Design I have several ideas for a turn-based game. Which would be the best to start with?

[Selek]

Hi all,

I've finished my first game in GML, a clone of Tetris. (I posted a screenshot in another thread earlier.) Since then I've been using pen and paper to design prototypes of a possible turn-based game of my own design. Mostly I'm interested in building my skills, but it would also be nice to make something at least slightly original. (I have a lot of experience as an artist, but I'm relatively new to the programming thing, though I was a comp-sci major for one semester in college decades ago. Back in the Jurassic era, when we used APL and Pascal and FORTRAN.) For most of these ideas, I have a theme in mind. I'd like to keep the project small and feasible to finish. It took me about a week to make Tetris. I'm willing to spend several weeks on the next project, but not months. Here are some candidates.

1. One idea is a small turn-based square-grid-based tactical wargame. Like a small, simpler, prototype version of Advanced Squad Leader or X-Com. My admittedly unoriginal theme would be WW2 platoon- or company-size combat. But any territorial-control theme might suffice: lions controlling the savannah; birds controlling the lawns with the best worms. I'm actually more interested in operational and strategic wargames, hex-based, with lots of juicy resource and building decisions -- but I know enough about the complexity of such games (not to mention hexes) to know I shouldn't start with those. From a programming standpoint, pathfinding (A* stuff) would be great to learn. I've watched some of sgt indie's videos on this, and read some theoretical articles on the subject.

2. Another is a deckbuilding cardgame, again on a small scale, in the same genre as Slay the Spire or Monster Train or the boardgame Dominion. (Possible themes include building a business, or running a hospital, or building a prosecutor's case.) Each hand would take just a few minutes, but would be grafted onto a larger metagame or roguelike structure. I have a few ideas on how to inject a little innovation into the genre. Maybe this is more manageable than option 1 because no grid or AI? Also it could start as just the deckbuilding loop and expand to the meta game if it were going well, or not.

3. Yet another would be a tiny 4X game (or 3X, with no eXterminate). Space setting, or perhaps a Wild West setting, playable in an evening rather than over 40 hours. Again I have one or two ideas on how to refresh the genre a bit. But I'm skittish about the 4th X because of AI.

4. Still another would be a simple tycoon game. It could just be a 2-hour game, not particularly replayable, of making early investment decisions and growing a business, like a sports club or a vaccine maker. I say not replayable because I find if I win such a game, I don't feel like replaying, and if I lose, I get annoyed and quit, lol. This strikes me as relatively straightforward to implement, but maybe less interesting to play than some of my other ideas.

5. Yet another would be a simpler variant of chess, or a more complex variant of checkers. The point would be to get more experience manipulating a 2D array or grid, selecting pieces, and learning something about AI. Or just chess itself, a game I have always loved. I sort of suck at it, so it would be amusing to see whether I could build an AI that would beat me. But if it would take months and months, I'm not sure I want to go there.

I'd appreciate any suggestions on which to try first! Thanks in advance.

 

[flyinian]

Hi all,

I've finished my first game in GML, a clone of Tetris. (I posted a screenshot in another thread earlier.) Since then I've been using pen and paper to design prototypes of a possible turn-based game of my own design. Mostly I'm interested in building my skills, but it would also be nice to make something at least slightly original. (I have a lot of experience as an artist, but I'm relatively new to the programming thing, though I was a comp-sci major for one semester in college decades ago. Back in the Jurassic era, when we used APL and Pascal and FORTRAN.) For most of these ideas, I have a theme in mind. I'd like to keep the project small and feasible to finish. It took me about a week to make Tetris. I'm willing to spend several weeks on the next project, but not months. Here are some candidates.

1. One idea is a small turn-based square-grid-based tactical wargame. Like a small, simpler, prototype version of Advanced Squad Leader or X-Com. My admittedly unoriginal theme would be WW2 platoon- or company-size combat. But any territorial-control theme might suffice: lions controlling the savannah; birds controlling the lawns with the best worms. I'm actually more interested in operational and strategic wargames, hex-based, with lots of juicy resource and building decisions -- but I know enough about the complexity of such games (not to mention hexes) to know I shouldn't start with those. From a programming standpoint, pathfinding (A* stuff) would be great to learn. I've watched some of sgt indie's videos on this, and read some theoretical articles on the subject.

2. Another is a deckbuilding cardgame, again on a small scale, in the same genre as Slay the Spire or Monster Train or the boardgame Dominion. (Possible themes include building a business, or running a hospital, or building a prosecutor's case.) Each hand would take just a few minutes, but would be grafted onto a larger metagame or roguelike structure. I have a few ideas on how to inject a little innovation into the genre. Maybe this is more manageable than option 1 because no grid or AI? Also it could start as just the deckbuilding loop and expand to the meta game if it were going well, or not.

3. Yet another would be a tiny 4X game (or 3X, with no eXterminate). Space setting, or perhaps a Wild West setting, playable in an evening rather than over 40 hours. Again I have one or two ideas on how to refresh the genre a bit. But I'm skittish about the 4th X because of AI.

