DMG tables

[Seananigans]

I'm just having a chance to mess around with rolling hoards from the DMG module and such. I'm noticing that where the FG module added in extra tables for stuff like rolling a random scroll of 2nd level, or a random Vicious weapon, etc., that some unorthodox RNG dice methods were used. For instance, the Vicious weapon table rolls 2d20 because there are roughly 37-38 possibilities, and the 2nd level spell scroll table rolls 3d20, similarly because there are roughly 59 possible spells of that level (I don't think these include Elemental Evil spells?).

While it's super awesome that these tables were painstakingly created, I notice that they tend to be alphabetically laid out (scrolls at least, the vicious table doesn't seem to have an organizational pattern I can recognize), rather than perhaps by rarity of usage or some such. Unfortunately when rolling 2d20 or 3d20 etc, your results are going to much more commonly be somewhere in the middle (similar to 2d6, the dice used for craps, ending up as a 7 most often). So rather than the middle items in the list being the more common things (scroll of magic missile, fireball, cure wounds, etc), instead they'll just be whatever happens to be in the middle of the list alphabetically.


If it's even desired by the FG team to shore this up, I see two solutions, a simple/easy short-term fix, and one that takes more effort but would result in something very akin to how the DMG already takes rarity into account (and is also future-proof).

1) The simpler short-term fix (I think?)
Rather than using 3d20 for a table of ~59 scrolls, nest the rolls similar to how the other nesting is done, to create a table with an equal statistical weight across the board. For example, split the 59 scrolls into 3 sub-tables of 20. Roll 1d3 to determine sub-table, then roll 1d20 to find the scroll. This will create an equal statistical chance of each scroll result, ignoring the very small difference created by the sub-table with 19 instead of 20. This sort of thing works best if the numbers aren't prime, as for instance if you had 57 scrolls, each sub-table would have 19, with a 20 result being a simple re-roll. The 59 issue is small though, and this method is still much more equal statistically than the 3d20 that is set up currently.


2) The effort-full version that's somewhat future-proof
Have all of these tables (vicious weapons, scrolls, gems, whatever you guys have added on your own) set up on a d100 basis. Use similar weighting methods that the DMG uses (rare items have only 1 or 2 in 100 result, more common things have upwards of 5 to 10 in 100 result). This will obviously require effort in designing weights, and some arbitrary decision-making on your own parts as to what would be more common or more rare. It does, though, come with the extra benefit of being future-proof for any future items/whatever that get added and that might be desired in these tables. For instance, a source of spells like the EE, which could then be squeezed into the existing d100 scroll tables, with a bit of weight-shifting. Personally, I'd love if the EE spells were included in these tables, as the tables themselves aren't even DMG content (in fact, spells aren't in there at all), and just FG guys being awesome and adding useful things to the module.


To conclude, the DMG module is well worth the money and a fantastic addition to FG, and I thank you guys for your hard work! If a slight re-work of these tables is desired, I hope my suggestions are helpful.


Small side-note bug report, re: the Vicious Weapon table, it currently rolls 2d20 but has a result of 1 (vicious club). This will obviously result in a DM never seeing a vicious club if using this table, unless a modifier is applied to the roll. I haven't been able to look at every table so far, so there could be more of this sort of thing (There are tons of tables, omg!). Regardless of whether you guys have a desire to re-work the tables as I've suggested, the bug could use a fix.


Thanks again for the amazing, hard, work!

[Zacchaeus]

Taking your last point first, that problem has already been noted and will be in the next patch.

To address your main point there is actually a third option which is to remove the dice altogether. In the table you mention with 59 scrolls if we remove the dice then there will be an automatic roll of d69 which would come out as random as it can be without looking at option 2 which as you say would take rather a lot of effort.

The only difficulty that I forsee is that I don't think that the parser will handle the parsing if we leave out the dice altogether and/or might give spectacularly bad results. I'll have a word with Zeus and see what comes out of that discussion.

[Seananigans]

Taking your last point first, that problem has already been noted and will be in the next patch.

To address your main point there is actually a third option which is to remove the dice altogether. In the table you mention with 59 scrolls if we remove the dice then there will be an automatic roll of d69 which would come out as random as it can be without looking at option 2 which as you say would take rather a lot of effort.

