Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Random_Dragon]

So, I found myself regarding the value of coins in Dwarf Fortress, especially in adventure mode, and the kind of buying power they have in-game. What better to compare them to ancient Rome? Tempting as it was to use the values of the olden days, when 1 denarius was 25 aurei, it's hard to get a reliable source of the prices for goods back in those days. So instead, I used a more extreme example of value, or the lack thereof, to compare DF coinage to.

As usual, if my math, DFWiki magic, or historical accuracy is FUBAR in some way, feel free to point and laugh.

EDIT: Some corrections to be found. Seems I was using half as much of volume as I should be, should use the k. mod. instead of the regular modius. This sharply changes the implications of DF flour pricing.


In response to the growing inflation and poverty in the Roman Empire, in the year 301 emperor Diocletian issued the Edict of Maximum Prices. This laid down price limits for various goods, which if exceeded would lead to harsh penalties for both buyer and seller.

Figures listed where given in a unit termed denarii communes, a form of notational currency. The physical silver denarius had varying values, weights, and silver content through the history of Rome, steadily devalued over time, and therefore d.c. was used in several imperial edicts to provide a consistent legal currency, and the coinage of the time would be valued in relation to it.

The values set for grain were based on the modius castrensis, a unit of dry measure (volume). This was roughly 12.93 liters. With handfuls of powder having a volume of 200 milliliters, that's 64.65 handfuls of powder in DF terms.

Unlike with flour in DF, the prices given varied from grain to grain. Oats and hayseed were relatively cheap 30 d.c. per k. modius. Sesame and hulled rice were an impressive 200 d.c./k.modius. Several items were set to a higher value if processed first, as the latter indicates. Hulled millet bumped up the value from 50 d.c. to 100 d.c.; while broad beans were 60 d.c. unshelled, 100 d.c. shelled.

But luckily for us, many of the prices for dry goods were fixed at a nice 100 d.c. per k.modius, including good old wheat. So the net result of this was that 1 nice, DF-style handful of wheat would run you about 1.55 d.c.

So, what's that mean compared to how much value a silver coin has in Dwarf Fortress? In adventure mode at least, every civ gives silver coins a value of 5 dorfbux, one hell of an increase over its raw material value. Handfuls of powder have an item value of 1, while the various flours tend to have a material value of 20. So, that's 4 silvers if you want to buy a handful of flour, while that modius would be an obscene 258.6 silvers (assuming we equate 1 silver as 1 d.c., second post gives a good reason not to).


But what about other goods? Meat's one instance where the average adventurer can live like a king, even if they aren't hunting their own game. The prices for meat in the Edict were by weight rather than volume, specifically the libra, equivalent to about 328.9 grams. 1 libra of beef or goat was fixed at 8 d.c, while chickens and several other animals were sold individually instead of by the pound. Pork, lamb, and venison were set to 12 d.c/libra. This would be a good price for our average, since most of the domestic animals have the same values for their meat.

1 piece of meat in DF has that familiar old 200 ml volume, while muscle has 1060 density, for a weight of 2120 grams, or roughly 6.446 libras. Yes, that value made me say "holy shit" too, until I realized that meat does indeed say it's 2 kilos in your inventory. If we assume that this density is accurate, this means a libra is about 31 milliliters. So by Diocletian's Edict, the 100 d.c. that gets you almost 19 liters of wheat would buy you around 258.33 milliliters of pork. By volume, pork was almost 50 times more expensive than wheat! Yes, this is way worse than when I thought these prices were using the Italian modius.

Now, compare to how meat has a material value of 1, and an item value of 2. That means 1 silver will buy 2.5 units of meat, yielding 5300 grams. That means in Dwarf Fortress, wheat is 10 times more expensive than pork by volume.

What about some booze? Wine is a classic Roman favorite. Prices were fixed for liquid measures in sextarii, being 546 milliliters. Since DF booze doesn't have quality levels, we'll use the figure of 8 d.c. for a sextarius of "vin ordinare" or ordinary wine.

Good old grape wine in DF has a material value of 2, and 60 mL of liquid has an item value of 1. So that sextarius would cost 1.82 silvers in DF, which is even cheaper than the price limit on beer.

Speaking of which, how about some nice, cheap beer? Want barley, or wheat? The former is a mere 2 d.c/sextarius, while the latter is fixed at 4. With most of the beers in DF having the same 2 material value as wine, so the value there is perfectly within the limits set by the Edict.


What about some clothes? Cloth was often rather expensive back in the day. You could get a winter (most likely wool) tunic, meant for a solider, for 75 d.c. Or how about a libra of wool for the same price? Or for only 72 d.c. you can get a libra of the lowest-quality linen. Want some silk? 1 libra will cost you 12000 d.c. Don't even ask about having it dyed purple. That's for use at the direction of the emperor only (on pain of death, naturally), and even then it'll cost a hefty 150000 for those 328.9 grams.

