RH Commute Testing...

[andyp_m 0]

I really appreciate all you all are doing to improve Sim City for all of us. I have one question regarding the testing of RH. Have you considered what effect the skill level of the city has on the path finding and commutes? It seems to me that if they are going to make playing the game easier or harder, one of the first things to change would be making the traffic model more friendly or more difficult to the mayor. Just one more thing you might want to test.

[Marc Collins 36]

Check out the attached pic. It is a modest sized city with C and R in the north and a decent sized industrial zone in the south. A major city is to the east and a newly developing one to the south.

I went in there to do some pathfinding testing so I could then apply the new temporary mod and see how it affected things and performance on my machine. But before I even do that, I wanted to note and ask here about something very strange--almost all of the buildings below the strip of forest have no workers!!!! The three power plants, the few commercial buildings and almost all of the industry have no workers, but all have freight truck connections. The few that do have workers show them coming primarily from outside this city.

Everything north of the forested line, including the I$$$, has workers and normal commuter relationships according to the query tool.

The I establishments are not abandoning or showing any problems so far in the normal query.

Anyone else seen something like that? By the way, this region and city pre-existed RH.

[redlotus 16]

Marc.  I can only think of three explanations for this happening.

 

Since this was a city created before you installed RH, it takes the game several months to sort out where everyone is commuting to/from.  The larger your city is, the longer this will take.

 

If you had opened this city before and run it for several months before saving it, then your situation is probably being caused by a lack of appropriate wealth/education level sims.  Your city's educational level has reached a point where the dirtier industries in the south are relatively unneeded.  But, since the patch 2 fix, ID (and sometimes IM) will not abandon for anything.  They may say that they have workers in their query boxes, but in reality, they don't.  The few low-wealth, not-so-bright sims left in the city are all taking jobs in the closer industrial and commercial buildings to the north.  One of the best tell-talke signs of this is that demand for dirty industrial is negative.

 

The last idea is closely related to the first:  if you haven't run and saved the cities to the north and east since you've installed RH, the game cannot get all of the information it needs to show an accurate route query for those neighboring sims.  Before you do any testing, you should make sure to run each of the connected cities for several months (or a year, just to be safe) and then save so that the commute engine can reorient itself.

 

Anyway, I hope this helps.

[the7trumpets 10]

qbert2 - amazing!

 

You are definately doing some awesome experiments, and with you on board, this should go more quickly than I feared

 

Marc Collins - I agree with the second hypothesis that redlotus offered (not enough sims of that wealth class and education level).  Unfortunately, Patch 2 made it so that we are forced to demolish dirty industry ourselves if we don't want them.

 

Usually by the time I want to get rid of them, I have lots of demand for HT, but HT hates being near ID, so they don't replace the ID buildings since the surrounding ID buildings have not abandoned.  It is annoying, but hopefully I'll get back on topic for this thread later tonight

[Marc Collins 36]

Jeez...I had no idea that I$ and I$$ didn't act properly--I$$$ certainly degenerates and loses jobs...

Are you saying that if I look at the I$ and I$$ jobs in the graphs, the jobs won't really be there...or that those numbers will belower than the actual developed and queriable lots in my city?

Hmmmmmmm...

[redlotus 16]

 

I think that my wording could have been a little better in my last post.  The jobs are still there and will be reflected in your jobs and population graph.  However, they are not really being filled by sims.  So the actual number of FILLED jobs will be lower than what the graph says.  Does this make any sense?

[qbert2 9]

Has anyone figured out why the commute time is greater than 120 minutes in RH but the max number of tiles that can be travled stays the same? IE: In pre-rh max road tiles was 193 with a max commute time of 120 minutes. In RH I've found max road tiles to still be 193 but the commute time looks to be somewhere around 145-155. Picture

[the7trumpets 10]

wow qbert2, this is pretty disconcerting, but I need to do some more testing and some math before I can say this is for sure happening.

 

 

Comparing our current speed tables to the tables in the original report, I see that the speeds now are actually slower (yes, slower) than what they were before.  However, the kph values in the traffic exemplar have not changed!  This means that the x10 multiplier used before is now closer to a x11.1 multiplier.  To know for sure, we need to do some math on the traffic speeds found so far, but it definately looks like the effective road speeds are now slightly slower.  Possibly this was done to cause people to need to put avenues in thier cities in order to lower commute times...who knows.

