In a sense, we have all played this game before. We have all created the deathcoaster in Rollercoaster Tycoon, or created a pit for all our Worms to fall into, or just run Mario to his death a lot.
So I figured, why not be explicit about having a game where the goal is to kill tiny virtual people? But since that gets boring after five minutes, instead it’s a game where the goal is to kill exactly the right percentage of tiny virtual people. And your tools are the tools of a dungeon master. Traps of all the usual sorts, minions, creatures, and so on.
But it’s not going to be entirely trial and error, oh no. The core game mechanic is probability. Please don’t run screaming! I promise it’s going to be fun. See, the game software will represent the probabilities in nice easy visualizations, and the only math required is simple arithmetic. I
stole found some free spike assets, so let’s use those for an example. Here are some red spikes:
Why are they red? As you build your dungeon, traps and things are color coded to probability. Red means what you think - a low probability that the virtual hero will avoid the trap; let’s say 10%. If this was the only thing in your dungeon hall, 10 out of every 100 heroes would survive. Well, let’s kill more of them.
Now I’ve put two of them together, and here comes your lesson in probability: conditional events. Your hero doesn’t even get to the second spike trap unless they survive the first one. So we send 100 heroes in, 10 of them survive the first trap. Of those 10, 10% survive the second spike trap - one lone hero. But what if the level goal is telling me I want 5 heroes to survive - 5% of my original hero population? We can mix and match different traps.
The yellow floor spikes here are easier, with a 50% survival probability. So 10 heroes come out of the ceiling spikes. Half of them survive the floor spikes. Bingo bango bongo, there’s our 5%.
If this all seems overly simple, that’s because I’ve been using overly simple examples. In a real version of this game, the following would be true:
-The traps would be far less lethal. This would require the player to string together a lot more of them to achieve the desired outcome.
-There would be frequent savecheckpointstations for the heroes. These will reset their survival probability.
-The traps will be much more finely graded in their lethality and the desired outcomes will be more precisely defined.
For later levels, another fun mechanic would be to have different types of heroes. Some are sneakier, others are stronger, others are faster, et cetera. Different types of heroes would have different probabilities for surviving various traps. Then the goal for the level can be multifaceted: 12% tank heroes, 7.5% sneak heroes, 3.6% fast heroes.
Of course you’d be able to share your dungeon halls online, and play as a hero yourself in other players’ dungeons. This in turn gives rise to one more category of dungeon hall: the Impossible category. No worrying about probabilities here. Just trying to make it as hard as you can for other human players.
What does the game look like? I have in mind a 2D or 2.5D sidescroller/platformer style. Mostly linear hallways, with some larger open spaces, but also branching paths. Let’s say you break your hallway into three routes. Well, if you have 300 heroes coming through, you’ll get 100 going down each pathway. Unless you make one of the pathways look easier. This is a good way to further adjust your death percentage and make more interesting dungeon halls.
I would be remiss if I didn’t mention the recent article about Guild of Dungeoneering, which is a very similar concept, but one that uses deckbuilding and stats instead of a raw probabiility mechanic. Their game is more complex (and, of course, actually real), my game is more like a puzzle genre dungeon builder. And, of course, their game is about training strong heroes, and my game is about killing them.
What are your thoughts? What would you like to see in a deathtrap designer game?
NEXT TIME: We move from D to E. Encroaching ECMs Encounter Engagement!
EDIT: Kinja really doesn’t like me trying to do one image at a time. If anyone has ideas for how to clean up the formatting of this post, please let me know.