4. Still another would be a simple tycoon game. It could just be a 2-hour game, not particularly replayable, of making early investment decisions and growing a business, like a sports club or a vaccine maker. I say not replayable because I find if I win such a game, I don't feel like replaying, and if I lose, I get annoyed and quit, lol. This strikes me as relatively straightforward to implement, but maybe less interesting to play than some of my other ideas.

5. Yet another would be a simpler variant of chess, or a more complex variant of checkers. The point would be to get more experience manipulating a 2D array or grid, selecting pieces, and learning something about AI. Or just chess itself, a game I have always loved. I sort of suck at it, so it would be amusing to see whether I could build an AI that would beat me. But if it would take months and months, I'm not sure I want to go there.

I'd appreciate any suggestions on which to try first! Thanks in advance.

I'd say go with option 4. Seems to be easier than the others.

 

[Selek]

Thanks for your reply. That's a good point. Anyone else care to weigh in?

 

[Rob]

for 1) - Depending on what you want to do, you'll need to be dealing with stats, so you'll probably need enumerators, arrays or data structures. You might also need A* pathing ( I see you already mentioned that and sgt_indie is how I learned to do it too! (with some additional reading of A* blogs) ), although mp_potential_step can sometimes work OK. If you want animations to play in sequence, you could add scripts to a ds_priority_queue and use script_execute to play them in order, like Friendly_cosmonaut does in her Farming series, or how I'm doing it in my Isometric Tactics battle series.

2) I've never made a deckbuilding game but I'm sure you'd want to have plenty of stats, which means arrays/enumerators/data structures etc. Structs too if you're using 2.3.

3) What about generating the map? You'll need to put a bit of work in if it's at least slightly random. You'll need a state system to allow different players to take their turns, even without AI.

4) Could be as simple or as complex as you want - but that's the same with any of these games I guess

5) I think this might be the easiest one to finish within the time frame you mentioned, particularly because there aren't that many features to implement when compared to the other games. Rather than making a complex AI, maybe they could have a simple ruleset like "always take a piece if you can" "always go for the check/checkmate" but even those 2 rules would mean the ability to make lots of different checks before taking a move.

A lot of games seem simple to players but have a lot more systems going on behind the scenes and can take quite a bit of work to put together - and the more systems there are, the more work you need to do / more thinking/design you need to do to make sure that everything's going to work together and you don't need to rewrite every other system every time you implement a new feature.

 

[Selek]

Rob, thank you for those thought-provoking comments. Heh, I hadn't given much thought to generating the map for a 3X/4X. A daunting proposition, to be sure. I agree that my fourth option could be relatively simple, but I'm not sure about it, perhaps because it requires a pretty vivid theme. I nonetheless did spend some time today with pen-and-paper sketching out decision trees for one possible tycoon game. I'm actually rather excited about fussing with chess for a bit, so I'm glad you think that's a worthwhile option.

I'd still welcome any other viewpoints about this. For me, this is always a difficult stage of any form of art.

 

[Yal]

One idea for 1): make a Fire Emblem or Age of Empires-style game about ants (maybe represented as humans with insect-like armor if you want higher marketability). Ants constantly wage wars with rival colonies, have drastically different body builds depending on the caste they were born into, herd cattle bugs, can lift stuff several times their own weight, and exist all over the planet except on the poles. There's plenty of fun stuff you could do with this. And have you ever taken a close look at a natural field? Flowers and weeds look weird close-up, so if you based macroflora (trees etc) on how common small plants looks, you'd have a really unique environmental art style.

 

[Selek]

Ants sound fun! I like that idea. "Bugs Life" (the movie) meets "Age of Empires." Yeah, giant flowers and mushrooms would be a cool environment. Thanks for the suggestion!

 

[Psycho_666]

Hey, how's it going?
Uhm... What is the goal of this exercise?

If the goal is just to learn more things no other project than the 4X would teach you that much information. You basically have everything in it. No wonder turn based strategy games are labeled as one of the most difficult to create. Path finding and AI and everything is in there.

The tactics game is pretty much the 4X without the base building, exploration and so on. You just have battles, bit you still have your AI and pathfinding. What you skip on is the hard progression system you must have when building stuff and training units only when certain buildings are constructed.

If the goal is to expand your knowledge and yet create something out of it, I would say go for the card game. At the end of the day, card game you can make with huge database for every card and programmed interactions between them. From that point forward it's entirety up to your imagination what situation you gonna throw the player in, will there be other things on the map other than the player, enemy and cards? Like wildlife you can kill to charge up your mana for a powerful spell or something, it's entirety up to you. You still have to dabble in AI, cause you can't just let the game play random cards with no logic.

I would say stay away from the Tycoon management games and chess... Both of them are absolute black holes. In chess every play have counterplay and you have to program that, and them you have to program forward thinking, meaning the AI needs to have a goal and follow it and adapt to what the player does, it's gonna be a nightmare to program, especially when you said you are not really good at chess. The Tycoon is pretty much the same. It's a management game. The options make or break a management game. You need a lot of extras and stuff. If you choose this option this is the effect, bit if you choose that option that will counter the first option you chose and this other effect will trigger...