In DF, a no-quality sheep wool tunic has an item value of 16, and wool has a material value of 1. Pretty cheap. 1 stack of freshly-woven cloth has a volume of 30 and item value of 7, which at wool's density is 150 grams, and a value of 1.4 silver (making a libra of wool about 3 silvers). Meanwhile regular old cave spider silk is the same price (what a steal!) while even GCS silk is only 12 silver to the libra. Linen has a material value of 2, so that's only 6 silvers per libra.

As for dying, no purple dye yet, but midnight blue dimple dye is perhaps the closest in color. Most of the dyes have a material value of 20, and (if the wiki's correct) adding dye counts as adding the item value of powder (1), but (unlike the norm) is subject to quality multipliers. So assuming your dwarven king orders a patch of master-quality cave spider silk dyed with masterfully-applied dimple dye, that's only 76.8 silvers per stack of cloth, or 168.39 silvers per libra.

Want some leather? Sadly the sources I could find didn't give an explicit volume for the prices on hides, but leather in DF is also quite nebulous in how much you can do with any given piece of leather, which rabbit and oxen leather both having the same utility.

Tanned oxen hide of second quality is listed at 400 d.c., while tanned deer skin is maxed at 100, wolf leather at 40, and goat leather of the largest size (whatever that implies in the translation) being 50.

In DF, leather has a fixed item value of 5. Oxen being a term for particular sort of bovine, we'll use cows for that. Lacking any changing to leather's default material value of 1, the whole hide will only cost you 1 silver. The same applies to goats and deer. Wolves have a material value of 2 applied, so this means their pelts will fetch 2 silver.


So now, a while back I mentioned the value of non-silver coins, with DF gold coins being only 3 times the value of silver coins. The gold coin issued (though in limited numbers) by Diocletian was the solidus, containing around 5.5 grams of gold. This replaced the aureus that had varied in size over time, but (unlike the denarius) retained more consistent purity. When Caesar was around he standardized the size of the aureus at about 8 grams.

Whereas in the early days 25 denarii (having 4.5 grams of silver when introduced) made for 1 aureus, Diocletian tariffed the solidus to 1000 denarii communes. By this point the silver content of the denarii in use was a mere 5% in a coin weighing 3.41 grams. That'd make for 170.5 grams of silver to equal 8 grams of gold, or 21.31-to-1 ratio by weight.

So in the old days, those 25 coins made for 112.55 grams of silver to 5.5 grams of gold, for a 20.45-to-1 ratio.

Now for Dwarf Fortress coins. Each coin has a meager 0.0625 mL volume, making the silver coins 6.55 grams in weight and the gold ones 12.075 grams. The end result is a by-weight value ratio of roughly 1.627-to-1.

Not that this trivia ignores some important considerations that're covered in better detail in my second post.


So what's there to learn from all that? Honestly, not much. Amusing as it'd be to make a "realistic pricing" mod, there's little real need for such. If one intends to do so, then perhaps that'd be useful information for someone.

Other that that, it's more for some amusing comparisons, because I am a neeeerd.

Source, english translation of Ēdictum Dē Pretiīs Rērum Vēnālium: https://www.academia.edu/23644199/New_English_translation_of_the_Price_Edict_of_Diocletianus

[Random_Dragon]

So, what would I do to mess with the prices of goods in DF, if I were to mod the values? There's a lot of reasons why I wouldn't want to use these values as a guide, but one can certainly try.

For starters I'd want to alter the values of coins, to a nice Roman-looking ratio of 1:10:250, or maybe 1:10:200 since gold is somewhat easy to get ahold of in fortress mode, with a little luck.

Then I'd want to adjust the material values of silver and/or gold. However, I don't want to make them perfectly consistent with their coin values like they are currently. The reason is that'd make the material values 2, 20, and 500. That's way overboard.

However, it helps if one considers the fact that DF coins are all the same volume, making the relative values by weight different. If I based the values off of that they'd be (very) roughly 2, 16, and 140 instead. Still rather high, and a bit weird. If I did set values, I'd much rather devalue gold to something under 100. While it'd be arbitrary, it's at least partly excusable due to the availability in fortress mode.

The only real goal is to think things father ahead than Diocletian did. The Edict valued a libra of silver at 6000 d.c., presumably of high purity. If you assume that the denarius of the day is roughly equivalent to the denarius commune (which may or may not be accurate, hard for me to get good info), it'd only take you almost a couple thousand of those 5% denarii to get that libra of silver.

