When will goods substitution be fixed?

  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Culture-level substitution values instead of unified ones
  • (disclaimer: the following text is written as an improvement over the original proposition to change the needs fulfillment to the price equilibrium model, as it is seen as unavoidable. The text implies that pops already optimize their purchasing behaviour to minimize costs per need fulfillment, which is not how the game works right now)

    After a lot of thinking on this problematics, I now believe that goods substitution coefficients (like 2:1 for oil and wood for the heating needs in the current implementation, see original post) should not always be the same and should be untied from price ranges (now the ranges correspond exactly like the need fulfillment coefficients).

    More than that, cultural preferences in consumption should be reflected not solely through obsessions and taboos, but mostly through having different substitution coefficients. For example, for the French wine should be much more preferable to tea as a luxury drink, with it being 2 times more effective than coffee and 5 times more effective than tea in fulfilling the need.
    The coefficients should be recalculated yearly to reflect changes in consumer behaviour, with higher coefficients than for the previous year given to goods with higher consumption.

    Obsessions, in their turn, should modify the needs (so the French need more luxury drinks in their consumer basket compared to other goods), and not directly modify the types of goods suitable for fulfilling this need. The desirable type of good bias would naturally come as a result of different preference coefficients.
    Alternatively (for lore reasons, as it makes sense that the French would be obsessed with exactly wine, and not "luxury drinking"), obsessions could still correspond to individual goods, and additionally to boosting the need share in the consumer basket, they could add bias in yearly preference coefficients calculations.

    What would this bring, on the positive side?
    The countries with different cultures (and thus different consumption patterns) would play significantly different and have different effective industry composition, with naturally different demand for various goods. This is a very important outcome, as the problem of "every country feeling the same" is arguably the biggest one in the game at the moment, and I'm really not a fan of hardcoded solutions like "let's add 20 events for each major in which real-life names would be used".
    It would also eliminate the ugly price locks, when for two goods, both used at the time to fulfill a pop need, the price modifiers must be the same (e.g. +34%), which would stem from price optimization for the consumer basket.
    Finally, it would bring some inertia to consumer behaviour changes, quite expected in some cases (like the heating which we discussed here a lot), while still applying economic pressure on pops to switch to more reasonable consumption.

    What problems would it bring?
    First, this would make computations a bit more complex. At a glance, the difference isn't major though.
    Second, the price lock removal would mean that it might be more effective for a pop to fulfill a huge portion of a need with goods in severe deficit, basically non-existant. The prices are capped at +75%, so if a good with base price of 100 at +75% cap (so 175) is twice as effective as a good with base price of 80, the cost per second good can not raise over 88,5 (+10%) while fulfilling consumer demand. It can only raise higher if the industry demand is even higher (so pops would stop buying the latter good). If the former good is in fact in huge shortage, this creates an undesirable outcome not present in current game or in basic price-lock option, when pops could buy goods in reasonable supply, but instead they buy what isn't there (as for pops there aren't, and probably shouldn't be, shortage penalties).

    How to deal with this problem?
    There's always an option to ignore it. +75% prices shouldn't be the norm anyway, and with, hopefully, economic system improvement they should become more and more rare for most goods (all goods except for those experiencing short-term technology-induced demand spike when supply didn't have time to build up yet).
    Alternatively, the possible price ranges could be broadened (like -75% to +300%, so four-times difference in either direction), although it would demand a ton of balance work to ensure that industries are still able to recover from death spirals.
    Ultimately, the pops could be made to stop buying the goods further than max price if some (even ineffective) substitution is available. However, this would make the system much less transparent, will again kill the effective price minimization and would probably be even more complex in calculation.

    Please share what you think on this.
     
    Last edited:
    • 3Like
    • 1
    Reactions:
    Devs' response
  • Hi everyone! In the latest DD's comments Wizz actually shared the reasons for it being implemented the way it is. This comment is purely to cite him, and I'll put my opinion in the next one for ease of further citation and clearer likes/dislikes attribution.

    The most significant reason as to why goods substitution uses supply is that having substitution be based on price leads to an endless back-and-forth where the prices never stabilize. The first version of the goods substitution system (before release) worked like this and we were never able to get it working, by tying it to supply we get something that doesn't change the calculation whenever pops make a small adjustment to their buy packages.

    There's more reasons such as performance (particularly when predicting impact of goods substitution from market changes, using a relatively stable value like supply helps a lot), the fact that people IRL did and do have purchasing habits which isn't just 'buy the cheapest calories I can get' and that supply also to some degree reflects availability of goods. I don't think the current system is perfect by any means, but I do think (based on experience of trying to get it to work for half a year) that trying to anchor it in prices is just not feasible.
    (between Wizz'es messages a fellow member of the public IsaacCAT suggested implementing a logarithmic weighing to counter the problem when one of the goods is huge due to availability and industry demand and the other isn't, so the latter gets very little pop consumption)
    Yeah, I've been thinking about whether using a non-linear formula here might be the way to improve it going forward. It's probably not going to see any major revisions in 1.5 though given the sheer amount of other things we are trying to balance right now.
     
    • 14
    Reactions: