What are some things that you’ve seen developers screw up in the past? Any pitfalls to be aware of? Do you have any examples of what you think are outstanding metroidvania games?
Any help would be appreciated, my friend has blown it quite a few times and he’s really trying to make this next endeavor a success. Thanks!
I would suggest that your friend try to keep the scope modest, especially if he’s gotten his ass kicked a couple times.
Perhaps it would be best to think of it more like a Contra / Ghosts and Goblins or maybe even a Mega Man 1 style game than a Metrovania game at first. Build out a basic level, work out the basic mechanics, and get people to play test it.
There isn’t that much different between a basic platformer and a metrovania, the latter mostly just being an ‘open world’ platformer. But open worlds need tons of stuff to do and it’s really easy to make a design mistake and get stuck with boring areas or lost players. If your friend got his ass kicked, don’t go for the open world platformer, go for a good solid linear platformer.
If he can get the platformer working and fun, and he has a few levels and is getting positive feedback from play tests, then he can build side missions into his existing levels and basically build out a linear quest line with some optional side quests.
Hold off on going full open world till version 2. I wish your buddy luck.
Guide To Making Metroidvania Style Games
One good way to gather intel is to take advantage of reviews for other Indie games.
There should be some good guidance on what to do in these reviews:
Axiom Verge
UnEpic
Environmental Station Alpha
And what not to do in these reviews:
Humanity Asset
Luna: Shattered Hearts Episode 1
Inescapable
Of course, you’ll have to take the average and read between the lines but still should be a good help in identifying the 3 to 4 most important things to do and the 3 to 4 worst most important things to avoid.
I’ll be sure to let him know what you guys have to say, thanks!
If your friend is really dead set on non-linear - try to convince him to do a Mega Man 1, where you can choose which bosses you go after. This is kind of a compromise, still clear level design where the player always knows his goal, but the player can also make choices about which goal he takes on next.
The player still has options and choices, but the levels will be way, way easier to build than a conventional metrovania and less likely to end with players confused or lost.
Actually I think Metroidvania’s place more emphasis on RPG elements as well. They take the platforming action of Metroid and combine with the RPG elements of Castlevania. Of course, the amount of RPG elements vary greatly in such games but that is the gist of the combo.
I don’t think he cares so much about it being open world. He really seems to like exploration, hidden things, upgrades and lots and lots of jumping, jumping, jumping.
Maybe something like the oft imitated world map of Super Mario World, where you can traverse stages, find secrets and then re-enter other stages would be the best approach. The levels are numbered, so it’s quite linear. To be honest, in my opinion, Metroid and Castlevania both get pretty boring after a while, simply because you’re just running around the same areas and there’s no clear way forward.
Lots to think about. For him. I’m sure.
I really think Mega Man is underrated in terms of it’s design. But Super Mario World is cool too.
But a Castlevania really requires absolutely top notch level design - the bored/confused player problem is real. Definitely not for someone who might not have already had a successful release.
Well, the super mario world map has been imitated by so many and even amateurs have pulled it off perfectly. For me, I’m not even sure I like that open style map you see in metroidvania games. I like straight-forward games.
The guy made axiom verge it took him 6 years, so tell your friend he better be in it for the long haul.
I just watched a 100% play of axiom verge today. The one thing I kept noticing was just how much content was put into that game, how detailed the effects were, how refined everything was. Quality soundtrack and a lot of high quality pixel art. I don’t think my friend is thinking anything along those lines…
At this point it’s down to NES level simplicity.
My view on content (as in needing a lot of it) is that it can be managed by making some decisions up front to go with the simplest graphics style (as in the one you can pump out the fastest and still get the kind of look) you can live with. Not sure if this will work for your friend but it is definitely a valid option.
That’s basically why I included Environmental Station Alpha in that “Good” list above. It has been called the best metroidvania so far by many people.
