Awesome, I think it’s a great exercise! I’ll be assuming that you’re newish to gamedev as a hobby. The more that you’ve done game jams in the past, the less you’ll need the advice here.

That being said, it is a lot of work to pull off a Metroidvania, even spread over a month.

Finishing a game in a month

My first bit of advice is to team up with folks. Some folks can do all of “art/level design/music/code within a month”, but it is a properly large amount of work, especially for a Metroidvania. There is always the chance that folks will have life plans change during the month, as well, so do be gentle there.

Try to leave enough time for art and audio to make it into the game after they’re finished. That is a process that will take time to set up, and, as someone who’s done last-minute asset importing and integration, you want to avoid it if you can. Generally, I’d give at 3-5 work sessions for getting things like animations and sound effects and music tied into the game, and I’d also budget some time for tweaking things to fit, unless you’re far better at pre-production than I’ve managed to be.

Every jam advice article warns about scope creep. These warnings are forged in the hours of lost sleep from overcommitting and procrastinating.

A good first step to help combat scope creep is to have two buckets for your ideas: The big version that has everything you’d like to add in an ideal universe, and the minimum viable product: The stuff that is essential for the game.

That being said, “This is the End Of The Demo” shows up in many MVM entries, even ones that do well. Remember that the player won’t know what your original planned ending was, so if you need to turn your mid-boss into the final boss, don’t feel the need to apologize for it, just do it, and then adapt bits of lore/dialog as needed. Being adaptable about the shape of your game is useful for reaching something that’s finished.

(There is an exception: It can be quite successful to imply a much larger final product than you have time to build, but build the bits you do have with a lot of polish. Fodio is a good example of this, though it’s not a path I’ve managed to go down myself, yet)

If you get stuck with a programming challenge, feel free to reach out in the MVM server, it’s got a helpful and experienced bunch of people

If you’re struggling to gauge how much you should aim for, I’d suggest the following targets:

Relatively new to game jams or low staff/time:.
1 boss 1 biome
2-3 enemies
2 upgrades

Starting to gain experience and have some extra free time:
2 bosses
2-3 biomes
3-5 enemies
4 upgrades

Heavily experienced and have a lot of time to spend:
3-4 bosses/biomes
4-6 enemy types, or trade off enemy art for more environmental stuff or hazards
4 upgrades

If in doubt: Start smaller, and be prepared to cut content mercilessly. Smaller than that,cutting even more than you expect.

As an anecdote: For EYES (in MVM 19), I had plans for easily twice as much content, and a whole mechanic set that didn’t make it in. As time passed, I ended up cutting at least two planned bosses, a biome, and two sets of mechanics. The game still ended up placing 3rd in the jam.

Some further considerations regarding game size: As your map gets bigger, you’ll need to add more features to your game to counteract the size of the map leading to confusion. At the most basic, distinctive level designs help, then signposting landmarks, and, if your map is big enough, you’ll want a minimap.

As your map becomes more open, and less linear, you’ll want to also add ways for players to figure out where to go next. This is where level design can help, if there’s a minimap, and only one door for your new ability, that’s a lot clearer than if 3 of the next 4 ways you can go lead to dead ends.

If you’re in a position where your game starts to take you longer than 10 minutes, first of all, congratulations! That’s a big milestone! Second of all, start keeping an eye on how long your game ends up. A game that takes 2 hours for someone to finish won’t see many folks reach the end. It can work if your game is well paced and varied, but that is a lot to pull off in a month.

Design Tips

If you’re stuck with how to start on level design for a Metroidvania, start by going small. A good place to start is with a 3x3 room layout. This gives you a small starting point, but it gives you a lot of ways to easily expand out as ideas start to flow.

Another thing to keep in mind is how players will navigate the overall map, as they gain abilities and more places to go. If you can, building a minimap is quite helpful. If you don’t have the time for that, setting up memorable points of interest helps. Another consideration is that your map doesn’t have to be super open: Sometimes having multiple linear sections that loop back around to a common starting point helps give that sense of exploration that helps make a Metroidvania special, but without players bumping around the map with no idea of what to do.

If you’re doing a collaborative map design, using google sheets to do the map layout can be useful. It also allows you to track who’s doing what on a per-room basis as well.

Your choice of art style matters! Remember that whatever art style you end up going with, you’re going to need to make at least 1 tileset, and a set of player animations, and animations for enemies and power ups, so choose something that the active artists can do in a decent amount of bulk. I ended up wasting nearly a week trying to make a 2.5D projection effect in Scrap Hive. It looked cool, but cost a lot of time I could have used to make the mapping better. EYES, by contrast, went with a 16x16 art style, since that leaned into the strengths of the artist on the project.

Alongside with art style, if you’re building melee animations for the main character, animating them on a grid that’s three times larger than the base character helps give plenty of room for bigger hitboxes and expressive animations. If your game’s tiles are on 16x16, you’d animate your character on 48x48.

General Dos and Don’ts

If this jam is the first time your team is making a game with the chosen set of mechanics (aka, everyone’s first platformer, or 3D FPS, or RPG), plan smaller than larger. A few rooms and a basic boss, provided that all of the mechanics work well, is much better than a larger area without clear boundaries.

Which brings up another point: If you have an incomplete experience, deliberate or otherwise, please make the edges/end of what is there clear. Knowing that I can put down the controller is far better than wandering around for 5 minutes wondering if I missed a thing.

You, the game coder/dev, should start by building solid core mechanics, and then move to anything risky in the game design. They don’t have to follow another game exactly, but your core movement should fit with your level design, aka jumps should only be as many tiles as players can make. My current favorite example of a non-traditional set of platforming mechanics in an MVM entry is probably Saturday Night Vania. The jump arc is super low compared to most platformers, but the level design is built around it, so it ends up working out well.

You shouldn’t build difficulty from tedium. There are many ways to add challenge into a Metroidvania, but damage sponges are boring. There is an element of taste here, but if you’re spending multiple moments at a time doing nothing but pressing a button to do damage, without having to dodge or otherwise make a reaction or decision, the enemy is probably too tanky. The same goes for super long platforming segments with death pits and few checkpoints. There can be ways to make it thematically appropriate, but this is one of those things you want to learn to do right, before you break it. Highly punishing platforming segments (or the same in other genres) also get tedious after a while. Variety helps make a pleasant experience. A good question to consider is “can I get this level design idea across in fewer ‘screen’s of map than what I’m currently using?” The more screens you have, the more you’ll need to find ways to mix things up, the fewer screens, the more your existing variety spices things up.

Tryhard mode: Aiming for First Place

Getting first place in an MVM game involves a few factors: The first is that you need to hit certain baselines: Audio needs to be present and implemented competently, your movement and level design need to be well-polished, and you should have enough content to tell at least a short story

Beyond that, the fight for the top place is often more defensive than it is offensive. The less folks have to complain about in your entry, the higher it’ll tend to be rated. Crucially, this means that the more content and features you put in the game, the more chances you have to implement something wrong, or in a way that will lead to lower ratings.

Another thing is that there are certain mechanics that are… hard to please everyone with. One example that can be controversial is wall jumps. Finally, there’s going to be an element of randomness. Maybe someone plays your game especially tired, and rates it poorly because they couldn’t get past an execution test. Usually, the top 3 or so game are very close in ratings.

Finally, you’ll increase your odds by having a well-performing web build, and by rating other folks’ games and leaving helpful feedback. MVM 18 had a lot of neat feedback sharing, and I’ve love to see it come back for MVM 19.

Good luck, and remember to sleep and drink!