15 August, 2020

A Treasure Type Comparison

Someone asked how the various treasure classes compared across editions (and with Labyrinth Lord), and since I had already done the work analyzing this for Simulacrum, I thought I'd post the results (along with what I wound up going with, as usual).

I'm only going to cover "lair" treasures here (the large ones).  I'm also leaving out magic items: I'm only interested in cash value for the moment.  Click on any of the images to enlarge, but I have a spreadsheet at the bottom that you can download to view all that info together.

All average values listed are in gold pieces.


You can find the B/X treasure classes on page X43.  Here's what they look like:

B/X lair treasure types and their averages

The order of worth is C, B, E, D, F, I, A, G, M, H.  There are two unusual items here.  First, there's Type M, which neither Advanced edition felt the need to bother making into another lair level.  Second, B/X hands out platinum on the table at a much higher volume, equal to the other coins (×1000); 1st and 2nd ed use only ×100.

Types B and E overlap pretty strongly (though E has a much better chance for magic items, a factor not being considered here).  The race to the upper range is rather poor as well, going 7.6K, 11K, 17.7K, 23K, 50K, 60K.  Unlike 1st or 2nd ed, the B/X designers listed the averages of these types in the text (p. B45).  Unfortunately, they were calculated incorrectly: the designers thought the progression was 5K, 8K, 15K, 17K, 25K, 50K.  Below are the major discrepancies:

Type F: Stated average: 5,000 gp.  Actual average: ~7,600 gp (~50% higher than stated)
Type H: Stated average: 50,000 gp.  Actual average: ~60,000 gp (~20% higher than stated)
Type I: Stated average: 8,000 gp.  Actual average: ~11,000 gp (~40% higher than stated)
Type M: Stated average: 15,000 gp.  Actual average: ~50,000 gp (~230% higher than stated)

1st Edition

You can find the 1st AD&D treasure classes on page 105 of the Monster Manual.  Here's what they look like:

1st edition lair treasure types and their averages

The order of worth here is C, B, E, D, F, I, G, A, H.  The oddity is transposing the escalating worth of A and G, compared to the other two editions.

1st ed races up the charts very fast, going 5500, 6700, 21K, 34K, 77K.  Its Type H treasure is also the clear winner amongst all editions, truly a treasure of legend.

The real money in 1st edition is in jewelry: any Treasure Type that feeds you this is a license to print money.  A piece of jewelry is worth, on average, 3,336 GP (allowing for the opportunity to get that sweet double-value result by rolling 1 on a D8 and then once more by rolling 1 on a D6; this can continue indefinitely, but gets vanishingly unlikely quite quickly so I didn't try to model for it past two instances).  This is the reason Type A and especially Type H generates such high values: while your dragon hoard is going to have tons of coins, as you'd expect, about half of its value is actually going to come from the on-average 25 pieces of jewelry you're going to find in it.

2nd Edition

You can find the 2nd ed AD&D treasure classes on page 133 of the DM Guide.  Here's what they look like:

2nd edition lair treasure types and their averages

The order of worth here is C, B, E, D, F, I, A, G, H.  This follows the general pattern we've seen, matching B/X specifically (minus the presence of Type M).

What's remarkable is how much poorer they are than 1st ed classes.  The letter codes are a very good match in terms of order of ascension.  However, the clustering here is all down low--excessively so--with nothing really up high to compensate unless you want to start stacking instances of H.  700, 1250, 1450, 2250, 2650, 4000, 5750, 7500, 15K.

If you're using the gold for XP rules, made optional in 2nd edition, you're going to make a lot less of your XP from cash in this edition.  As monsters are worth far more XP in 2nd, and there's an arbitrary story award called for that can compensate to some degree for a lack of coin if the DM feels like it, this isn't as big a deal as it appears.  But trying to run a game primarily with gold for XP and relying on these treasure tables would go poorly past the lowest levels.

The reasons for this are 1) the removal of the princely jewelry category, and 2) the removal of the fixed platinum category.  In place of the latter, you instead get a result that can be either electrum or platinum, this being determined by what the DM feels like handing out.  This gives a DM a bit of flexibility in moderating the treasure amount rolled, even if it does somewhat defeat the purpose of allowing random chance to guide things, which is primarily the point of these tables in the first place.  The introduction of the art objects category (which is supposed to include jewelry) and the increase in value of copper and silver does not even remotely compensate for the above.

Adaptation / Conversion Notes

First, here's the spreadsheet with the above analysis on it (note that there's two sheets in the single file, and that I've also included Labyrinth Lord conversion notes):

Treasure Type Comparisons

So what does this mean at the table?  We have some wild variation here, and so no nice and clean way of converting between treasure types from different editions: you'll have to convert one at a time.  In the linked spreadsheet I've given some suggested conversions, though they often involve using two or even three of another table's types to get an equivalent.

The value of gems is pretty consistent.  A B/X gem is worth about 70% (on average) of a 1st or 2nd ed gem, but since the average values are generally low (194.5 vs 275) the variance isn't too big a deal and you can generally be comfortable converting a gem hoard from a module in one edition to another straight across (or just going with a clean 3 B/X gems for every 2 Advanced gems).

