home / content / blog

game ideas

posted 2023.10.29
send feedback

I very slowly make games, and my main goal at this stage is to make a lifestyle game. That is, to be satisfied playing almost exclusively that one game. Some games commonly played as lifestyle games are chess, go, poker, bridge, Magic: the Gathering, Dota 2, League of Legends, and Counterstrike.

This means I have to design a game I’ll like. I have some ideas for doing that, but they’re not finished games, so clearly some work needs to be done. (Or maybe Auction Tic-Tac-Toe was the answer all along…) It would be very tailored to my specific taste, so here’s two statements that will give you a pretty good idea of my taste:

I am soliciting suggestions for how to progress with these ideas and examples of existing games doing similar things. And I may follow through on these ideas one day if the design becomes compelling enough. But it would be way easier if someone else did it.

Mayhem Manager (in progress)


I like sports management games. That is, games where you manage a sports team. Signing, drafting, trading, and coaching players to try to win championships in a simulated sports league. The most popular ones are Football Manager and Out of the Park Baseball. But I think almost every game in the genre makes the same few mistakes:

  1. Most importantly to me: They tend to reduce players to a single number, signified in nearly every game as “OVR” (overall). A 90 OVR is always better than an 87 OVR is always better than an 80 OVR. Except in some rare cases relating to the positions they play, like if you have two good second basemen and no good catchers. But at the same position, better OVR is nearly always strictly better. Players of these games often talk about “team composition” as if it’s important, when really it doesn’t matter. But it should! Players should be distinct enough that they can be much better on one team than another. Instead they all wash out into a set of 3 numbers (OVR, age, and contract).
  2. A more minor quibble is, I find that these games (save for OOTP) give information at much too precise a level. I just think it’s ridiculous to say one guy has 84 accuracy while another guy has 82 accuracy.
  3. These games are either way too luck-based or way too easy.1 And usually they are way too easy. At which point championships aren’t enough and I have to make my own fun, like trying to go undefeated in the NBA or get a RB to score 40 touchdowns in a season. In some games the difficulty can be tweaked, but not in a very fun way.

I’m also inspired by deckbuilder games like Slay the Spire and Dominion. I like picking things that have synergies with what I already have.

The plan

The idea here is to alleviate the main flaw of sports management games (players reduced to numbers) by adding synergies like you might see in a deckbuilder. (Or in many games that let you customize your build.) Real sports aren’t nearly synergistic enough in this way so I am inventing a new sport, gladiator combat with magic, for the game.

In Mayhem Manager I am currently addressing point 2 by only rating anyone on 6 stats out of 5 stars in half-star increments, which is not very precise. And I plan to address point 3 by just writing better AI.

This is designed to be a good party game or strategy game depending on how you approach it. Silly flashy graphics and fun abilities for the party side, tough decisions and interesting synergies for the strategy side. Playable multiplayer and/or against AI. I would also like to have a sort of “global” league system with promotion/relegation, because I think it would be cool to have a sports management game where a serious player can try really hard to make the top league.

Open questions

One question is how fighters and equipment in the game should synergize. One point of inspiration is party compositions from class-based FPSes and party-based RPGs. (MOBAs should be included but I don’t understand or enjoy them.) The roles in these games tend to be DPS/tank/support/pick, which maps roughly to wizard/fighter/cleric/rogue in an RPG. The other point of inspiration is deckbuilders and TCGs, like tribal decks or combo decks in MtG, where cards synergize either explicitly or by exploiting an interaction between them.

A second question is how to make the fights feel fair or entertaining. It would be a completely different game if you could directly control your fighters, not to mention technically impossible for me. So I currently have it as an autobattler, which means the fighters must be controlled by AI, which means the AI needs to put on an entertaining and fair-feeling show. I think Salty Bet does a good job of this, but it’s aided by the total incomprehensibility of its fighters’ movesets.

Three Birds One Scone


Three Birds One Scone is the proposed sequel to the demo game Two Birds One Scone, in which you are a triangle and you throw scones at birds to prolong your life. The goal, like in an infinite runner or a Super Hexagon-type game, is to survive as long as possible. You throw your scones in an arc at birds that briefly appear on the screen, and each bird you scone adds time back to the timer before you lose. Hitting multiple birds with one scone provides exponentially greater rewards. There are also squares flying across the bottom of the screen which you have to dodge, because you lose half your remaining time if they hit you.

The plan

One order of business is to give the game real graphics. Another is bringing it to a browser, so you don’t need to download and run a Python file to play it. And another is adding a high score tracker and a global leaderboard. But beyond that, I would like to [REDACTED: SPOILER].