I’m not saying it doesn’t look cool (it does to me and it took the dev 3 years to make this) just that if a person used a similar art style they can probably knock out a lot more content a lot faster than if they used different a visual style (which just happens to be my view of the art stuff):
In terms of art style maybe something like that, where its like low poly stylized
Look at hyper light drifter. I think to make really good looking pixel art its going to take along time or a really good artist.
Yes, simple is the key to feasibility at the indie level. Sure. But even a game like that has a lot of graphical depth and variety. Take a game like the original Super Mario Brothers, though. That’s a game that is feasible for someone to make. And indeed, look at all the freeware mario clones out there. It’s hard to limit scope. But, it’s the only way.
Just a couple thoughts from the artsy guy of the bunch. 
I think it’s really important to scope to what you think you can do in the timeframe you are comfortable with.
IMO - Design of the characters and (lesser extent) world is super important to gain attention to your friends game.
Along that train of thought, one of the most important things (excluding mechanics) for metroidvania/platformer games is the player character and the enemy characters.
Game play mechanics aside (because those will be driven by the type of platformer created) to separate your friends game from the crowd - the main character has to be kick-ass memorable and visually pleasing (story is part of this) and the enemy characters have to be visually pleasing, memorable and for bosses they really should be as cool and unique as they can be.
Relating to the classic originals, Metroids, Castlevanias, Marios, Mega-Mans, Zeldas, all those games have iconic main characters and just as iconic enemies, especially bosses.
This can also be related to in modern platformers - Shank, Bro-Force, Axiom Verge, Guacamelee, Ori, Odin Sphere - others.
With some quality characters, enemies and (of course) solid mechanics - I think that game - regardless of how long it is or how pretty the environment is - would be a fun play through for most any gamer.
Regarding the environment -
One thing I think is very rewarding to players in the original games (first series Metroid 1 & 2, Castlevania 1-4, Mario 1-4, Zelda 1-3) was the ever changing art palette.
In Metroid the further Samus progressed the more the game (environment) changed, and the same with Castlevania, each level was basically the same (same gameplay) but with different visuals.
I don’t know how manageable this could be for a lone wolf indie, however it’s an awesome thing to offer the players if you - ahem - your friend - could manage it.
I think creating a back end level editor (or finding a nice one that works well on the asset store) could really ease the work required for level creation, but - that still doesn’t ease the development required for the art that is needed.
For progression - Metroid is simply an awesome go to reference for how to build game progression in a set environment that doesn’t change.
First she finds the morphball that allows her to access a lot more areas of the levels. Other add-ons and unlocks that she recovers allows her to progress in the game. Same with Mario - and his tanuki suit, cape and yoshi. All these unlocks allow Mario to traverse other parts of the levels and progress further.
Castlevania is pretty linear, but it’s just the aesthetics, mechanics and story that make these awesome.
Zelda on the other hand could be considered one of the first open world games, really. You could go anywhere in the game - but if you went too far away of course you might run into some baddies that were too tough for you to handle. And you learned by playing how far you could go before you needed to turn back.
Mario is really kinda open world also - with the overworld map. Once you cleared a level - you could return there if desired, and sometimes required to gain access to other areas/hidden levels, once you were able to with a new mechanic.
Tell your friend good luck - and when he can - post wips here so he can get feedback on his works. 
So far, so good. Main character is adorable (it’s a she) and has a very different feel to her than most platforming characters. Seems like the idea is going to be “back to when it all began”. Simple controls, not too many abilities and happily ignoring a lot of gaming tropes. It’s targeted at the newer generation, abandoning the current stream of games from the golden years thru the console wars, where we end up with games as they are today. It’s very floaty and ethereal. The exploration is going to the forefront while combat and all of the stuff we grew up with is taking a back seat. And it terms of the story… just like we’re still scratching our heads about a fire-breathing turtle dinosaur kidnapping a princess, I think this story will leave people similarly confounded. Anyway, today’s little prototypes are promising. I see a real foundation to build a game around, for once.