Thoughts on Knowledge Skills

The use of knowledge skills in a game can sometimes be a tricky proposition. A lot of ‘old’ systems which lacked a skill system did away with them entirely, and it was either down to player knowledge or maybe a straight Intelligence check to see if a character knew something that could have bearing on an adventure.

As a fan of skill based systems, and a split between player knowledge and character knowledge, I’m in favour of characters making skill checks to see whether they know something or not, so YAGS makes use of them quite heavily.

The trouble can come when a knowledge check is crucial to the adventure. It’s not just knowledge checks, but any investigatory check where success will provide information that is needed to proceed. This can be different from a combat where there are often multiple rolls involved, and a single bad roll generally doesn’t result in failure.

In a pure sandbox game, where neither the players nor GM care whether they ‘proceed’ or not, it probably doesn’t matter. If the PCs have to give up on something because everyone failed to spot the footprints, or didn’t recognise the cult symbol that was scrawled on a wall because of a single bad roll, then in this case it may not be a problem. But even so, it can be annoying if if keeps happening.

In games like Gumshoe, characters automatically succeed at gaining information. It fits that style of game, because finding out information is what drives the plot of these games. You also tend to either have the skill or not, there isn’t a range of skill values you can have.

YAGS is more progression based, so a character can have a skill from 1 up to 10 or more, so there can be a wide variation in what individual characters know. For the Fantasy version of the rules, there are a number of standard Knowledge skills, such as Arcana, Common, Geography, Warfare etc. Knowledge Common is what pretty much everyone has some of. The others start at zero and can’t be used until points are put into them.

So I’d like a way that rewards characters who have invested points into skills, without completely preventing a group from finding out things in a way that halts progress and makes things boring or frustrating. Non-martial characters don’t get to automatically win a fight just because it’s important to the plot, so non-knowledge based characters shouldn’t be able to automatically know everything for the same reason.

So the way I want to do things, is to have multiple clues and multiple ways of finding them. Some clues will give hints which can be followed up on, others will provide the full set of information immediately.

For example, the party find a closed stone door deep in a dungeon, with a strange symbol carved upon it. When they try to recognise it, there could be a number of different skills they could try to use, which provide different amounts of knowledge depending on what Target Numbers (TN) a character achieves.

Knowledge Common
TN 10: It’s a cult symbol, but you can’t remember any details.
TN 20: It’s the symbol of a cult of demon worshippers, generally considered a bad omen by the locals.

Knowledge History
TN 20: It’s a symbol used by the cult of Asash, who used to operate in this area about 50 years ago.
TN 30: They were driven out by the Paladins of Sielzen who are based in the nearby cathedral.

Knowledge Arcana
TN 20: It may be a ward of some kind, but divine rather than arcane.

Knowledge Religion
TN 10: It’s a symbol used by certain demon worshipping cults to mark the location of their shrines.
TN 15: The cult was the cult of Asash, which practised human sacrifice in exchange for eternal life.
TN 20: Their shrines were sealed with demonic traps and wards, and there were words that could be used to get around them.
TN 30: You know the types of wards, and what to look out for around warded places, so you might be able to spot and avoid them.
TN 50: You know the words needed to disarm the wards.

So the characters have multiple chances to know something, even if they don’t have exactly the right skills. Since everyone has something in Knowledge Common, they may get enough clues to know who to go to ask if they want more details (go find the local church, and ask there, rather than taking it to a local wizard who will probably know nothing).

A character with knowledge of History may know how it relates to local events in the place, and get some good ideas of who will know some real details on the cult. Finally, a student of Religion will know some real detail, since this is the skill that is really needed. They are unlikely to know the really useful information – how to bypass the traps – but a highly skilled character will get rewarded if they roll well.

This sort of approach avoids the type of binary pass/fail results that can leave PCs with no knowledge about things, without just giving them everything automatically. If they roll badly, they have an idea of who they could talk to if they want more information, and if they roll well, they find out useful details they wouldn’t have otherwise got, making it worthwhile investing points into such skills.

Sometimes low checks could give common knowledge which isn’t entirely accurate. I think I’d prefer such checks to be missing information rather than providing the wrong information. This could be just as deadly sometimes, but they do say that a little knowledge can be a dangerous thing.

It’s more effort doing things this way, and it’s definitely not a new idea. Some of the later D&D 3rd Edition monster books gave graduated lore results for knowledge checks about monsters, something I think is very much worth doing, and some adventures in Pathfinder take the time to provide graduated results about topics PCs may come across.

I plan on taking this idea and adding knowledge difficulties to all monsters in my game. This would make it easier to take into account the differences based on the rarity of creatures. For example, information for goblins might be something like this:

Knowledge Checks
Common (0): Goblins are ravenous, small, humanoid vermin that attack and kill humans and livestock. They are spontaneously created from rotting and diseased vegetation in forests.
Common (10): Goblins grow from a fungus that spreads through spores, so can emerge almost randomly from dark moist forests. They are barely intelligent, highly aggressive and always hungry.
Common (15): Goblin nests normally consist of a dozen or so goblins, but larger ones are possible. The youngest and weakest are green goblins, and as they age and grow in size their skin reddens and hardens. They have a rudimentary language of their own.
Common (20): The largest and oldest goblins can be almost as tall as an adult human, with dark reddish bark-like skin which acts like tough armour. They will sometimes use weapons they find, though rarely make their own. A goblin’s wounds can become infected and spread the disease that causes them.
Nature (10): Goblins are a type of fungal creature that sprouts from fungal patches often found in forests. They are spread through spores, which can settle and mature into the first goblins within a month.
Nature (20): Goblin spores are normally spread by diseased animals, though can be spread via the wind. The latter is rarer though, so culling animals in the vicinity of a nest is a good way to contain the infection.
Any wounds caused by a goblin can also spread the infection.
Nature (30): Most infected humans are just carriers. A few die. An even rarer few can undergo a transformation into a goblin hybrid. An infected pregnant woman can also give birth to a hybrid, but the normal outcome is a still birth.

An alternative way of handling things might be to have multiple rolls, and building up points towards some total. This makes it more like a combat where everything hinges on a single roll, but would be more complicated though and seems onerous from a player’s perspective to answer the simple question of “what do I know?”. It could work for a more research orientated task – such as searching a library for information, where it could take hours or days to complete the work, or investigatory skills where PCs are searching a place for clues.

So for now, I think, my plan would be as I’ve described above, with graduated target numbers with the possible option of multiple skills providing different types of information.