Thoughts on modelling the activity of NPC adventuring parties

After the events of my most recent session as DM, I'm left with a problem and opportunity.  The party intends to take a few days out before continuing their exploration of the ruined temple.  That's fine - they certainly have the right to proceed at their own pace.  However, such breaks in tempo can't be free of consequences.  When a site potentially rich in treasure is discovered, other adventurers may seek it out - and in the absence of another group of players to sic on the place, I'm obliged to use non-player characters.

The tables for drawing up a random party in the Non-Player Characters section and Appendix C of the DMG are largely adequate to task, though I did modify the chances to make space for the Barbarian and Cavalier before rolling.  I suppose a pre-gen party from a module would also have done the trick, but there's something quite enjoyable about discovering the PCs' rivals with the help of the dice.

Far more difficult is fairly simulating the actions of a rival party in the adventure location.  I took a quick glance at solo rpg supplements, but these all seem to be based on creating a virtual referee rather than virtual protagonists.  So I'm going to improvise.  Picking a direction for the NPCs to go in is simple enough, morale rules can handle their approach to combat, and it's usually easy enough to see when a party has pushed their luck far enough.  (If not, the game will tend to correct misjudgement!)  However, while the NPC's abilities and Intellect give some cues as to their problem-solving ability, if this is going to be a way of representing the world rather than just me "deciding to mix things up", I need to leave the final decision in the hands of the dice.  Come to think of it, I probably need to draw up some odds on this kind of intervention by NPCs so that rival parties don't just turn up at my whim.  But for a trial run I'm content to put off that bit of work.

OK.  Time to just throw down some numbers.  My assumption is that an NPC's level is the closest thing to the "play skill" factor that their PC counterparts enjoy. Although at low levels a certain degree of natural inclination might alter the NPC's (or player's) ability to solve puzzles, the most significant factor is experience.  For now, I think that an NPC can try to solve a problem only once per expedition, crudely representing the way that players tend not to be able to come up with fresh ideas in a session but often come back next time with new solutions.

Modifiers to solve chance
5% as the base chance
+2% for every character level
-1% for every point of Intelligence under "6"
+1% for every point of Intelligence over "15"
-5% for "Dull" intellect
+2% for "Ponderous" or "Scheming" Intellect
+5% for "Brilliant" Intellect

Modifiers to solve time
1d4 turns, modified to a minimum of 1 turn
+1 turn for "Dreaming" or "Ponderous" Intellect
-1 turn for "Active" or "Brilliant" Intellect
-1 turn for a character of level "10" or above

Of course, this doesn't quite work for situations in which the NPC party must first notice the problem - especially when not noticing a problem has consequences, as with a trap.  Again, I think play skill (level) is the most important characteristic, but some abilities and traits ought to have an effect.

Modifiers to notice chance
15% as base chance
+2% per character level
-1% for every point of Wisdom under "6"
+1% per point of Wisdom over "15"
-5% for "Dreaming" Intellect
-2% for "Foolhardy" Bravery
+1% for "Cowardly" or "Craven" Bravery
+2% for "Scheming" Intellect

Although these odds are relatively low, the cumulative effect of a party's problem-noticing and problem-solving prowess is significant.  The NPC party I've created to rival Aegis, Hulk, Kweeth, and Vhondrel has about a 1 in 2 chance of noticing problems and better than 1 in 4 to solve problems.  We'll see how they go.