Midnight Fields - Developer Review

<- Back to blog | Posted on 19 February 2017 by Adam Summerton

Midnight Fields was our entry into ‘Finally Finish Something’, a game jam that we also hosted.  This is the post mortem of the game, rather than of the jam itself, but you can read about that here. For this post mortem, we are trying something a bit different… Usually we put together a rather formal document in the traditional post mortem style, but we found that none of us really enjoyed doing this or referred back to it much.  We decided that from now on we are going to do these as an informal live chat and record the results here on the blog.  So below is the result of our first attempt, taking part are:

Edge: Made the game and wrote the music.  Also ran the jam and did the website & Twitter accounts
James: Created the main character and enemy designs. Did all the animation. Designed the levels
Adam D: Made all the tiles and scenery. Designed the HUD and UI

Edge: Hello and welcome to the Post-Mortem/Review thingy for Midnight fields!  So are we all ready? Do we know who we are and why we're here?

James: Those are hard questions

Adam D: I'm ready and I know who I am, I think.

Edge: So the idea of this is that we sort of interview ourselves.  We all have to be Parkinson

Adam D: Okay cool, do you want to start us off?

Edge: Yes, I have a question to begin… Did you have an idea of what MF was going to be before we started it, and did it turn out like you expected?

Adam D: Yeah I had an idea of how I thought it might look, obviously the core mechanics were laid out before we started the jam as you'd done multiple prototypes, so I guess I was influenced by that. I had images of Halloween and retro shooters, it certainly captured a lot of that. It wasn't quite as I'd imagined it because there is input from you and from James of course, so there are always surprises but that's a good thing.

James: I did have a fairly clear idea of what the game would look like. I guess it helped having worked on it before and knowing what needed to be improved. Generally I think it turned out like I imagined, I always feel like I could do better with the animations but I feel like that with every game.

Edge: I guess I have done about 10 versions of this already, so it's different for me.  I feel like it ended up looking kind of 'too good' if that's a thing.  I imagined it very lo-fi retro, and it became quite pretty.  I don't think that's bad, just a long way from what I pictured

James: When you say Lo-Fi, what do you mean?

Edge: I guess more like a Jeff Minter game?  I thought it would be super low detail, like original Space Invaders or Jet Set Willy

Adam D: Interesting, yeah the original prototypes had quite a different look. Projects always seem to work out differently than we imagine.

Edge: Are you Googling Jeff Minter games?

James: Yes

Edge: You are right to do so

Adam D: It's cruder pixel art (I mean that respectfully). I remember Jeff minter games and know what you mean. That crazy space giraffe stuff and things to do with llamas.  I remember there being a lot of llamas.

James: That style is just a long way from what I want to do, it always takes me a while to adjust to the pixel style. I think it's just in my nature to want to add detail. If I'm honest the lo-fi style is not something I want to head towards.

Adam D: It isn't a style that I naturally gravitate towards either, I'm okay with it but I don't really know how to make artwork like that and make it look like it all fits together coherently.

Edge: It's not a style either of you have ever really employed, so it makes sense that we didn't go in that direction. Maybe it's a style I'll save if I ever do another game on my own without artist support

Adam D: Yeah it is hard to work in a style that we aren't used to and doesn't fit with how we work.

Edge: It's probably best to stick to what you know. I suppose I feel like the gameplay is fairly crude and we could have gotten away with it more with lo-fi graphics. I think my feeling after playing the prettier finished thing is that I want the gameplay to promote exploring the level more, and for the levels to be more dense (which would have been bad for shooting things).

Anyway were you guys happy with the graphics overall then?  I thought the enemy designs were really good

Adam D: I really like the enemy designs and the animations were very impressive. Did you just do them all in gimp James?

James: I was happy with the way it looked, I think if I had more time I would have liked to do more creative weapon animations. I did all animations using Gimp to make the sprite sheets and didn't really use Spriter.

Edge: I think my favourite was the demon that spat littler demons

Adam D: Cool, I especially like the floaty octopus thing.

Edge: Was it easier without Spriter?  I guess because the resolutions were quite low?

James: Yep, sprinter is good for higher resolution stuff but I don't find it helps with the pixel style. I much prefer doing it by eye.

Adam D: That's good too, makes me think of Diablo a bit. What's your working process for doing animations in gimp James? Do you test them at all, or do you draw them out and then see how it works in game?

Edge: Do you test them in Construct or wait until it's in game?

James: I've started making them as a gif at first just because it's easier to view as an animation. I go through a lot of variations before I'm happy. Once it seems to work as an animation I convert it into a sprite sheet. I only test the sprite sheet in Construct to see if I haven't cocked up the spacing on the sprite sheet.

Edge: That ended up working really well

James: A lot of the time I can be happy with something until I see it in game and then it looks wrong, Like the tree enemy, I think I did 4 designs of that before I was happy.

Adam D: Nice, I'm still a bit intimidated doing animations by eye in gimp, I've only done it for really simple stuff. I'm used to Flash and having layers and symbols and all those features.

James: I guess it's just practice, I think I could do it for a shorter jam, maybe I would need to do less animations.

Adam D: Speaking of Ludum Dares, this is the second month-ish long jam we've done. How do you guys feel that they compare to doing 48-72 hour jams?

Edge: Bizarrely, I felt like it followed the exact same pattern, just drawn out over a month.  I was still cramming like crazy at the end

I think overall it produces a better game, but I was thinking towards the end of the month "how in the hell do we do this in one weekend?"

James: I do prefer them but I agree, we do seem to cram everything in at the end. I think we still find time management a bit of a problem

Adam D: It is odd, it's like the more time you have, the more time everything seems to take. I do feel that MF seems noticeably more polished than say Proxima B.

