Planning my first Car Tycoon Challenge

[Packbat]

Okay - so I copied out all the stat weightings for Gasmea in today's build (I don't remember 204 or 205) into Excel and screwed around for a bit trying to pick out categories which are both (a) a wide variety and (b) fairly distinct from each other. My first tentative list:

  Code:

        Driv  Spor  Comf  Pres  Safe  Util  Prac  Offr  Reli  Econ
Offroad  ==    --    ==    ==    -     +     ==    ++    +     ==
Utility  ==    --    ==    ==    ==    ++    +     ==    +     +
Commut.  +     --    +     -     +     --    +     --    ++    ++
Family   +     --    +     -     ++    --    ++    --    +     ==
City     ++    --    -     -     ==    --    ++    --    ==    +
Muscle   --    +     -     ++    ==    --    ==    --    --    --
Track    -     ++    --    --    --    --    --    --    ==    --
Sport    +     +     ==    -     ==    --    --    --    --    --
Luxury   ==    -     ++    ++    ++    --    --    --    --    --

(note: no stats are negative - minus-signs just mean a stat is underweighted or unweighted.)

My thought is, by basing my stats at least somewhat on the in-game stats, I can make it easier for people to design to target demographics - the numbers won't be exactly the same, but a car fine-tuned for Muscle is going to be better in Muscle than a car fine-tuned for City.

[koolkei]

getting more complicated. and excited every update

this might gonna be the very first challenge that accounts for tooling cost and engineering time and cost

[Packbat]

Hmm, interesting - the engineering time limit puts a pretty hard cap on how expensive a car you can build. I'm still planning to use the budgets for the in-game demographics as a guide, but I'll have to make things a little bit tighter at the high ends of the distribution to make those decisions meaningful.

[Leonardo9613]

Finally went through this. And it seems mostly solid, apart from a couple of things. Firstly, desirability/value being given in dollars is just wrong, it feels wrong, maybe make it a "adimentional" value. Secondly, how do you plan to input all of this data? By hand? Bayer had a lua reading algorithmm so the process was mostly automated, putting all of the values and planned calculations by hand would probably be a pain in the ass.

As I said, apart from these minor issues, the planned market modelling looks pretty straight forward and effective. Also, that gives a suggestion to the devs, why not make those charts as part of the market screen? It's a lot better looking that the current list of percentages.

[Packbat]

Re: desirability in dollars: I guess I can see how it is surprising, but it's basically akin to the economics idea of consumer surplus: the customer paid $X when $Y is how much money they would have been indifferent about spending ($Y+1 and they just go without), and the difference is how good a deal they got. It's a little less straightforward in this case because I wanted to measure the annual return on investment - basically, how much they might have spent to rent the car for a year - but the idea of measuring preferences in money is pretty standard.

Re: inputting the data: ...I have a high tedium tolerance and a lot of free time. I'll definitely come up with a better system for future competitions, but entering 30-50 cars over a couple weeks doesn't seem daunting to me. I have a good workflow.

Re: data viz.: the in-game calculations have 10-14 numbers, and not always the same 10-14 numbers, and there's not a lot of room - trying it isn't a bad idea, but I don't know that it's feasible. Also, it was really convenient to me to have the real numbers.

[koolkei]

soooo any possible ETA to when this will be opened?

[Packbat]

Hmm - let me make a to-do list:

• Final balancing - one of the test cars I built is performing a bit too well, and I need to figure out if the scoring system is broken or if it's just that good. Also, I haven't set Idiosyncrasy scores for the market demos.
• Fix the break-even calculation sheet on the Market Calculation spreadsheet I'm making for the players to use.
• Make the spider charts for the market demos.
• Make the contest post.
• Run a test of the full calculations using my test cars, so I know I have the method down pat.

...I think if I can get the first four cracked in short order, I'll start the contest tomorrow, and let entries run from December 11th to January 11th - four weeks plus one weekend - or until the maximum number of slots is filled. Otherwise it's starting next week.

[Packbat]

...interesting. Developments since this morning:

• I've decided to change the cost-sharing rule to be larger plus one-third smaller, instead of larger + half smaller. The latter makes it too expensive to share production lines.
• (That said, shared production lines still have two major advantages for a manufacturer: first, you can make the 99-trim problem work in your favor, at least a little, and second, it reduces excess capacity if one trim is successful while the other is not.)
• I learned why 0% in-game markup is where it is - it's hard to break even on a car in the first year! Granted, cars are usually built for more than one year, but they're also usually given refreshes quite frequently, and generally speaking investors expect some return on their capital.
• ...that said, although I was sorely tempted, I refuse to compromise on the principle of the five-day work week! ...at least for this competition. Twenty-first century labor management practices will have to wait.
• Also, thanks to timeanddate.com's Business Days Calculator, I know how many workdays there were in my target year - at least, in the US. (Country-by-country variation will also have to wait.)

Assuming I don't hit any major roadblocks with the spider charts and the writing, I can probably get this started sometime tomorrow - probably not in the morning my time, but sometime.

[Lordred]

Godspeed Packbat.

[Packbat]

Okay, we're scrubbing tonight's launch.

As I said, I wanted to run a CTC in the spirit of yore - which is to say, in the spirit of Der Bayer's CTCs - but in my head I was assuming that the entire game would take place over the course of one year. What I see when I look back at Der Bayer's competition, though, is entrants breaking even on investment with factories running at 20-25% of capacity, whereas in my game right now, even at a 30% markup, you would need to run at 90% of capacity to break even in one year ... and going from one factory to the next larger usually means a 50% increase in capacity. If I ran the contest in the current rules, most entrants would either go broke or sell out their entire runs. That might be realistic, but it's not fun. (Especially given how much work I would have to do to handle all the sold-out cars.)

Plus, I designed all the market demos looking at cars sold with 0% markup, and for the cheaper demographics, 0% to 30% markup is a big difference.

The solution might be as simple as explicitly making the contest take place over a three- to five-year run for all models entered, but I want to make sure it is (and decide if I want to revise the market demographics) before launch.

New ETA is next week.

[Leonardo9613]

The CTC assumed a three year cycle, so the tooling and engineering costs were diluted over that time.

[Packbat]

Aha! That would explain it.

...I'm still going to take the time to make sure I've got my ducks in a row and do it right.

[Packbat]

...well. These spider charts are going to be a lot weirder than previously anticipated. I designed the criteria weights in each demographic based on the in-game markets and my experience with the Hoon Competition, and ended up with weights ranging from $250 to $10,000. The final charts look less like the original example I posted and more like ... well:

Muscle-Super radar chart.png (12.28 KiB) Viewed 5833 times

Edward Tufte, please forgive me.

(Yes, I know Utility and Practicality are switched. What with all the formulae already programmed in, I don't know an easy way to fix that, and I don't want to introduce bugs trying.)

[koolkei]

take your time. as we all know by now. delayed is still better than unfinished.

[Packbat]

Appreciate it.

I want to make sure that the competition is open for a while but still over before the next major update, but last I heard they were talking February for that, so I have breathing room. It might not go up tomorrow, but if I can start it before the New Year then I'll be happy.