I've been playing a game of Gearcity on the 'hard' level for the last couple of weeks. I'm really enjoying it, but I have some concern with the AI. Almost all of the competition is continuously making a loss. I understand (from reading a post elsewhere) that this is probably mimicking reality, but it makes is very difficult for a human player.
My main bone of contention is that these companies that are performing poorly, are under pricing their vehicles. They seem to sell them far too cheaply. This means that when I'm selling vehicles in the same segment and market as the AI, I can't compete.
With the current number of AI competitors, this is manageable. I'm concerned when the game has it's full complement of car companies, that the AI behaviour will make it next to impossible for a human player to compete.
AI has been significantly changed in 1.11, so wait for that release before making conclusions. (It's still a little weak, but much better than 1.10)
I keep trying to find a balance between turn speed and AI strength. 1.10 was way too geared for speed and not strength. 1.11 is probably too geared to strength, as it's probably too slow to support 500 companies.
We'll find the right balance at some point.
(01-11-2014, 07:54 AM)Giacomo Wrote: [ -> ]I've been playing a game of Gearcity on the 'hard' level for the last couple of weeks. I'm really enjoying it, but I have some concern with the AI. Almost all of the competition is continuously making a loss. I understand (from reading a post elsewhere) that this is probably mimicking reality, but it makes is very difficult for a human player.
My main bone of contention is that these companies that are performing poorly, are under pricing their vehicles. They seem to sell them far too cheaply. This means that when I'm selling vehicles in the same segment and market as the AI, I can't compete.
With the current number of AI competitors, this is manageable. I'm concerned when the game has it's full complement of car companies, that the AI behaviour will make it next to impossible for a human player to compete.
Edit: I see Eric.B responded before i posted. I think this post is still relevant though.
This should be an interesting discussion.
I agree on the mimicking reality part. Ive heard it mentioned many times that most of the Car companies failed at this time.
Part of the problem is that many of the yet to be implemented AI you mentioned are some of the largest in the business. I think the massive Ford Company has been mentioned for example.
So currently what we are seeing is a large number of historical failures
Ill comment more on balance later probably, but for now, i think a lot of it depends on the intention of the
game designer.
Are the current companies supposed to be mostly successful or even partly successful?
Are the companies which will be added later supposed to be the large successful ones?
Finally, a suggestion
I think it might help to provide a clear example of a time the AI is "underpricing" their vehicles for a loss.
You might also want to mention the quality of the vehicle and the quantity of them being sold compared to others there.
This will help clarify your point as well as eliminate other explanations for the AI behaviour.
Arakash.
Good point. It could just be my perception of the AI selling too cheaply. I could probably go through the car review magazine, find the cost of the AI models and compare these to the prices they are selling them at. I've made the assumption that the companies have been geared to fail (in the historical fashion), by selling their vehicles too cheaply.
Anyway, looking forward to 1.11. I'm desperately trying to finish my 1.10 game before the next version comes out. I'm a micro manager and I can spend ages on each turn!
If I'm not mistaken the AI would reduce price down to 10% above unit cost if it could not meet demand.
This has been changed to a much higher number, around 40% I believe (In line with retail store markup at least)
However I've made the AI much more aggressive with price increases and fixed some bugs in production expansion. This should solve two issues, 1) AI not building enough cars, 2) AI not being profitable.
It still needs some work, but it's getting there.
(01-11-2014, 01:55 PM)Eric.B Wrote: [ -> ]However I've made the AI much more aggressive with price increases and fixed some bugs in production expansion. This should solve two issues, 1) AI not building enough cars, 2) AI not being profitable.
It still needs some work, but it's getting there.
I hope i don't sound like im exaggerating by saying that sounds fantastic. Im looking forward to seeing it.
Feedback on the AI would be appreciated.
This should be another "suggestion" thread response but it seems to me that a report that shows all AI competitor car models and prices against your own would be good. Again, data is king in a business sim and you need to know what the competitors are doing in order to make changes in your own business model.
Going through the magazine for each of the 200 AI companies would be tedious.
A report that compares you and say the top 5 companies at any given city PLUS an average car model price would be all that is needed. Then you can decide to sell higher or lower or at the average price.
You can check AI competitor models by going to the world map and selecting a city you're interested in. Select "Competition". Then select a vehicle type and it'll show you all the vehicles sold in that city of that type. Along with some data.
I think a competitor report has been suggested already in another thread(In one of the Suggestion forums i might add), i remember because i was trying to suggest what it should contain.
I will add some other methods i use to monitor the Competition to what Eric has already pointed out (that you can get information from the world map competitor button in each city).
There are also Generated reports that give you information on your competitors.
Notably:
- Best Selling Vehicles of this Year
- Best Selling vehicles in (region)
- Vehicle class market share leaders (a very interesting report that shows your nearest competitor.)
Some of the market share charts are also useful, though most(like share in city) dont track market share over time, which limits their usefulness.
I plan to comment more on the AI once ive played the new version a bit more.