The only difficulty that I forsee is that I don't think that the parser will handle the parsing if we leave out the dice altogether and/or might give spectacularly bad results. I'll have a word with Zeus and see what comes out of that discussion.

Ah ok cool, I've never made a table so I wasn't aware simply taking the dice out is an option, and that it'll just run a simple 1-59 RNG number. You mention difficulty with the parser, but these tables I'm speaking of are all made from scratch by the FG team right? So I'm not sure what you mean by that. There's no table of vicious weapons or 2nd level scrolls to parse out of a PDF or whatever, to my knowledge?

[Zacchaeus]

Ah ok cool, I've never made a table so I wasn't aware simply taking the dice out is an option, and that it'll just run a simple 1-59 RNG number. You mention difficulty with the parser, but these tables I'm speaking of are all made from scratch by the FG team right? So I'm not sure what you mean by that. There's no table of vicious weapons or 2nd level scrolls to parse out of a PDF or whatever, to my knowledge?

The parser is the tool which is used to create the Dungeon Masters guide and other 5e things like the PHB and the adventure modules. Everything in the DMG is created within that tool, including all of the tables etc. Nothing is created by hand. So there are certain limitations on what can and can't be done depending on whether the tool can cope with it.

On thinking about this further I'm not sure your first option is technically feasible. The mechanism doesn't exist within the tables to select from a menu of tables. So for example say on a roll of 20 you got a level one scroll; the call from the originating table would be to roll on the level one scroll table. You can't say roll a d3 and select from three possible other tables. You would have to nest them so that the call would be roll on the first table and if you get x result roll on another table and so on. In order to have an even chance of any one of the tables getting a hit you'd need to weigh up the odds and that could be even more complicated than your second suggestion. If you see what I'm getting at :)

[Seananigans]

Well, I'm certainly no expert on how FG modules are made, and have only heard of this parser secondhand. I just know statistics! So my main point here was just that as things stand, certain items/whatevers are weighted more than they should be, and it seems to be accidental, rather than intentional like my #2 suggests.

To further my own knowledge though, as I'm confused why you think my first suggestion isn't feasible, I'll try to explain what I meant in more concise terms. Currently, the DMG's treasure hoard creation is set up with, in some cases, nested tables. You roll a Hoard, get a d100 result, and then depending on that result, it might roll on several other tables. This functionality seems like it could work just fine to facilitate equal-weight determinations for these tables I've mentioned.

So, as an exact example:
Hoard roll d100, result is some coin, and 1d4 magic items
-Triggered nested rolls for coins, and triggered nested roll 1d4 for how many magic items (let's say it rolls a 1)
--Triggered nested d100 roll on magic item table (let's say it comes up a 2nd level scroll)
---Triggered nested roll of 1d3 (or 1d6 using sets of 2, whatever) to find a result of 1-3 for the level 2 scroll sub-table (let's say it rolls a 1)
----Triggered d20 roll on L2 scroll sub-table 1, which contains L2 spell scrolls Aid to whatever the 20th alphabetical one is

This method seems well within what is possible, since in creation of the FG DMG module, roll-nesting is possible, and there was creation of many tables from scratch that don't exist anywhere in the DMG (such as the level 2 spell scroll table, as mentioned). This would simply shoe-horn in another nest level, and split the current spell scroll tables into smaller parts, facilitating equal statistical weights. Is there something I'm missing in understanding how this all works?

I'll disclaimer that I could be way off on which one of my suggestions is easier for the team to do, but based on what's involved with option 2 (determinations by hand on how to weight certain items to facilitate a good weighted d100 table), I just assumed #1 was the easier.

[Moon Wizard]

We'll take a look at this internally. Most likely, it was an unexpected situation, and we just need to remove the defined dice for the table in the DMG files. Then, the ruleset will simply generate a random number in the correct range.

Cheers,
JPG

[Zacchaeus]

You are correct Seananigans; I wasn't thinking straight at all and was coming up with a much more complicated scenario. Yes, you could, instead of rolling directly on scroll table x interpose another small table with three entries with each entry going to a different table each with 20 scrolls.

However, as MW says removing the dice altogether simplifies the procedure and would create a lot less work and obviate the possibility of introducing a plethora of errors.