引用自 stevo：

The AI popularity modding is not documented properly so far, hence my questions, even before the patch comes out.

Seems fairly well documented to me.
http://wiki.gearcity.info/doku.php?id=modtools:componentseditor#ai_pop

1) I assume that the order of records in the table is not important? What I would like to do is to set the AI popularity of component A year by year, then - component B year by year, etc. Is this OK, or do I have to sort everything by year?

Only the order of the components' year matters, order of the components do not matter. So you can have 1900 A, 1910 A, 1920 A, 1900 B, 1910 B, etc OR you can have 1900 A, 1900 B, 1910 A, 1910 B. You shouldn't have 1900 A, 1920 A, 1910 A, as it will lead to interpolation errors.

If I am not mistaken, the mod tools should automatically sort them for you.

2) The popularity of vehicle types and fuels among customers is always map-specific and stored in the turn events file - is this correct?

No, only manual vehicle type popularity changes are specific to a map. You can make specific changes on specific dates for fuel popularities in turn events files. You can also do that to other components, but it won't have much in game effect.

Is there no way to set it for any map in a separate file?

Vehicle popularity is defined in Car Types Tab, Fuel popularity is defined in Fuel Tab (erroneously named AI Popularity, It should be consumer popularity).

The componentsstartingAI file is only for AI's/Assisted Designer sub component selection. It has no effect on player related stuff. It's used to make the AI make historical component selections.

The exception to this is the Fuel Pop Index (Table calls it Overall Pop). Is also used to interpolate fuel popularity automatically. So you don't have to implement a lot of data in the Turn Events Files for every map.

Please note, fuel popularity is dynamic in the game. While you can manually make a fuel popular using interpolation, the dynamic markets in the game can counter that based on how popular fuel types are in the game.

Any changes to vehicle popularity outside of the starting values and the dynamic systems has to be done at the turn events files, due to the fact that not all regions are the same in every map.

You can modify specific map turn events files with mod tools. However, the mod will only work on those maps.

3) Am I right that I do not have to set the popularity for EVERY year and use interpolation instead?

That's what interpolation means. (End Value - Starting Value) / Years. It will grow or shrink that much each year.

4) I also think that some randoness is included? E.G. when I have several turns with the same popularity value, and no change in proces / competition (competition only nakes decision once every three months, if I remember correctly) the sales will fluctuate a little anyway?

There is a bunch of randomness in the game. It would be pretty cut and dry and boring otherwise. These component pop files are just suggestions to the AI on what components to pick for their designs. No more, no less. The exception is the ability to interpolate fuel popularity. Fuel Popularity is also influenced by the map's fuel prices, in game fuel market share, random events, and specific turn events.

Thanks for all the explanations.

I am a little lost, though,

Can you please tell me which files and tabs I should edit to do the following:

-set and adjust car types and fuel types popularity among buyers over time (this is not region specific, I suppose?) - is this in the turn events file? I can't see anything about fuel types there, only about car types,

-set and adjust AI popularity of all other components? Is this only the components file/AI Popularity tab, or maybe somewhere else as well?

Thanks,

引用自 stevo：

-set and adjust car types and fuel types popularity among buyers over time (this is not region specific, I suppose?) - is this in the turn events file? I can't see anything about fuel types there, only about car types,

As stated in my previous reply:

Starting Vehicle Popularity is in the "Car Types" tab.
Starting Fuel Popularity is in the "Fuel" Tab, It is the "AI Popularity" spinner box.

Vehicle popularity changes over time is done in the map's turn events file. This is in the Turn Events Editor, "Car Popularity" tab.

Fuel Popularity changes over time is done in the map's turn events file ("Components Tab" in the Turn Events Editor) OR using the "Pop Index" spinner box in the "AI Pops" tab with the Components Editor.

-set and adjust AI popularity of all other components? Is this only the components file/AI Popularity tab, or maybe somewhere else as well?

AI component selection (aka, AI Popularity) is the "AI Pop" tab.

OK, then. Which file takes prioroty over the other? Turn Events or Components.xml?

Putting the market share influence aside, is it enough to edit the Components.xml (the "Car Types", "Fuel" and "AI Pop" tabs. ), or will it get overridden by the turn events?

Also: will it cause any errors if the dates in the popularity lists do not match the start/stop year of a component? It will be easier for me to change the popularities every 5 or 10 years for everything rather than observe the exact dates for every component - I hope it won't crash the game?

Thank you for everything!!

Components.xml is the starting value.
Let's say sedans are set to North America 0.10 (10%) in the components.xml file.

You start a game in 1900. Sedan's popularity will be 0.10 (10%).

Now if you put 1.01 in Region 1 in the turn events file on 3/1901. Then when you get to 3/1901 in the game, the new popularity will be 0.10 * 1.01 = 0.101 (10.1%).


Let's say you put 0.5 (50%) for fuel's gasoline's popularity in the components.xml file. When you start the game in in 1900. Gasoline's popularity will be 0.5 (50%).

Let's say you interpolate gasoline's popularity from 0.5 to 0.75 from 1900 to 1925 in the ComponentsAIPop file. That means every year it will go up 0.01 (1%) in popularity ((0.75-0.5)/(1925-1900)). So when 1/1901 rolls around it will be 0.51 (51%) when 1/1902 happens it'll be 0.52 (52%), etc. If in the turn events files you have an event that has a 0.95 effect on Gasoline on 3/1902. Then on 3/1902 gasoline will be 0.52 * 0.95 = 0.494 (49.4%) in popularity.

Hopefully that clears it up for you.

In summation. Components.xml is the starting popularity. Turn Events are for events that happen on a specific turn. Interpolating fuel popularity in the componentsaipop file is the automatic way to adjust fuel popularity yearly.

Also: will it cause any errors if the dates in the popularity lists do not match the start/stop year of a component? It will be easier for me to change the popularities every 5 or 10 years for everything rather than observe the exact dates for every component - I hope it won't crash the game?

No it won't crash the game. It is used to generate interpolation values as mentioned before. And only for fuel types. Everything else has zero bearing on how popular a component is. It's just for AI companies picking what to use.

OK, thanks, it's clear now, both files introduce separate factors.

The turn events factor - 0,95 in your example - does it work only in one specific month (turn), or does it stay there and modify the final populairty till the next entry in the Turn Events file?

Turn events are events that occur on specific turns.

Good, thanks!

I am so glad to see the update!!

The engine layouts work fine now, but:

1) the AI popularity records still can't be removed in the mod tool,

2) the stopping distance works as it should and this is great. On the other hand - the safety/performance ratings are not affected - there's just a tiny difference in Luxury rating of vehicles that have chassis with very different braking parameters. Even though they have significantly different stopping distances. I assume that demand is calculated based on ratings only, not directly on figures like max speed or stopping distance?

引用自 stevo：

1) the AI popularity records still can't be removed in the mod tool,

It's working fine for me. Click the record you want to remove in the table, then hit "Remove AI Component Pop".

How are you trying to do it?

2) the stopping distance works as it should and this is great. On the other hand - the safety/performance ratings are not affected - there's just a tiny difference in Luxury rating of vehicles that have chassis with very different braking parameters. Even though they have significantly different stopping distances. I assume that demand is calculated based on ratings only, not directly on figures like max speed or stopping distance?

As mentioned in the previous thread, here is the formula for vehicle performance and safety.
https://psteamcommunity.yuanyoumao.com/app/285110/discussions/14/1743343017623880533/#c1735463620089892888

Braking has a small effect of on the ratings. 5 rating points at 50 ( off the top of my head i'm not sure if that's in meters or feet) and linearly less every unit above that. What are your actual braking numbers? If one is 240 feet and the other is 200 feet, there will be less than 1 rating point different between the two, which you will not be able to see due to ratings being integer values.

Anyway, this system is working fine, I verified the numbers after the fix for your last complaint.

1) Double checked now - still can't remove the original records in AI Pop tab. I CAN remove those I added myself, but not the original ones (and they refer to components that don't exist in my mod any more). I'll try to edit the file manually and see.

2) My two vehicles stop in 70 and 60 metres respectively. It's a pity that the ratings don't reflect this, because IRL this was something extremely important for earlier cars, cable brakes were terrible and it made a huge difference (I've driven cars from the 20s and 30s myself, with mech and hyraulic brakes).

But I will cope with it - I am going to differentiate manually the performance ratings of respective subcomponents, on top of the braking ratings. This should help.

BTW, I am satisfied with the results my mod gives for components, but after I assemble a car from them, the differences in respective fields (performance, fuel economy, safety, etc). seem to fade away. Because of that I will probably introduce even wider rating gaps between subcomponents - they might look weird when looking at bare components, but will make more sense with complete cars. And GearCity is all about cars, not components :-)

Sorry for the late reply,

引用自 stevo：

1) Double checked now - still can't remove the original records in AI Pop tab. I CAN remove those I added myself, but not the original ones (and they refer to components that don't exist in my mod any more).

See, that's different than what you were telling me. You were saying you couldn't remove anything. Details matters. You can't remove things that have been removed from your components.xml files. I did not test against this, because I didn't know it was an issue. If you email me your two files, I can duplicate it and resolve it for SP4.

2) My two vehicles stop in 70 and 60 metres respectively. It's a pity that the ratings don't reflect this, because IRL this was something extremely important for earlier cars, cable brakes were terrible and it made a huge difference (I've driven cars from the 20s and 30s myself, with mech and hyraulic brakes).

10m/30Feet is not a whole lot of difference. Braking is 5% of the safety and performance ratings. I don't know too many people who bought vehicles for their braking specs. It's not really that important of a metric so long as its in reasonable range.

