>>At the moment, leaning towards self-tuning, but suspect that the end >>result of attempts towards tuning won't be completely successful. > >If I understood what you're planning, I doubt you'll get enough >transactions to generate a meaningful average. In other words, the >player's input will not be stable, and will tend to drive the values >off the scale.
With a small number of (or zero) players, this will certainly be true. (if you're talking about transactions at shops....) Remember though, I've proposed nothing yet about tracking transactions at stores. The model calls for goods to be produced as players provide raw materials. Players stop providing materials, the shops run out of stuff to sell. I've proposed nothing about changes in pricing either, though attempting to change prices through supply & demand could undoubtedly lead pricing which Just Gets Silly. One possible (and inelegant) solution on that would be to have hard caps on pricing (top and bottom) set as percentages of the 'normal' cost of the item in question... Leans towards the 'Disneyland' feel you mention below. (if you're talking about transactions in the sense of inputs to system vs. outputs in goods, which at this point is simply MOB killings...) The system won't work at all without input. I've got an idea of a baseline (as mentioned in previous postings) where the MOBs in the supply chain each start with enough resources to stay busy for a bit, but the system grinds to a halt without input. It'll take some odd number of MUD days for this to happen, but... (and that's a matter of initial tuning right there... how many MUD days should that be?) Amounts of resources stockpiled due to previous player actions would naturally work to increase the amount of time it'd take for the system to grind to a halt... again how much? question for initial tuning... >I once coded my stock market so when it got to 0, everyone lost all >their shares. This was a very unpopular "feature". I suggest a market >that swings up and down a bit, and allows player to interact by >waiting for better prices to sell, or perhaps being forced to take a >loss due to circumstances, etc., but always returns to a preset >average value. I'd not yet considered a stock market (or anything like one as of yet), though this kind of problem could certainly show up in other places. >My gut feel is, if your economy needs people on to work, then it >might have problems. How many people is enough? If there's an >'enough', doesn't that mean there must be a 'too many'? And as a >purely practical matter, if you need people to make it work, how >will you tune it? Since I'd mentioned possible stabs in that direction before, you now leave me trying to decide if you haven't fully read the previous postings or if you're trying to get me to think things through more completely. I'll assume the latter... At this point I'm pretty much resigned to the idea that self-tuning won't be perfect, and the metrics I'd mentioned before won't be completely adequate. Different metrics will need to be explored, and some code changes will be needed if/when the number of players involved changes drasticly. Yes, I'm resigned to continuing maintenance work rather than having a model that fits every possible situation. Oh, if this weren't so... For the purely practical matter, it will be difficult... Best I can do is try to simulate. As an Imm wandering an unpopulated MUD, it won't be difficult to give the system enough input to see it start working (goto, slay, give self some gold & visit the shops) & watch the results. Won't be as good as actualy having players mess with things though... and I've also got a few testers already. Not sure how much time they'll be able to spend, unfortunately. >Otherwise, I like John Mariotti's approach. It's sort of like >Disneyland. You generate the appearance of a 'working economy', >which will actually seem more realistic than a 'real' one, given >the scale of the setting. True, it's a lot of work, but it can >be very rewarding. As mentioned previously, I like his approach on handling pricing, and may take some ideas from there. If I were to follow the model I've layed out and then follow his model for pricing though, this could lead to some wierd results at stages, since pricing would be completely divorced from actual supply & demand... (okay, it already would be since I'd have pricing as static to start, just seems to be it could this'd make it even wierder) >I wince while writing such discouraging opinions as this. I spent a >lot of time trying to spite older coders who told me things couldn't >be done, and it was good for me. :) Some people think reinventing the >wheel is a waste of their time on this planet, others think it builds >character, but here, I fear you'd be reinventing a flat tire. I'd prefer to look at this more as being warned of potential pitfalls than as discouragement. :) I don't believe this to be an insoluble problem, just I'm not sure of the true degree of difficulty yet. >Sandi "I have not yet begun to code!" >[EMAIL PROTECTED] blades.inetsolve.com 3333 Many thanks on your feedback :) Later, d.c. | dwight a campbell - [EMAIL PROTECTED] - http://rats.darktech.org - KD7KJG | | yet another linux guru in training & reality avoidance therapist at large | | "When you live in a sick society, just about everything you do is wrong." |