In-game of course, coins can already be abused via melting to get more bars than were used to make them, so by game logic breaking the laws of physics can't be avoided, let alone breaking the economy. So fudging the numbers so you couldn't get more value out of the coins by melting them in reality (not in fortress) is more a flavor detail than a practial one.


Setting explicit values for goods using all this math won't work very well for a couple reasons. Like I said earlier, this is more of a mental exercise, so taking these values as a guide is a bit silly to begin with. The first problem is of course that DF coins have a higher purity than the 5% denarii of Diocletian's day. So right away the values, if applied, should be 1/20th what the Edict suggests. That's already ignoring the different masses of the Roman coins versus DF ones, and the fact I can't find any confirmation as to whether a denarius back then was in fact equal to 1 d.c. at that time.

The second issue is that real-world volume and weight doesn't mean much in DF terms. Since we're focused on adventure mode, consider meat as an example. That one piece of meat may weigh over 2 kilograms, but adventurers will happily scarf it down in one bite, and they tend to eat about 3 or 4 pieces a day. Compare that to it being over 6 libras in weight, and compare that with the common Roman subsisting on 2 libras of bread a day instead, and pricing it off of real-world weight becomes absurd.

The values instead would need to consider in-game units that amount to the same. One piece of meat might be priced as a libra or as half a libra. In contrast, with a modius of wheat making enough bread for a week all the way up to 10 days, the real-world figure isn't that off in Df terms, with 40 handfuls of flour likely lasting 10 days if the adventurer had some way to render it edible, or if flour could be eaten raw.

Wool and linen could be priced more sanely, enough to make them rather costly, while leather would have to be priced very selectively. One positive of this is that the hides of larger domestic animals, often being more expensive in the Edict, would have their prices better reflect how much more costly they are to produce in fortress mode, as many are slow-growing grazing animals.

Silk however is right out. Pricing it relative to wool or linen would produce something massively expensive. It also wouldn't be realistic. You don't have to import silk from faraway lands, and production of cloth in general is easier than in real life back then. To a dwarf, silk's main limitation is having to gather it instead of growing pig tails or sheering wooly livestock. They could still be made more expensive than hair or plant cloth, but not excessively so.

Finally, booze. Once again, how long a given amount of liquid will last an adventurer needs to be considered. What info I could find suggests that the average Roman would have to make 13 or 14 sextarii of wine last him a month, which is easier when you consider that it was normal to water down that wine to some extent for actual drinking, as much as a 3:1 ratio of water to wine at home.

Of course, an adventurer drinks it straight, like a barbarian. Since adventurers tend to need about as many units of liquid as they need units of food, and 1 sextarius is about 9.1 DF units of liquid, that means they tend to go through that 13 sextarii of wine in a bit over a month, so the much lower consumption of liquid cancels out the lack of dilution.

So, all that math just to say that the price of booze likely doesn't need to be changed much, if at all. Wine is cheaper and lasts an adventurer a bit longer than it'd last a Roman. It could be make a bit more expensive without much consequence, but it isn't worth the hassle.

[Larix]

That's an interesting source, but i think the edict is only good for getting a rough impression of relative values. It was a maximum price edict, not a list of prices actually found in the marketplace. Especially with agrarian products - especially grain - fluctuations in supply caused massive fluctuations in price regularly, so the maximum price may well have been something like double the usual (and three or four times the minimum) going rate. As i understand it, the denarius communis was also a pure accounting currency, not an actual coin, and thus had no silver (or anything really) content to calculate with.

If you wanted a copper-silver-gold system where gold is more than 100 times as valuable as silver, you'd need a world where gold was much, much rarer than silver and copper: in Earth's crust, silver is 20 times as common as gold and copper ~700 times more common than silver (14000 times as common as gold), and indeed, the yearly extraction today amounts to 2500-3000t for gold, ~25000t for silver and ~ 15 million tons for copper. If in DF you could expect one bar of gold per ~1000 bars of copper (in the world, not necessarily in your fort), it would make sense, but with gold at roughly 1/5 the frequency of copper, the 15:1 value ratio looks fine to me.

There are two factual errors in your writeup: the measurement of grain in the edict was not the "italian" modius of ~8,8 litres, but the "kastrensic" modius, which is double the volume.
And standard muscle meat density is 1060 grams per litre. For some reason, you took the tenfold value (which would make meat about as dense as silver). 200 ml of meat weigh 212 grams, a bit over half a pound. A modius (k.) of grain would presumably weigh around 10-12 kg, i.e. around 30 roman pounds. That much pork would cost you 12*30 = 360 denarii, vs. 100 denarii for wheat, a mere 3,6:1 price difference (which to me points at the "maximum price" issue - as a grain, wheat would suffer from much sharper price swings, so the legal max was set higher relative to the normal price, while for meat it wasn't quite so far out).