That being said, the spec values are specifically used for racing.

BTW, I am satisfied with the results my mod gives for components, but after I assemble a car from them, the differences in respective fields (performance, fuel economy, safety, etc). seem to fade away. Because of that I will probably introduce even wider rating gaps between subcomponents - they might look weird when looking at bare components, but will make more sense with complete cars. And GearCity is all about cars, not components :-)

Just remember most components get capped at 100 ratings. The vehicle's use the component's ratings more than sub-component ratings. None the less, the sliders for the vehicle make up at least half of the vehicles ratings. So that may be the issue you're running into.

As for the AI Pops - sorry for that, actually I didn't realize this myself, because earlier on I had been only trying to remove the old records, without adding new ones. And I didn't know it mattered.

Regarding the brakes: you are 100% correct if we only speak about modern times. Today no one cares about the brakes because they always work as we expect. It is never an issue any more, but it wasn't the case e.g. before WWII (and Gearcity is all about car HISTORY, so it should'nt ignore it). People were installing SEVERAL additional brake systems to Model Ts to make them adequate for city traffic in late 20s (I recommend you Google "Model T Rocky Mountain Brakes" to give you an idea). Also after WWII the difference between simple hydraulic drums and servo-assisted disc brakes was enormous (just try to compare early VW Beetles with a Citroen DS, they were both on the market at the same time). Having tested about 70 classic cars from anywhere between 1926-2016 for my blog I am maybe too much of a freak, but I find it hard to accept that this game does not recognize one of the most important fields of automotive technology development. Of course, this is nothing against you or the game - I am just trying to find out a way to fix it and I have already discovered a bug. That's why I prefer to share all my concerns with you. I will just keep on tweaking the numbers :-)

You are very right that the sliders play a major role. Actually this is OK, because they also influence the costs greatly. I already figured it out with fuel systems: I couldnt avoid getting ridiculously powerful late-game engines until I nerfed the power of the moden injection systems (I made them WEAKER than the old carburetted engines), reducing their cost at the same time. This way, when you adjust sliders to get the same cost, modern injection systems give higher power, as they should, but you cannot make it outrageously high (bacause the slider is already at 70 % when you reach the same cost). Maybe I'll use the same method here.

Thanks for everything!!

引用自 stevo：

As for the AI Pops - sorry for that, actually I didn't realize this myself, because earlier on I had been only trying to remove the old records, without adding new ones. And I didn't know it mattered.

No worries, I will work on it for SP4.

Regarding the brakes: you are 100% correct if we only speak about modern times. Today no one cares about the brakes because they always work as we expect. It is never an issue any more, but it wasn't the case e.g. before WWII (and Gearcity is all about car HISTORY, so it should'nt ignore it). People were installing SEVERAL additional brake systems to Model Ts to make them adequate for city traffic in late 20s (I recommend you Google "Model T Rocky Mountain Brakes" to give you an idea). Also after WWII the difference between simple hydraulic drums and servo-assisted disc brakes was enormous (just try to compare early VW Beetles with a Citroen DS, they were both on the market at the same time). Having tested about 70 classic cars from anywhere between 1926-2016 for my blog I am maybe too much of a freak, but I find it hard to accept that this game does not recognize one of the most important fields of automotive technology development.

The majority of early cars were not hitting the required 100km/h to do the game's 100-0 tests. The game gives you the maximum as long as you stop in a reasonable amount of distance. That's 150feet at 60-0mph. Using your model T as an example, it goes approximately 190 feet at 45-0mph. As such, from 60-0 it'd would lose 1-2 points out of 5 for braking.

There isn't much more I can do but adjust those numbers. But that would at the same time weaken everything else. Is braking a more important safety measure than the chassis frame? Or the vehicle design? Or safety features such as seat belts, air bags, etc? If so how much of a difference is it? Also how does it balance with the rest of the game, weight is an important con of the safety sliders for a vehicle. Braking systems are not that weighty, thus making everyone max out safety more because there is less cons...

Of course, this is nothing against you or the game - I am just trying to find out a way to fix it and I have already discovered a bug. That's why I prefer to share all my concerns with you. I will just keep on tweaking the numbers :-)

I don't mind, I have to push back however, because 1) I can't change the game at everyone's whim every time. 2) Pushing back tends to lead to better ideas that are more thought out. (As you try to persuade me.)

Thanks for everything!!

Quite welcome, I look forward to trying your mod! :)

Hi,

I am afraid I need to report a problem again. My game crashes after ending a turn. Can this be a proble with my mod...? I am sending you all relevant files in an email.

Also, I noticed that again, in late game the rpm values are through the roof. Small gasoline engines with default slider positions achieve max power at over 7000, while maxed out sliders give over 11000. Is this intentional? We had this problem some time ago and it seemed to have been fixed, now it looks as if it was there again.

Thanks!