[+Guest]

Strange....

[Marc Collins 36]

Could that be what's happening on a massive scale in my city? That almost an entire existing I$$ region is not really there? And therefore the route query tool is just showing the truth? Do these buildings provide tax revenue? Isn't this a massive problem for anyone interested in realism??????????

[the7trumpets 10]

----------------

On 9/27/2003 9:54:42 AM Marc Collins wrote:Could that be what's happening on a massive scale in my city?  That almost an entire existing I$$ region is not really there?  And therefore the route query tool is just showing the truth?  Do these buildings provide tax revenue?  Isn't this a massive problem for anyone interested in realism??????????

----------------

The problem is that people (like us) complained about massive abandonment  (caused by poor pathfinding in the first release of the game) and couldn't play it without R$$$ abandoning due to commute time.  As a response to this, maxis put a desirability threshold in patch 2, which precludes buildings from abandoning if they are at a high enough desirability.

The result is that ID and sometimes IM will not leave for anything.  The only thing I can think of is to modd the desirability factors for ID so that higher land values and other factors kill thier desirability.  This would keep the entire map from having bright green desirability.

[Melon 0]

So one of the bottom lines here is that the commute x10 bug (or x11 now?) is still in the game?

How soon does the most excellent mod squad expect

[Marc Collins 36]

OK 7 now I am really worried. I started a *new* city adjacent to the one I described above and it too has this problem!! I have IM and ID all over the place that is fake??? How could a big swath of IM develop from scratch in a new city and en mass have no jobs???

[Skippydam73 0]

Ok here are two things I gathered last night with something weird that I will have to post in a day or two.

For the mod DM did for the better and perfect, the better worked just great. They all found the closet and fastest road. I turn off my sound so I did not notice any glitches or slowdowns. I did not try the perfect because everyone was finding the closet path.

Second, I was starting to see a very low desirable for my adjoining city. I started all over on my seattle region and it took a lot longer to start building up my industrial area than before. I wonder if that threshold Maxis adjusted in Patch 2 somehow got taken out by accident? With the Better pathfinding mod, it helped some but not much.

[Marc Collins 36]

But I have a new city with high desirability and demand for IM, a big zoned area for it, lots of R$ and R$$ population, the IM lots developed almost immediately (as expected), are jammed in at maximum coverage--but they all have no actual employees!?!?!?!?!?!

This is incredibly stupid if it was designed that way.

The bus stop is not used in that area...which I also had happen pre-RH, but now I guess we can figure out the reason for that seemingly strange occurrence.

My unmodded RH pathfinding doesn't seem to have any problems getting sims to travel quite far or to other cities to get a jobs. But if the development that is already there has no jobs, then all the pathfinding in the world won't help.

For me, this is a vastly bigger problem than the pathfinding tweaking for the obvious reason I just stated.

[Skippydam73 0]

----------------

On 9/28/2003 9:50:52 AM Marc Collins wrote: But I have a new city with high desirability and demand for IM, a big zoned area for it, lots of R$ and R$$ population, the IM lots developed almost immediately (as expected), are jammed in at maximum coverage--but they all have no actual employees!?!?!?!?!?!

This is incredibly stupid if it was designed that way.

The bus stop is not used in that area...which I also had happen pre-RH, but now I guess we can figure out the reason for that seemingly strange occurrence.

My unmodded RH pathfinding doesn't seem to have any problems getting sims to travel quite far or to other cities to get a jobs. But if the development that is already there has no jobs, then all the pathfinding in the world won't help.

For me, this is a vastly bigger problem than the pathfinding tweaking for the obvious reason I just stated.----------------

This is going to have to be tested to make sure. Now remember, it takes four months average for the cycle that the simulator takes on seeing if that sim has his job or not and if it needs to find a new one. The whole thing does behave differently and hopefully as T7T and others do testing on the transportation, we can figure out if its working correctly or not.

[Marc Collins 36]

After sim years I can confirm that this not only affects ID and IM, but IHT and utilities and institutions (and probably everything else depending on the particular conditions). My hospital has workers and then it doesn't, then it does again. Power plant operating for years with no workers...same for schools.