If you try that with jewelry, on the other hand, you may wind up turbo-levelling a party if you're not careful.  In B/X, jewelry is worth about 1,050 gp on average.  2nd edition does away with jewelry altogether, but replaces it with art objects that are worth almost exactly the same as B/X jewelry (average value: 1,100 gp, albeit with a much higher possible maximum).  In 1st edition, as we've seen, jewelry averages to triple those values.  ×3 / ÷3 are pretty easy values to work with, however, if you're dealing with converting material from one edition to another.

Playing Around with Treasure Types

My main issue with the canon tables is how chaotic they are.  They don't rise alphabetically from low to high, and they crowd too much down low, while having a poor spread elsewhere.  The first issue is largely just an aesthetic concern, but then again there's no real reason to preserve a less intuitive format for the sake of tradition.  It's not conducive to memorization.  If we're going to bother with treasure tables (I like them, but not all old-schoolers agree there), then I think they need to be recoded to sort low to high, but more importantly to also offer a better spread of values, so that it's easy to have a treasure type on hand for all expected levels of the game.

I could easily see someone dropping the number of types: there's no compelling reason to have this many, though the extra granularity is nice.  However, remember that treasure types also include magic items: you want to be able to allow for a nice, steadily progressing distribution of both money and random magic items.  Still, I'm sure you could toss one or two without any real loss.

The only bit from the 2nd edition tables worth using, I think, is the art objects idea, which can be a lot of fun, allowing as it does for very unusual items to be added to a hoard in a systemic fashion.  I also prefer the lower jewelry values of B/X and 2nd ed, especially as it then allows me to cleanly transpose jewelry and art objects as desired (since statistically they're about the same value).

For my own game, I've gone with the following:

Simulacrum Treasure Types

I didn't bother with B/X's Type M because I felt I had enough granularity at this point, and because topping out at H for Hoard (as is traditional) pleases me; sometimes it's the little factors that matter.  The overall values assume a silver standard (and then also 1 sp = 10 XP, in some ways making it a copper standard).  I like this because it makes small amounts of coin truly valuable: my players actually care about finding a small purse of silver, which I really enjoy.  It also eliminates the "wagonloads of gold" issue that plagues so many D&D campaigns and which has generated so much writing on how to fix or compensate for it.  If you're using the gold standard instead, you'll need to add a zero to every value.

The major change here is replacing the multiple rolls per type (to determine variance in terms of hoard size and precise composition) with a single D20 roll (or 2D20 if you want more variance; see below). Basically you're guaranteed to get a set amount of treasure, and the roll determines its makeup: whether a mountain of copper, a handful of platinum, or a pile of brass busts or whatever.  I just really got tired of making all those separate rolls when generating a hoard and settled on this as a way of combining speed with variance.  There's a separate D20 table if you want some swing to the overall hoard value, but that's still a lot quicker overall.  In terms of personal treasure, I've gotten rid of the "monster has 4 copper" categories as not worth bothering about.

What To Do With Coin Sizes

Some might argue that making coins more valuable (and thus handing out fewer of them) undercuts one of the major resource management elements of old-school D&D: encumbrance and coin weights.  However, I feel that while encumbrance and resource management are vital considerations, the principle does break down after a certain point.  Wealth is the major arbiter of advancement in old-school games.  As such, no matter how onerous you make it, players are going to simply just going to grit their teeth and deal with it: they have no choice if they want to progress.  Additionally, there's only so many all-copper hoards and giant teak cabinets and roc attacks on the PC wagon train and the like to screw over treasure hauling before it all appears adversarial, and it's difficult to make an encumbrance system even remotely realistic that doesn't just let you haul off mountains of coins anyways, leading to a lot of bad encumbrance systems (something I'll look at more closely when I finally finish that encumbrance post I've been working on).

I've come to the conclusion that there's already enough other, better ways of dealing with resource management (e.g. light, time, food, ammo, spells, fatigue, HP, every other aspect of encumbrance) than constantly worrying about this one element, especially when you consider all the distortions created by insisting on masses of heavy coins.  A lot of systemic ugliness--coin weights, encumbrance rules, knock-on economy effects--can be solved by simply going with fewer, more valuable coins, and not thinking that this is letting players "get away with something".  If the players truly earn their treasure (something it's almost entirely in your hands to ensure), just let them have it.

Some other material on the topic:

OD&D treasure types analysis:
An analysis of OD&D treasure types (registration required):

An analysis of 1st edition treasure types:
Another analysis of treasure types:
OD&D: https://batintheattic.blogspot.com/2012/08/delving-into-ad-types-part-i.html
AD&D: https://batintheattic.blogspot.com/2012/08/delving-into-ad-treasure-types-part-ii.html

Moldvay Basic treasure types by monster:
B/X treasure type densities:

You may also want to check Dragons #137 ("Treasures of the Wild") and #207 ("Trifling Treasures"), as well as Kellri's Old School Encounters Reference #4 and Treasure, by Courtney Campbell.  Lastly, the 2nd edition sourcebook Forgotten Realms Adventures has a fun Treasure chapter (pp. 130-146) that really expands on art objects and gemstones, for those who want to get extremely specific in those regards.