Do I need to go roguelike?

Using the term in the loose manner it seems to be applied these days. i.e. incorporating some manner of randomised procedural level generation and or controlled randomly generated enemy types etc. (not necessarily high difficulty / permadeath / or turn based gaming etc ).

Do you think it is essential for indi developers to incorporate “roguelike” elements into their games?

It’s clearly a key marketing point and sales driver in many of the successful indi titles in recent years.

Ive played a few shmups recently that randomise the enemy/bullet patterns so its impossible to memorise them which is almost counter to the whole Shmup culture. Also something like Enter the Gungeon. It’s great (and I do recommend it, its also very tough you have been warned), but for me the random generated dungeon and enemy types/spawns are the one thing I didn’t like about it because the focus always becomes about play skill rather than any type of exploration /discovery.

But it seems to me that roguelike stuff is all the rage these days and you need it in your game.

Im asking as Im at this point in my Shmup game engine that I need to either commit to fully scripted enemy/bullet patterns or get my head around trying to code in some roguelike randomisation.

Just wondering what you guys think…and if you are planning some roguelike elements for your games..?

Good question. I think it boils down to what you wanna do. Do you want to make it more roguelike just to satisfy a certain type of player or do you wanna make the game how you would want it to be? Anyway, make the game you wanna make or is capable to make.

I would say... take all the good things YOU like from other games, weather it's procedural or not, random or scripted. I kind of enjoy both.... sometimes it feels good to predict enemy shooting patterns, you feel you learn and get better. On the other hand you sometimes want more randomness.

Why not use both for different kind of enemies (robots and humans)?

Or why not have them use scripted but with some randomness?

I don't see why you need to commit to one or the other just because everyone else does it. Why not use the best of both worlds?

Cheers tunepunk

I think that is my problem.

I don’t know what type of game I want to make.

I started messing with C2 just for relaxation, like someone might take up painting, never had a mission to actually finish or release a game.

So never worried about these things.

But now im getting ideas about actually completing and releasing a game these thoughts introduce a lot of confusion.

The idea of having random enemy patterns in a shmup is almost sacrilege to me.

If the experience is not the same for every player then score based / time based competition goes out the window…Is that acceptable?……I don’t know…maybe I am just too old fashioned in my thoughts….What is the yardstick for measuring success in a shmup against others if it is different for everyone on every play?……..…hmmmm

I like your idea of trying to incorporate some radomocity in another way though maybe through weapon upgrade order or level order…need to go away and do some serious musing on the philosophy of good video games and what actually lights my fire.

One thing is for sure, Im liking the look and feel of Archer so looking fwd to play the full release.....

Don't worry about roguelikes too much. It's true they're popular, but there's plenty of market left for other types of games. I agree with tunepunk. Got a game you really like? Make something like that, only better

By the way if you want to examine a game that does roguelike and speedrun/score-based competition really well you might want to check out Crypt of the Necrodancer.

.... to check out Crypt of the Necrodancer.

it looks bizarre... i will check it out cheers....

Maybe this needs a new perspective. Roguelike today is not Rogue from 1980.

Roguelike today (i think) is in the first place BIG MAPS. And all the other things perceived as typical roguelike are just solutions to be able to offer BIG MAPS.

Procedural (random) generation is way more easy then storing big maps in arrays. Ever tried to make an array that can represent a big map ? You will be surprised by the limits. Random (seeded) is an endless mathematical world that can be calculated on the fly and therefor does not need to be stored anywhere. Not in memory, not in the designs.

With procedural generation there is not need to design each little corner in a big map.

Procedural generation is the way to go when it is about 'only render what you see'.

It is tilebased, because tiles prune probability's.

Permadeath is just a solution to the problem that rises with storing all progress.

Do you think it could be applied to a shmup? One of the reasons most shmups are so short is the amount of time it must take to script enemy waves. you could spend hours creating and fine tuning an end to end pattern for one single wave that a player will kill off in a second. If those waves could be procedurally generated and still be fair, fun and interesting then it might be worth a shot.

Where to start tho.....

If you're making it competitive i guess randomness is not very desirable. If the player does well, they might be accused of just getting a lucky roll, in the randomness. But it all depends on what type of game you end up creating.

I think that the comments from tunepunk & ErekT about 'Where to start tho.....' are the best advise that can be given.

I can add a personal view if you want (and wanna accept that). But that defies the meaning of their comments: "Keep from existing games (genres) only what you like". But i think that a shmup is about fast reactions and the player memorizing levels and enemy attack patterns.

TBH whenever I read "Procedurally generated" I see.... "A hollow game where you need to make up gameplay to have fun"

Games i like have been properly designed. Good levels, good story unless it's something like DayZ where almost everything goes.

i.e I like a set dinner rather than a buffet.

Roguelike is a bit different.

You can decide how you want your character / player to build upon skills / experience.

Even games like the original Fallout is Roguelike imho. A lot of time Roguelike is mistaken for an RPG-like nowadays imho.

There is quite a bit to be done with some controled randomness in shmups and if done right it doesn't need to hamper competitive play.

Parsec 47

The game generates enemy waves and bullet patterns, not randomly, but not completely scripted either. You definitely don't play the same patterns each time and it is mostly about twitch skill. I haven't heard anyone complaining about that when playing for highscores.

Warning Forever

You fight a single boss which evolves as you destroy it and as you die. The evolution of the boss is not quite random, but they are diverse enough that each player has a different experience and from what I remember people played for high scores since there is quite some strategy and execution involved in surviving for the longest time possible.

Zanac

This one is quite cool. The game detects what attacks are killing the player and adapts itself to attack with more of the same. You can purposely kill yourself with easy to dodge attacks so the AI will attack you with easier patterns as you get farther into the game.

I would say 'Yes' only because I play Roguelikes.

But at the end of the day, execution is what matters.

The last Roguelike I played was Darkest Dungeon and that game was great not just because of permadeath and strategy but because the execution was perfect. Everything from the music to environment to commentator's voice got it right. Heck even the disposable troops in that game had their own distinctive personalities.

Thanks guys, appreciate your responses. A lot of food for thought.

DiegoM . Played that Parsec 47 for an hour or so, haven’t heard of that before. That’s my kind of shmup! Simple with cool bulletML.

After some pontificating on what the heck I’m doing with Construct 2 I have decided to forgo any type of procedurally generated enemy patterns for NetOne shmup 1.0. Mainly because, between work and family (and a little gaming), I just don’t have the time to figure out and integrate all these things. I’m already feeling out of my depth trying to get the basics together!

So, for now, I will focus on getting the 1.0 core engine complete, nailing the gameplay mechanics, and keeping the feature set small enough that I might actually complete this sometime within the next year.

...but definitely I will revisit Roguelike for 2.0…