Tweaking the random encounter tables

I have been reading the Alexandrian and thinking about dice probabilities (both highly worthwhile pursuits). His series on hexcrawls is fantastic and has given me a great format to follow for my random encounter tables, where before I was really haphazard (every region had something different). I can trust that guy to present information that is useful, battle-tested, well thought out and incorporates knowledge of the underlying math.

The Numbers

As for dice probabilities, I have always flip-flopped on how to spread out the chances of running into each monster on a random encounter table. Rolling 2d6 gives you a 'pyramid' distribution (it's unlikely you'll meet those monsters on 2 or 12), which I always liked. You can hide some really evil, nasty beasts under those numbers. But just what are the chances, really? Well I looked it up. Bookmark that one, you'll thank me later.

This pyramid is actually steeper than I thought at first glance. Rolling a 7 is six times more likely than rolling a 2. I think these numbers (6/36, etc) are the limit of the kinds of probabilities that are easy to imagine and manipulate in your head, without writing them down. In fact, I think it's too granular, in a way. Let's compare two possible random encounter tables and see how much I care about the details...

This is the table for the Rainy Jungle as I used it last session. You can see there are a few oddities that could be fixed:

2d6

2-3  -  giant bee     (3/36, 8.3%)
4  -  1-2 giant geckos   (3/36, 8.3%)
5  -  1-3 neanderthals   (4/36, 11.1%)
6  -  1-3 giant botflies   (5/36, 13.8%)
7  -  1-3 goblin hunters  (6/36, 16.6%)
8  -  1-6 capybaras   (5/36, 13.8%)
9-10  -  1-3 steam beetles  (7/36, 19.4%)
11  -  1-2 orchidmen   (2/36, 5.5%)
12  - roll 2x and combine   (1/36, 2.7%)

Why are steam beetles so popular/? Because the probabilities are in 'chunks'. I can't freely add or subtract 1/36th to any result, except by adding 2 + 3 or 12 + 11. If I have less than 11 results, some have to group up and this creates odd 'lumps' in the probability pyramid. The 2d6 table works well when you have exactly 11 results. Any less and it becomes really awkward to use. 

I *could* move goblin hunters to the 9-10 spot and steam beetles to 7, but this seems needlessly fiddly and annoying to format. Is there a better way?

Let's change it to a straight d20 roll and see what happens:

1d20

1  -  1 giant bee    (1/20, 5%)
2-3  -  1-2 giant geckos    (2/20, 10%)
4-5  -  1-3 neanderthals    (2/20, 10%)
6-8  -  1-3 giant botflies  (3/20, 15%)
9-11  -  1-3 goblin hunters   (3/20, 15%)
12-14  -  1-6 capybaras   (3/20, 15%)
15-18  -  1-3 steam beetles   (4/20, 20%)
19  -  1-2 orchidmen   (1/20, 5%)
20  -  roll 2x   (1/20, 5%)

The probabilities are almost the same (within 3.3%), and I can fit the same number of monsters. But all of a sudden the table is much more adaptable. I can modify a result in the middle of the table (like making the goblins 5% more or less common) without having to rework everything!

The small downside is that I can't have a 2.7% result for the REALLY rare monsters. The smallest unit is 5% but that's a fair trade for versatility, ease of use and less mental math. 5% is the chance of a critical miss, so there is a nice symmetry in it also being the chance of rolling the blood-mad Wights of the Murderous Moors, or whatever harshness is on your table.


The Monsters

Can you feel... the danger?

Random encounters haven't happened often in the last few sessions (my dice seem to love going easy on the players). Given their rarity, I have decided to move away from the more 'naturalistic' monsters. Having the only random encounter in a three- or four-day expedition through the steaming jungle wilderness be with mundane capybaras (pictured) is a bit weak. 

Even if they should be there for 'realism', the encounter can at least be with giant capybaras the size of VW buses, or vampire bats instead of a swarm of normal bats, or sneaky camouflaged giant chameleons with 25' long tongue attacks (that's a cool one actually...) instead of plain giant lizards. Capybaras can be the kind of thing you hunt for dinner.

So here's my rectified random encounter table with better monsters and everything streamlined a little bit. Thanks again to the Alexandrian for this.

1d20

1-3  -  1-3 goblin hunters
4-5  -  1-3 neanderthals
6-8  -  1-3 giant botflies
9-11  -  1-3 steam beetles
12-14  -  1 man-mantis   [2-HD version of giant mantis]
15-16  -  1-2 giant chameleons
17  -  1 jungle bear (hairless)  [too good not to use!]
18  -  1 giant bee
19  -  1-2 orchidmen
20  -  roll 2x and combine

Now the monsters are a bit more exciting. Nothing you could actually encounter on Earth (a few variations are okay). I'm going to do all my random encounter tables in this format. It's customizable and has just enough granularity for me to work with, not so much that it takes any mental strain to manipulate.


*****

I'm on fire this week. Let's see what's on the playlist next?