Space smuggling game


The reason I came up with this idea is I couldn’t think of any roguelikes that don’t revolve around killing a series of enemies.

The closest points of comparison I have for the mechanics of moving between planets and managing your crew are FTL and Pixel Piracy, two games I’ve played for maybe four hours total.

I have a similar complaint about party-building games as about sports management games: that party members are reduced too easily to a collection of stats, which are too precise as well. You just met a guy and you instantly know he has 14 strength?

The plan

There is a galaxy with a bunch of planets. You start on the fringes of the galaxy and your goal is to make it to the center. All you have is a ship, but you can’t afford the fuel to get to the center. So you have to pay your way by smuggling contraband.

The core gameplay loop is:

  1. On a planet, you talk to a contact who offers you smuggling jobs, like “deliver X cargo to Y planet”. You can accept or reject. You can also accept much lower-paying legal shipping jobs. You may also meet people on some planets and hire them to your crew. And you can buy fuel and ship upgrades (or whole new ships) on some planets too.
  2. You load the cargo onto your ship and set off for another planet. It takes more fuel to travel longer distances.
  3. When you arrive at the new planet, you go through space customs, where your ship will be inspected. If your crew doesn’t conceal your cargo well, it could be confiscated or you could be put on a list. Or thrown in space jail. If you get thrown into space jail you lose.
  4. Now that you’re on the new planet, you can deliver your cargo, get paid, and go back to step 1.

Your success at not getting caught will depend on the amount of contraband you carry, the quality of your ship, the way you arrange your cargo, the composure of your crew, and your strategy. You can use tactics like distracting, bribing, and intimidating inspectors.

As you progress toward the center of the galaxy, shipping will become more lucrative, but inspections will become more strict. You will need a more advanced ship, a more prepared crew, and more believable fake documents. But if you stay too long on the fringes, you will arouse more suspicion there.

When you meet people, you will only get a general impression of their personality and skills. Over time, you’ll get a better sense by observing how they perform—will you need to fire someone who blew your cover? Or was that just a one-off event? What if someone isn’t as skilled as you’d like, but they’ve been with you for a long time and you want to have someone you can trust?

Open questions

How should the customs inspections work? More like FTL or a dialogue tree? Should they require a great deal of skill to handle, or merely result from the skillful preparations you already made? I would like to give the player a great sense of tension as they are being inspected, as you might if someone is nearby when playing hide-and-seek.

Thematically, you’d expect this game to be uneventful except for occasional disastrous moments (mainly getting caught). How can these disastrous moments be made to feel fair? Is this even an issue?



Many of the most elegant and popular strategy games have very simple rules, given depth by their use of space. Chess and go are prime examples. Modern strategy board games seem to neglect this compared to older ones (though not completely, of course). And card games, maybe my favorite type of strategy game, often involve no spatial elements at all. Magic: the Gathering has none, Dominion has none, and Hearthstone only has a handful of cards that involve adjacency. The others I’ve played have none I can think of, except maybe the movement in Clank!.

I’m also inspired by my favorite Magic: the Gathering format, three-card blind. In this format, your deck is only allowed to have three cards total. The lack of cards forces you to trade off between the number of cards you play, the power of those cards, and how early you can play them. This is because you have to use some of your cards (lands, generally) to play your other cards (spells, generally).

The plan

The game is so named because it’s about teams of robots sumo wrestling each other. The prototype version of SUMOROBO uses a hexagonal board, like the board in Settlers of Catan. (It was playtested on the exact board from Catan, in fact.) The rules were:

  1. You draw a card each turn (from a deck of playing cards ace through seven).
  2. You can place an ace down on the near side of the board and it’s your robot with 1 strength. Or you can place a two if you discard a card, or a three and discard two cards, and get a robot with that much strength.
  3. Then on each turn, for each card in your hand, you can move a robot by one space. You can push the robot in that space if yours is stronger. Two robots can push one robot (or one can push two, or two can push two, etc) as long as your robots have more combined strength than the ones you’re pushing. A robot pushed off the board is gone forever.
  4. You win by getting a robot to the center of the board and keeping it there for three turns straight.

The vision is that your cards would have abilities beyond just moving and pushing. Some could do special moves—imagine a robot that could PULL! And some could buff your other robots, or what have you.

Keep in mind this was a haphazard prototype. I’m not so much convinced by this ruleset as by the general concept of a card game with strong spatial elements.

