier".  This was tested with NetworkAddonMod_Traffic_Plugin_A_Hard.dat, as well as B_Hard and C_Standard.

Thank you for your review and time.

After mott posted his excellent tutorial TE Lots, Transit Switches, and You (http://sc4devotion.com/forums/index.php?topic=2763.0), I posted this reply and request:

Quote from: RippleJet on October 25, 2007, 04:40:40 PM
I will implement the transit switch entry cost in the formulas for the "X Tool" for the creation of custom stations in the future!
And I suggest that the next NAM changes that property for all in-game stations accordingly! ;)

Stations modded with PIM-X have since had an Entry Cost of 0.30.
After CLR1SC4D's excellent investigation and testing above, I suppose it should be lowered to 0.27 though.

However, making custom stations with an Entry Cost is futile as long as Maxis' own stations still have an Entry Cost of 0.00.
The picture below shows Sims walking some 17 tiles to a Maxis bus stop instead of taking any of SimGoober's closer ones.

(https://www.sc4devotion.com/forums/proxy.php?request=http%3A%2F%2Fi232.photobucket.com%2Falbums%2Fee198%2FRippleJet%2FStupidSimsTakingMaxisBuses.jpg&hash=6386f5dd8870359ccdc52529e1cfc03d773819ab)

PS. Thanks Barby for the picture! :thumbsup:

In this picture there are four SimGoober 1×2 bus stops, all being closer to the Sim's house than the 1×1 Maxis stop in the lower left hand corner.
The Sims actually walked past one of SimGoober's bus stops on the way to the Maxis bus stop.
All these stops by SimGoober do of course have an Entry Cost of 0.3.

I know NAM doesn't contain any lots... at least the latest update did not change the Entry Cost of any of them.
Should I in that case make a separate mod that changes the Entry Costs for the in-game stations and stops?

---

Secondly, regarding RTMT stations, and in order not to bring this discussion up in the RTMT Team's board...
If the pathfinding is zeroed at each station the Sims pass, should we in that case even use the in-game ELR and subway stations?
Both the Elevated Light Rail station and the subway station are TE lots made byb Maxis (although the subway station is a little bit different).
Has any testing been made covering the subways? Wouldn't the pathfinder prefer subways over buses and cars in that case?

there is so much custom content and a variety of modded stations available now.. indeed why even use any of
the original maxis stations.. ,wouldn't that solve the problem ?

Using the Cam and with such large cities now, i find using any of the maxis stations virtually irrelevant now,
mostly because of the low capacities they have..

Thanks for your investigations into this ripple :)

Brian

Quote from: RippleJet on September 11, 2008, 06:52:58 AM
I know NAM doesn't contain any lots... at least the latest update did not change the Entry Cost of any of them.
Should I in that case make a separate mod that changes the Entry Costs for the in-game stations and stops?

I would certainly recommend that this be done - I think everyone's in agreement that there's no case in which the Entry Cost should be zero.  And if you're doing that, you might want to consider updating the stations' capacities and cost to match the simulator and difficulty being used.  This is clearly a separate issue, but it's something that would make the Maxis stations much more useful.