I would say go for the card game. You can have a nice decent card game with something like 50 cards as long as you make the interaction between them interesting. And a card is just an entry in a database, so 50 is not really that huge of a number.

Look at a free game on Steam and mobile named "The Night of the Full Moon". It's extremely simple, all it have is cards with different effects, and damn, it's super cool how they built it up, so you stack effects and stuff.

 

[Selek]

Hey, @Psycho_666, thanks for that awesome post and that great advice. I think you identified my goals: to learn from a project, but also to create something out of it. I mean, it would be cool to make an original game and share it with friends, play it myself, or even sell it online. Your post has helped talked me out of the tycoon game. (In fact, I sat for a couple hours trying to make a pen-and-paper prototype and found I wasn't getting anywhere. It's hard to make such a thing interesting without lots and lots of content, it seems.)

You've got me excited about the card game! Also, that genre plays to one of my strengths -- making my own art. Plus, I just like card games. They make you think but don't take too long.

As it happens, I have spent a couple days working on a chess game, and I'm pleasantly surprised how far I've gotten. I enforce most (not all) the rules for both me and my rudimentary computer opponent, which at this point is mostly random. I'm using arrays, not objects/instances, which has been a good learning experience. But, as you say, it's a black hole. I think I'll dip my feet in a little deeper to see if I can learn something about minmax AI, then move on to one of these other projects. Quite possibly the card game. I will think on it more. Thanks again!

 

[Selek]

An update on my doings: I ended up getting rather obsessed with my little chess game, as it's fun to make. I've got the AI searching one move ahead now, doing the minimax thing. It's amazing how much it improved just by looking one move ahead rather than simply capturing every piece it could, lol. I've also given it some scripted openings, for both black and white. I'm going to increase the search depth, add a few more openings, and maybe give it some basic endgame play. Then I'll move on to one of the other projects I've discussed here.

 

[Selek]

Look at a free game on Steam and mobile named "The Night of the Full Moon". It's extremely simple, all it have is cards with different effects, and damn, it's super cool how they built it up, so you stack effects and stuff.

Thanks for the tip. I played through the first boss and really enjoyed it. A bit easy, but I was on the newb difficulty level. Yep, that's just the sort of card game I had in mind. Lots of fun. I've always enjoyed that deckbuilding genre. "Monster Train", the boardgame "Dominion," and "Slay the Spire" are some of my favorite card games.

Anyway, I'll play the game more tomorrow. Good fun!

 

[Selek]

Update: So I've followed the advice of @Psycho_666 and spent a week designing and prototyping a strategy game that has deck-building as its core mechanic. I've built the core game loop, which makes it very easy to whip up new cards and test how they work.

I have some ideas for a new take on the meta-game, but I'm concerned that my core loop is unoriginal: draw cards, play a subset of your hand, discard the rest, enemy attacks, rinse and repeat, then pick a shiny new card (from a menu of 3) after you win the battle. Of course, the meta-game can add originality: Slay the Spire used a roguelike meta, while Monster Train has elements of tower defense. Innovating is, for me, harder than any code- or art-related task.

 

[Psycho_666]

discard the rest,

Why? Keep them for the next round.
Or if your balance is based on that, you can use discarding as a mechanic. You may want to add a card that allows you to not discard your hand or keeps one of your cards for the next turn or something like that.

A card game is as boring or as exciting as you make it.
I don't know what your ideas are, what the setup is and so on, but you can do so much with duel style of card game.
Resource cards, equipment and item cards, offensive cards, defensive cards, buff cards, debuff cards...
The fun part of the card games is that literally one card can change the game because of the way it interacts with the other cards.
Card games are one of those genres where the content make the game. The wore content you add, cards in this case, the more interesting the game will be.

 

[Selek]

Why? Keep them for the next round.
Or if your balance is based on that, you can use discarding as a mechanic. You may want to add a card that allows you to not discard your hand or keeps one of your cards for the next turn or something like that.

Yeah, I had in mind later adding a card or ability that would allow you to hold one or more cards rather than discard them, for a price. But I hadn't thought of allowing the player to keep one or all cards from the get-go. Allowing her to keep just one card might make for one more interesting decision -- which to hold, which to discard. Food for thought!

And yes, I already have cards with buffs, debuffs, attacks, evades, etc. Right now I have the protagonist squaring off against one or more enemies, sort of like Slay the Spire, but with some significant differences in mechanics and hand configuration. I'm still debating whether to add a tableau in the style of Magic the Gathering or Race for the Galaxy. That makes each match longer, slower, but also arguably more interesting. But it also makes for a more complicated and crowded UI. Monster Train does something along these lines, but with more of a tower-defense vibe, whereas I am going more for an assault-the-enemy-fortress vibe.

Even in my simpler form, I'm finding that nonstop testing is essential. It's amazing how hard it is to make each battle reasonably close and tense. This means I'm constantly tinkering with card values, which is easy enough to code -- but my sprites aren't keeping up, as I'm lazy about updating them. That's one project for today -- edit my card sprites so they reflect the current state of the code. Or have the code display essential info about each card, for now.