Open questions

The design of the game feels like there should be a clear first-player advantage. There hasn’t been a decisive advantage from the few test games played, but I think that’s just because no one is skilled enough for it to matter. Like how bad chess players win with black nearly as much as white, because white’s advantage is dwarfed by the advantage granted to the opponent by each blunder. I don’t know how to balance this advantage but I don’t think it matters yet.

I’m concerned that this game would have very long turns, which is true of most strategy games with near-perfect information. I don’t want people to need chess clocks so I want to find a way to reduce the amount of lines of play people are encouraged to calculate.

I’m also concerned that the game board that is too small. A larger gameboard takes better advantage of space because it makes locality matter. Basically I think locality lets you lower the processing power needed to think about a deep game. For example, I don’t have to worry about a pawn in the bottom-left of the chessboard very much when I’m launching an attack in the top-right. I’m pretty convinced this is an issue, but the question is, can you make sumo work on a larger board? Is it even worth keeping the sumo part? What would it be instead?

Be The Factory


As with the previous idea, I think space is an underutilized element in games. In this case, I’m inspired by grid-based factory creation games like Factorio and SpaceChem. Or perhaps redstone from Minecraft. I think there’s a beauty to having a mess of interlocking pieces that make Production Go Up.

I’m also inspired by the trend in board games of having to fulfill randomly selected scoring rules, like “2 wood = 1 victory point”. I don’t like victory points though.

The plan

You’re a factory and you get offered contracts to produce goods. The money from those contracts funds new equipment and upgrades to your equipment. You place the equipment in a grid that runs every day, producing ever-more goods. If you miss too many contracts, you go out of business. Your goal is to stay afloat as long as possible (or to last until a set time).

Open questions

How permanent should equipment placements be? I want to force questions like, “Do I place this where it helps the most now, or do I save that spot for something better?” How can some spots be made much more important than others? Should you be allowed to sprawl as much as you want, or be limited to a small space? Would it get repetitive running the same factory over and over? And if you have a good start, what stops you from cruising to a win without much thought? One benefit of making it about lasting as long as possible is that you can never rest on your laurels.

How many types of goods do you need to make? How do you reduce the influence of luck from whether contracts match your specializations?

How can tradeoffs be introduced between short-term and long-term goals? Would this cause snowballing?

Other assorted ideas

Settings game

A puzzle game where the entire puzzle is within the settings menu. The player would have to change the right settings in the menu to complete the puzzles. I don’t know if it would be based around a puzzle “formula” like Baba Is You or if it would be a series of relatively distinct puzzles. The hard part about this kind of game is coming up with novel, interesting puzzles, so I suspect it would be a short game. I also have a very fun plan for how the game would start, which is [REDACTED: SPOILER].

Rule interactions game

A puzzle game where you have to exploit interactions between seemingly sensible rules to solve the puzzles. So far I have the rule that inspired the whole idea and one more that I think could work. I think it might communicate at least “the vibe”:

Theft roguelike event

An event in a roguelike where you come across a farm with a bunch of supplies. You can take them, but when you leave, a powerful guardian fights you who is angrier the more you took.

Build-altering roguelike

A roguelike where you can see the next few enemies you’ll face and have to tailor your build heavily towards them to have a chance of success. You would have frequent loot options but a tiny inventory, so you would constantly have to make tough tradeoffs between general value and value against the enemies you will immediately face. One open question is, how will you progress long-term? Would it feel hollow if none of the choices you make last very long?

Game with a playbook

Football coaches spend a lot of their time designing playbooks. I’d like a game that encourages players to “draw up plays” themselves. One of the main goals of a playbook is deception: misleading opponents about what your play will do so they will not match up against it correctly. How could a game be designed to encourage this? And especially to encourage players to develop this individually and not netdeck? (Does that matter?) I can’t tell if the answer is, “this does exist, it’s called deckbuilding,” or this is really an unfilled niche.

High score non-competitive game

There are some singleplayer and co-op games where the goal is not to “win” but to get a high score. But most of these are based on reflexes and precision. I’d like one that’s not real-time, involves a great deal of strategy, and is (preferably) playable as a board game. I’ve played a few of these, but they tend to be too simple or too boring. And there’s a strong tradeoff between high average score (unexciting to strive for) and chance of getting an extremely high score (very luck-based). I think this could be mitigated if it were about lasting a long time instead of maximizing points in a fixed time. Is this type of game doomed or has it just not been made yet?

  1. Football Manager and OOTP are more difficult and I respect them for that, but their seasons take much too long to simulate for my taste. Front Office Football 8 is also respectable in this department and I have played about 100 hours of it. But I find its scouting a little tedious.