The dividing line between Trap, No treasure and Special, Treasure is at 50%, and I tend to assume the probability on a d6 in percentage terms goes:
1 - 17% (17%)
2 - 16% (33%)
3 - 17% (50%)
4 - 17% (67%)
5 - 16% (83%)
6 - 17% (100%)
My reason is, if you're rounding up or down, 16.6 recurring is closer to 17 than 16, but 33.3 recurring is closer to 33 than 34. So you get tiny adjustments as you go - 1 adds a tiny bit, 2 loses a tiny bit, 3 and 4 add tiny bits again, 5 loses and 6 has what's left.
It gets really gritty when you start taking sixths of sixths and calculating percentages. But guestimating can produce something not too clunky that's still workable (basically, my smallest unit is 6%, approximating a 2/36 chance such as Empty, Treasure)
So the table could look more like this:
Chart 3b
01-17 Monster; Treasure
18-33 Monster; No treasure
34-39 Trap; Treasure
40-50 Trap; No treasure
51-56 Special; Treasure
57-67 Special; No treasure
68-73 Empty; Treasure
74-100 Empty; No treasure
Compared to the other table it gives a slightly larger chance of getting an 'Empty, No treasure' result but a slightly smaller chance of getting a 'Monster, No treasure'. That was one of the things that I noted in the original table of course - I deliberately shaved the 'Empty, No treasure' chance as it was a) the largest single group and b) inherently the least interesting.
The other way I thought of to do it was to say that each of the original 36 entries had a base 3% chance of happening. This of course produces a table with 108% on it, so like the occasional addition of 16 instead of 17, 8 of the 3%s need to to be 2%s instead. This will give us 100% again. Like trying to map 16s and 17s onto 16.6 recurring, I'm trying to map 3s and some 2s onto 2.7 recurring. I'm still keeping the same 17/16/17/17/16/17 format, so essentially I need to have five 3s and one 2 to get to 17, four 3s and two 2s to get to 16.
I'm happy that my jumps are 3, 6, 8, 11, 14, 17 (this is a whole d6); 19, 22, 25, 28, 31, 33 (another die), 36, 39, 42, 44, 47, 50 (third die, we're halfway there!) and I suspect I can just replicate that line with 50 added to it - 53, 56, 58, 61, 64, 67 (die 4); 69, 72, 75, 78, 81, 83 (die 5), 86, 89, 92, 94, 97, 100 (die 6). I did check this afterwards and it's right (I prepared an excel spreadsheet to multiply 2.7777777 by 1-36, then list the result as whole numbers, then calculated the difference between the previous number and the next step, so it now looks like this: yellow marks a new die, magenta is a 2, but of course sometimes a 2 is a new die; they're just magenta).
3 3
6 3
8 2
11 3
14 3
17 3
19 2
22 3
25 3
28 3
31 3
33 2
36 3
39 3
42 3
44 2
47 3
50 3
53 3
56 3
58 2
61 3
64 3
67 3
69 2
72 3
75 3
78 3
81 3
83 2
86 3
89 3
92 3
94 2
97 3
100 3
We know where the steps are here (though I'm collapsing the choices of '1 - Monster, 1-3 - Treasure', '1 - Monster, 4-6 - No treasure', '2 - Monster, 1-3 - Treasure', '2 - Monster, 4-6 - No treasure', down to what is in effect (because actually, Moldvay does it) '1 - Monster, Treasure', '2 - Monster, No treasure'.
The first parts (up to 33% probability) look the same - 17% probability of 'Monster, Treasure', 16% probability of 'Monster, No treasure'.
The next part, up to 50% dealing with Traps, have 2/6 determinations with treasure and 4/6 without. They also look the same.
Specials (with or without treasure) take up to 67% - again you could bundle them together if you don't accept the logic of the Trap/Special analogue. Otherwise, they go the same as I have them above (and below).
The last 5/6 is the 'Empty' set from 68% - in this case, 'Empty, Treasure' changes from 73 to 72 and 'Empty, No treasure' increases from 74-100 to 73-100. That's the only change between the two tables.
Mathematically, this is the closest to the original.
Chart 3c
01-17 Monster; Treasure
18-33 Monster; No treasure
34-39 Trap; Treasure
40-50 Trap; No treasure
51-56 Special; Treasure
57-67 Special; No treasure
68-72 Empty; Treasure
73-100 Empty; No treasure
Is it easier? I'm not sure. It's shorter, because it's a single table and has fewer entries. It seems like it's simpler, but it still involves rolling 2 dice (because with dice, two d10s is still the easiest way to get a percentage I think). If you're using an electronic generator then it's much easier to get a single d100 result and consult one table than a d6 result and consult a table then another d6 result and consult another table. I tend to work that way - dice at the table for sure, but when I'm sitting around fiddling with world-building I'm happy with numbers spewing out of my screen.
Anyway, for the benefit of whoever might want to use it, please accept any of my 3 percentage-based variants on Moldvay's room-stocking table.
And I think I'm done with it for a little while.
Except for using it of course... I'll get back to that soon.
No comments:
Post a Comment