Tribulation

Tribulation: Artist's Statement

My name is Sebi Samadi and I am a 5th year business student at Northeastern. The game I designed is called tribulation, which core focus is navigating obstacles and being able to overcome adversity in a journey. You are one of a hundred runners facing five trials, with a gatekeeper to fight between each one. I chose this idea because I always liked a game that was in constant motion, this was specifically inspired by games like subway surfers, geometry dash, and temple run, but had a nice twist where there is a battle in between each round

The first decision I'd defend is throwing the original game away. My first version was a top-down arena fighter, and it worked. It was also too hard, and it felt too close to the example games we were given. I asked Claude to redesign the game, instructing it to look for some more classic inspiration in games where the map is changing as the player continues. The first game was also too hard so I asked it to decrease the difficulty. It was able to rebuild the game completely by just these simple prompts. ​

The second is the difficulty. After the rebuild the runner was too forgiving, so I asked for it to be as hard as the first version. A game about tribulation that you clear on the first try doesn't mean anything. I clarified though that although I wanted it to be as difficult as the first one in terms of passing through a specific round, I didn’t want the player to die as easily, because this was the challenge with the first game. I realized that the health bar was very small.

The third is the Resolve meter. Every stumble fills it, and when it is full you can spend it to break through an obstacle. I liked this option and defended it because I think it adds some good individuality to what could be considered a generic game concept (although I think most games fall into that category). There is also different abilities for the player, like gaining health bars, or going through a period of extra fast speed.

The collaboration was real in places. I gave Claude the context of the other games as a base so it could understand the ability and criteria in which it needed to design this game. So I think it had a good understanding of the level of advanceness, in terms of coding and design. The collaboration came in more when I started to manipulate the game within each version it gave me. I think I landed on 5 total versions, prompting it between versions to change little things. For instance when I wanted to add a battle round in between I clearly prompted Claude to do so. “Between each trial, add a short side-on duel against a gatekeeper. I have to beat them to open the gate to the next trial. Keep it to three moves: strike, jump, duck.”

I also had to have it compare between trials so that it could get a more accurate reference of what I was referring to. “Trial 1 is too easy and trial 4 is unfair: obstacles appear with no time to react. Give me at least half a second of warning on every obstacle and make the speed increase smoother between trials”​

Ultimately I do think the game is mine, but I can recognize that it didn't take much in terms of actually building it. Could someone replicate an exact game without having the code? Probably not because I don’t think anyone would have the same exact thought processes with prompting it. But I do think claude did 99% of the heavy lifting here.

AI disclosure: I used Claude code in the app visual studio code to help me complete this. I believe it’s the same exact thing as claude code, but runs a lot faster on my computer for some reason, when I used claude code in the actual desktop app it is very glitchy. I didn’t have many rounds of iteration, maybe 5 total and although the prompts were intentional they weren’t to extensive by any means. I was able to give Claude good insight by letting it observe some of the sample games and asking it to examine the code for reference on what would be the criteria to create a game that although is designed differently, falls under the same category of design and complexity. I had to correct minor things like the game formatting, difficulty, abilities of players, and other things similar to that.