So next I thought it could just be that the sim isn't looking far enough afield for workers, so I used the new perfect pathfinding mod....no effect.

Looks like SC4/RH is just one giant fudge factor to get the game to play reasonably well.

[the7trumpets 10]

The old pathfinding engine did not look for workers for the jobs available, it looked for jobs for the sims who needed work.  Seeing as how the no job zots still appear over the houses, and not the jobs, it seems reasonable to assume this has not changed.

 

The result of this is that new jobs that develop are only filled with jobs immediately when there are sims within that city who are unemployed, or would preffer to work in the newly available jobs (because of distance or wealth/education level).  In regoinal play, this means that if you start a new neighbor with lots of IM demand, while the neighboring city has lots of R$ and R$$, the new neighbor dedicated to IM will not show any sims traveling to the jobs until you save that city with some IM population, open the residential city and run it until the pathfinding engine finds routes to the neighboring city, and then re-open the IM city and let it find routes from the city border to the IM jobs.

 

The simulator is also fairly forgiving on how many jobs you can create before demand stops, especially on the new easy and medium levels.  This means that even within the same city, the simulator will let you create something like 15% more jobs than you have need for.  At this point, R demand will rise.  If R demand is negative, there are too many sims for the number of jobs that are available, so it seems reasonable to assume that the actual 'equalibrium' point (where all jobs are bing filled, and no sims are unemployed) is when you have strong I+C demand, and R demand is sitting right at about zero, without being negative.

 

I hope that makes sense

[+Guest]

Think you could mod it so it also looks to fill existing jobs before creating new buildings? Or can that not be changed?

[Sharr 6]

I think that job-searching and pathfinding are two distinct problems that probably should not be addressed in this thread simultaneously. We are off the original topic. I do understand that finding jobs drives the need for pathfinding, but that is a whole other problem in and of itself.

If you want to make a job-searching engine thread, by all means, be my guest. But let us stick to the actual observable pathfinding here.

[+Guest]

Sorry. I kinda figured they would be kinda the same, but I'm wrong. Go figure. Anyway, T7T, can you give us any kind of time-table on a release?

[Marc Collins 36]

Thanks 7...I hope that's all it is. What's interesting is that the new perfect pathfinding mod seemed to have no effect on this or anything else. 95% of sims still work across the street or around the corner from their residence. The other 5% seem to be able to pathfind a very long way without any mods in my cities.

[stamp 4]

half of my commercial sector has no commuters, yet they have jobs filled.
note: the side that has no commuters is at the edge of the city (not the map)


each building has jobs. every one of those buildings' jobs are filled (using the building query). but when i use the traffic query, there are no workers. and its not just that building, its pretty much all of the buildings that doesn't have a blue tinted road (better part of the screen).

So, no workers, yet there are jobs. i have NO CLUE what the problem is. I just put in the perfect temp path mod, and the results are the same.


its been like this for AT LEAST 10 years, and it still does this.
Any suggestions?

[Sharr 6]

My suggestion is to not post here unless you have some testing results of observed pathfinding, or commentary on testing results expressed here. If you have nothing of substance to add to the actual testing of the engine, please post your issues with jobs on a different board.

I will delete anything not relevant to this thread if this keeps up. Thanks.

[Skippydam73 0]

----------------

On 9/28/2003 10:16:40 PM Sharr wrote: My suggestion is to not post here unless you have some testing results of observed pathfinding, or commentary on testing results expressed here. If you have nothing of substance to add to the actual testing of the engine, please post your issues with jobs on a different board.

I will delete anything not relevant to this thread if this keeps up. Thanks.----------------

Sharr,
I started a thread to hopefully get everyone to keep them separate. Its called City and InterCity Job testing and the link is below for everyone.

https://www.simtropolis.com/idealbb/view.asp?topicID=39958&sessionID=AD81B509459D498993C2FF3F74714975

[the7trumpets 10]

Thanks for keeping us on track Sharr, I was beginning to feel the same way

 

 

I have gone through all the values I have tested, and it does seem like the x10 multiplier which was previously used is now a x12 multiplier.  It is confusing why maxis changed this, but it does appear this is the way it is

 

Attatched is an excel workbook including the current tested commute times, calculated speeds and showing the x12 multiplier.

