Take

Spoilers below. I encourage you to download

and play the demo before reading further.

Scope out the target

This prompt was one of those infuriatingly generous ones. Essentially, make whatever you want, and impress us. One of the main reasons I seek out work at game companies, rather than making games on my own, is that while I’m great at working outside the box, I still need the box to exist. So, in lieu of more specific requirements from the prompt, I gave myself some requirements. I would push myself to work in the terrain editor, something which I hadn’t used before. I would work to add verticality to my level, something I am actively practicing. Finally, and most importantly, I would create a complete (but very simple) gameplay mechanic utilizing discrete reusable actors.

Make a Plan

Gameplay mechanics didn’t come with the assignment, nor with the template, so I had to make them. I didn’t have too much time, so to make sure I had time to work out the kinks I kept the mechanic simple. I wanted to be able to interact with nearby objects in the world so I’d need a few bits of code. First I altered the template character blueprint to give me an off-shoulder view and perform a line trace from the camera identify what the player was looking at. Then I created a new actor type for objects which could be looked at, PoI Actor. I had to create a simple highlight effect on the PoI to tell the player which objects were intractable, and I settled on a simple gag of upscaling the base mesh and giving it a blue transparent texture. That worked for Static Meshes fine, but I ended up needing to also create a different PoI for the Skeletal Mesh so that I could get the highlight to animate and use poses. But I digress, both would have the same basic functionality: Highlight if looked at. Broadcast if either interact button was pressed. And have some data about the state of the object like the look at text. The final part to set up was a text overlay in the character blueprint that I could use to display text to the player, by sending plain text strings to from anywhere in the code. With those in place I could just start plopping these PoI Actors around the level where needed.

Next was getting back to the gameplay. My very first two puzzles got skipped right away. I had ideas for a puzzle based around getting access to the party, but those were stretch goals. The top priority was to create a completable heist first, then I could add in complications later. The first puzzle then was also the beginning of the final heist, steal a means of escape. Steal a valet ticket and get the valet to bring you the car. This would set off a one minute timer which I’d hoped would cause the player to scramble to try to find and steal the artifact even without a plan. Having failed that the player would then know the parameters of the challenge.

About that artifact, have to set that up, don’t I? I’ve already made a nice place of honor for it, but I need a security system to prevent it from being stolen. Simple enough, one big trigger around the object. I could have made it so that the player could interact with the artifact whenever they wanted in order to start the heist, set off alarms and countermeasures, but instead I wanted to give the player the clear message that they weren’t going anywhere without a plan. So instead, when the player gets near the artifact they a bit of internal dialog about how the character was planing the heist.

This also let me create a hint system. If the player knew about some step of the plan to escape, the character would presumably know about it too, and would be wondering about the next step in the plan. The slightly awkward part of this mechanic was that sometimes the player needed to interact with something, and sometimes, they just needed to know that it existed. For the sake of expedience I erred on a simple mechanic, once the player interacts with a PoI it gets flagged as known. And with that it was just a matter of creating a cascade of if the player knows about X and Z, then the the character’s hint will be about Y. Finally there did still need to be a way to steal the artifact. I made it easy, if the player gets close to the artifact while the heist timer is running, instead of some interaction, or dialog, they just steal the artifact and set off the alarm.

Execute the plan

With the beginning and end in mind, and the means to achieve goals through the point of interest system, I had all the tools I needed for a successful heist.

Steal the ticket from the guest on the ramp

I set up the PoI so you had to stay close to him while he walked in order to pick his pocket. Unfortunately making him move for the demo would have cost too much dev time since no other characters moved. So the player just gets to enjoy the view for a moment.

Lock the door on the second floor from the inside

This started as a simple concession to the architecture I had developed. If there were doors upstairs there should be a guard upstairs too. But as the plan continued to unfold I saw that it could also be an excuse to make the player walk past the solutions for the next set of problems. See the pool from the porch and see the booze in the bedroom.

Get the booze from the bedroom on the second floor and spike the punch

This puzzle had the most adventure game logic of all the puzzles I included, but it led to a lot of fun things like the party music getting turned up, and more people dancing. A fun and visible change in the world which few of the other puzzles really matched.

Push the party goer in the pool

I tried to do a lot of telegraphing for this, so that you knew that being the only one wet would make you stand out to the guards, and that this dude was your fix for that problem. If I had more resources and time I'd have liked to make him and his friend talk about crazy parties ending up in the pool. Hearing them talk would help the player notice and locate them better.

Give the ticket to the Valet

This is when the heist begins, put 60 seconds on the clock and GO! You could do this as soon as you stole the ticket, but you would fail. I don't really like a learn by failure gameplay model, but I do think that if a player is going to fail the lesson should be clear. In this game you need a plan.

Push the tablet in front of the door

The tablet tipping over will only happen if you are already on the clock, otherwise it would be too easy to fail before you’ve even started. The fact that I stopped you from experimenting like this is what originally inspired the hint system. With the hints in place when the player 'used' the tablet before the heist starts, the hints would add that step to it's ruminations going forward, responding to the players actions, but not by forcing them to learn through failure.

Grab the Artifact and climb up to the second floor

As part of my design process I usually ask myself, what was a complication, what was the solution, and what did the result dictate. In this example, The player just blocked their exit in order to lock the guard out, so now they need to find another way out of the room, so the only way out is up. So, I added a few variously sized display cases.

Jump into the pool to get downstairs fast

In order to make this a gameplay element, and not just a bit of fun, I added fall damage to break the character’s ankle if you tried jumping down anywhere but into the pool. There were no deadly heights involved, so all that happens is you get a hobbled walk anim and a fail message if you do hit the deck instead of the pool. Breaking your ankle was one of only two ways a player can fail during the planning phase.

Climb down the ivy

The ivy puzzle was the first I thought of while designing the space, though initially it was going to be the ingress to the party. Unfortunately, since it was so obvious to me throughout development, I think I may not have telegraphed it as a potential solution. That's where I missed being able to have others who could playtest.

Get to the car in time

The Valet gives you one minute before the car shows up, and 10 seconds from then to get to it and get out. This had two functional effects. You knew that your plan had to be executed quickly, even if the planning didn't. And since the timer is actually counting down to when the car arrives, there is always a moment of tension and relief as you pull off the heist with 0 seconds on the clock.

I didn't intend to only develop the one correct solution, I wanted to create a scenario filled with natural causes and effects. For example the player could steal the artifact, climb to the second floor, then walk past both of the guards who were distracted by trying to break into the locked gallery doors. But they would only be able to as long as the player character walked, remaining inconspicuous. Of course if the player walked from the second floor, around the porch, down the stairs, past the pool, through the guardhouse, and down the hill, they would never make it in time. But I wanted to let them try. Since I was now checking when the player ran past the guards during the escape, I made the guards yell at you for running whenever they saw you. They wouldn't arrest you unless the artifact was stolen first, but their mild rebuke would tell the player that the guards were watching, and they were paying attention. Setting up this element led to another fun idea, if you jumped in the pool the guards would also notice that you were all wet which lead to the second of the two fail cases which could occur before you stole the artifact. Party fouls will not be tolerated.

Twist ending

It may have been obvious to everyone else before this point, but it wasn't until the whole project was finished and submitted that I realized that for all of my original thinking, as laid out above, I had essentially re-invented the Hitman series’ primary gameplay loop. Which is odd since at that time I had yet to actually play a Hitman game. Chalk it up to ludo-narrative carcinization.

Galleries

It’s just like playing the demo, but less fun.