[Random_Dragon]

Ah, thank you then. Since this was more an incidence of mental gymnastics, I'll admit that the edict's only useful if the math implies an item is disproportionately above the price ceiling. Using d.c. also makes it hard to translate into usable DF prices, I'd need to double-check what the coins of the day were tarriffed at in d.c., but even if it gives a rough idea of how much physical silver was equated to how many d.c., it'd be rather iffy math.

What is peculiar is another source, which I used at first before finding the link in the first post, seemed to state that pricing was based on the Italian modius. Did make some modifications once I found that second source as it seemed more complete, but that was one difference I failed to notice. With meat though? I was surprised by that too, but...



Which would explain why my smaller adventurers seem to be unable to ever carry enough supplies for long journeys.

Regarding the rarity of gold, I did expect that gold is more common than in reality, but exactly how common it is was something I couldn't get easy math on. If it's about 15:1 then that does make it pointless to make gold coins worth more, yes. Which admittedly was the line of thinking that started me into looking this stuff up, so there goes the main thing I thought would be interesting.

[Fleeting Frames]

Well, when it comes to metal available in histfigs' hands, not even divine metal has 1:14000 ratio with all other metals combined; though it'd be a start. Of course, that's not so bad issue when you can water it down with other metals.

[Random_Dragon]

Well, when it comes to metal available in histfigs' hands, not even divine metal has 1:14000 ratio with all other metals combined; though it'd be a start. Of course, that's not so bad issue when you can water it down with other metals.

True. Idea here is that gold at least seems to be rarer than silver to some extent, and much rarer than copper, but I can't figure out how much it is in game math.

[Random_Dragon]

So, I decided I'd run some testing on he actual in-game effects of making coins more valuable in DF. First I gathered some vanilla fellows and examined their pouchs. Order is gold coins (15 value), silver coins (5 value), and copper (1 value). Then of course the total.

Spoiler (click to show/hide)

45, 5, 17 (total 67)
15, 40, 21 (total 76)
45, 5, 58 (total 108)
75, 15, 18 (total 108)
30, 15, 70 (total 115)
30, 5, 91 (total 126)
60, 55, 26 (total 141)
15, 90, 42 (total 147)
15, 60, 72 (total 147)
60, 70, 67 (total 197)
30, 95, 99 (total 228)
60, 95, 96 (total 251)

12 tested
Lowest total: 67
Highest total: 251
Mean: 133.583
Median: between 126 and 141
Mode: Tie, 108 and 147 (2 instances)

After this I decided I'd test a simple, D&D-style 1:10:100 ratio. Some reasons for that, namely to mean I don't have to fudge material values TOO much relative to what their coin valies are, compared to 1:10:250 or even 200.

Order is gold coins (100 value), silver coins (10 value), copper (1 value), then total.

Spoiler (click to show/hide)

100, 20, 15 (total 135)
100, 10, 31 (total 141)
100, 10, 59 (total 169)
100, 10, 86 (total 196)
100, 70, 32 (total 202)
100, 30, 78 (total 208)
100, 20, 97 (total 217)
100, 50, 67 (total 217)
100, 50, 76 (total 226)
100, 80, 91 (total 271)

10 tested
Lowest total: 135
Highest total: 271
Mean: 198.2
Mediamn: between 202 and 208
Mode: 217

Could only get ahold of 10 people this time. One thing that became apparent is that the game REALLY wants to give NPCs at least 1 of each type of coin, enough to push the amount of pocket money the average NPC has to a higher amount than before.

So then I figured maybe a less extreme example, 1:10:50. Still a massive step up in terms of how much a gold coin is worth compared to value, but enough that my earlier concerns about giving gold a sane material value requiring number fudge isn't as big a concern.

Order is gold coins (50 value), silver coins (10 value), copper (1 value), then total.

Spoiler (click to show/hide)

Next test:
50, 10, 5 (total 65)
50, 10, 6 (total 66)
50, 20, 23 (total 93)
50, 10, 29 (total 99)
50, 50, 27 (total 127)
50, 80, 5 (total 135)
50, 50, 37 (total 137)
50, 60, 27 (total 137)
50, 60, 94 (total 204)
50, 70, 98 (total 218)

10 tested
Lowest: 65
Highest: 218
Mean: 128.1
Median: between 127 and 135
Mode: 137

Again didn't find 12 NPCs right off the bat. Average is a lot closer to vanilla, lower in fact. Close enough to suspect it's just RNG plus low sample size, whereas the prior test was enough of a step up to tell something was definitely wrong.

Despite the lower value, the game still is reluctant to give more than 1 gold coin to NPCs, unlike in vanilla where gold coins are more common. Even when there's enough silver to convert into a second gold coin, and even when doing so still leaves more than 1 silver left over.