 

 

Other than that, here's what we know so far:

 

• Commute time graph STILL displays the trip from home to work doubled, even though the trip from work back home is now calculated.
  
• Trip from work back home is calculated, and affects the traffic congestion dataview
  
• A sim must have a non-walking path to and from thier workplace in order for the zone to develop.  However, once the zone has developed, they can walk to and from thier job.  Anotherwords, the determination of whether a zone is 'connected' to a job does not consider a walking path valid.
  
• Commute x10 multiplier or bug, which attempted to account for the fact that sims work at the closest job available when in reality people do not do this, has been increased to a x12 multiplier.  However, the limit to the length of the commute time as well as where sims change from short, medium, and long commute times, has remained proportional.  So, the distance they can travel has not decreased, but the times in the graph have
  increased by 20%.
  
• Old concept of sims starting the commute on the tile adjacent to thier house and ending at the network tile adjacent to thier workplace is still in tact.  I have noticed that it seems like sims can no longer get onto a road from thier backyard, as they could previously, but more testing is necissary to verify this.

 

Anybody know anything else?

[the7trumpets 10]

A coupple more things to throw in the mix:

 

MT buildings appear to have thier pathfinding done differently.  Previously, a network tile from each network the sim was trying to switch between had to be adjacent to the same tile in the MT lot.  However, now as long as the two types of networks border the MT lot anywhere, sims can switch between the two modes.

 

 

Inter-city commuting is beginning to boggle my mind.  I started testing it, and it looked rather strange.  Basically, it looks like it's behaving the same as pre-RH with one major difference.  As soon as the simulator realizes it's an inter-city commute, it deletes all of the commute times from the neighboring cities, and only incorperates the commute time within that single city.  This means that for a commute which was previously 20 min in a R city and 50 min in a working city, the pre-RH commute time in the R city was 50 min, but post-RH the commtue spikes up to 50min, and then comes back down to 20 min within a month of sim time.

 

If this is true, it has a very cool effect, but a very dumb effect as well.  It makes vast commutes accross the region possible.  BUT, the size city you are using directly affects what your commute time is if you zone realitically, with R being very separated from working zones.  Since the commute time for regional commutes is now simply the time it takes the sim to go from thier house to the city border, and does not encorperate the commute past the border, small cities will theoretically have 1/4 of the commute time of large cities.  So, if you want to zone realistically and have low commute times, use small cities?  How dumb is that...yeesh.

 

Anyway, more testing is necissary to fully confirm this, it's only a casual observation at the moment.  I just hope this is not the way maxis tried to make inter-city commutes possible.  It's quite possibly even more unrealistic than using a x12 multiplier to account for the fact that people don't work at the closest job available.

[Skippydam73 0]

T7T,
I played around a little this morning using a large map. I had the avenue only halfway across the large map from R to I and the commute time was 70 minutes. It was 87 till I upgraded from road to avenue.

Also you mentioned about the sim only accessing the road in front of the house and not from the yard. I think I saw in my game that it did still do that. I will have to look in the morning to see and I can post a SS if I find it. I believe the house had a road behind with a street in front of it.

To add to your interesting about intercity, I have had pedestrians cross from their house to the next city. I can post a SS of that too. The one option now I wished we had with the query was to show the route AND the time it takes for their round trip. That would definitely help out more in deciding how long it actually takes for their routes.

[pitch 0]

This is not about pathfinding, but about commuting in generall.

There is no interchange between Avenues and On way streets.

Here is a pic of what I mean:

 

Number 1 and 2 no commuting (No road acces zot). Number 3 commuting is possible

[the7trumpets 10]

----------------

On 9/29/2003 4:08:35 AM pitch wrote:

 

Number 1 and 2 no commuting (No road acces zot). Number 3 commuting is possible

----------------

Now THAT is a problem I know I can fix   Just give me a little time to sort through the sc4path files.

 

However, I'm REALLY Starting to get worried about inter-city commute stuff.  It appears that inter-city commutes spanning more than two cities act extremely weird.  Basically, the demand keeps going up as you get more sims and jobs, but the sims can't figure out how to commute through the middle city.  The result is tha many sims move in, then get a no job zot, and then abandon.

But, lots more testing needs to be done on this inter-city stuff, because it's obviously reacting much differently than in original SC4.