Edge: But does it feel 15x better?  Because that's how much extra time we had

Adam D: I'm not sure I can quantify that but the you're right, I felt like that on Not Alone too, everything was just crammed at the end. All the progress I'd made up until that point didn't seem to make a difference.

James: We seem to always rush the same things like level design, tutorials and testing, no matter the length of the jam.

Edge: I supposed with moving house, life, work, babies etc, we probably didn't spend that much more actual time on it

Adam D: Yeah life definitely gets in the way.  You can shut everything out for a weekend just about but not for a month.

Edge: I think my failing is that I concentrate on things as a list of tasks, almost in isolation.  I have a test level and make the entirety of the game mechanics in that, and think that we'll make more levels once everything works in this one

Adam D: How do you think you could work better?

Edge: I don't know - I feel that isn't the best approach, but not really sure how to change it

James: Maybe with the longer jams saying we need to complete stages of the games at certain times so we have enough time to test at the end, but like you said, life has a habit of getting in the way.

Adam D: Levels are definitely time consuming things that we always leave to the last minute, but without having the mechanics in place it is difficult to design them.

James: I think getting one level working as a game is a good idea, we just need to do it sooner so we can make more

Edge: The other thing is that I didn't spend much time just tweaking moving, firing & animations until they felt right together.  I did the 3 things separately.  Spent a lot of time on weapons.

Do you think we put too much in?  I mean 16 different weapons?  That was too many

Adam D: We are always ambitious, possibly over ambitious. It is just what we do.

James: I feel we always put too much in every game

Adam D: I think that Half-Life approach to game design is good. Get one level working and make it fun, really fun. Then go from there.

Edge: I guess I agree.  But I rarely decide what the core of the game is, I don't think the core idea of MF was a fun-feeling shooter.  I spent all the time making the weapons work

James: So do you think we should spend more time making the core idea/mechanic more fun and  work better, before we move onto more complicated stuff?

Edge: Yeah, if I had thought about it and done that, I would have made it fun to shoot everything with a basic gun and concentrated on that.  Once that was fun, then add more variety.  That's what I mean when I say I get caught up with a list of features in isolation

Adam D: Yeah maybe working more on the core mechanic/feel of the game is important. I always marvel at really nicely designed minimalist puzzle games, I have no idea how people come up with that stuff. But they can be great examples of polished core mechanics.

James: I guess it's harder to make a well made, simple game that has a great feel and core mechanic.

Edge: I was better at that with Morphling and I think it shows - the main character felt good, and everything else was tacked on at the end (enemy behaviour, levels etc). I need to be slapped and reminded of this

(metaphorically slapped)

Adam D: Yeah it's easier to add in lots of extras to build up a game, but if the core isn't there then it won't work. Morphling and Rift Knight work because they were very focused on feeling nice to play and control.

James: Maybe we rush ahead too fast before we get the core done. We have so many ideas sometimes the time limit scares us and we don't concentrate on the basics.

Edge: That is true I think James.  Sometimes the idea can just be "shooting things will feel cool" (if the game has shooting obviously)

I have to admit this as well... I think I am actually pretty good at platformers, but I am super average at top-down shooters (The movement in MF is not good)

James: You are very good at platformers, they are always our better feeling games in terms of gameplay.

Adam D: So I was going to ask how you guys think we can apply that idea to our game design process? For different types of games, like city builders, platformers and so on...

Edge: Difficult question that, I mean with a city builder it seems like the UI is a massive part of it, whereas in a platformer it is secondary.  I guess agree on 1 or 2 basic interactions up front, and then make them good

Adam D: Maybe starting with the core idea that defines the mechanics is good. "Shooting things will feel cool", "building a city will be intuitive and immediate, colourful and lively" and then starting to work out the interactions and most important parts from there.

I think From the Sea works because we focused on the core. We said it's Miyu but without all of the other stuff getting in the way of exploration.

Edge: That is a good way to go.  I know with Toy Box Metropolis I immediately started work on the different types of building, before even thinking about the interface.  Everything was dependant on the type.  It was an insane decision, but then I went too far to scrap it and start again

Adam D: Yeah there isn't much time to back track in game jams.

James: That's something we learn with practice I guess, it's easy to stick with platformers because we are good at them. It's only when we break our comfort zones that we will learn these things.

Edge: To guide us back on track a bit here, what did you guys think of the enemy AI in MF? They all had different behaviours and I wondered if that came across?

Adam D: Yeah we wandered off a bit there. I thought the AI was nice, I spent a lot of time blasting so didn't always notice all of the variations. The shape-shifting trees were cool, and having shooting enemies as well as ones that charged you.

Edge: The trees are the best gameplay-wise, they make you change what you're doing when they surprise you

James: I noticed some more than others. I think more extreme behaviours may have helped or more variation of attacks. It felt that some just chased you at different speeds sometimes.

Edge: Shit, you found me out James.  Because that is exactly what they did!

James: I knew it (safety wink)

Adam D: Busted. Yeah perhaps fewer enemies with more distinct behaviour would be good, you've got to be careful of performance though, complex AI can eat it up.

Edge: Honestly if I could redo it then I would stop a lot of the constant chasing, it was a bit much. Only the pink ones should really always chase you.  I suppose I was thinking it was supposed to be fast-paced, so you shouldn't get any peace

James: Maybe some that teleported in front of you, or dug underneath, just something different from chasing and shooting.

Edge: It doesn't feel very nuanced to play really, I think I am a bit disappointed gameplay wise.

Adam D: Binding of Isaac has some nice examples of varied enemy behaviour.

Edge: On the plus side - all the AI does exactly what it should with no bugs.  First time that's ever happened.  Thank you finite state machine

Adam D: Nice, I had horrible trouble with the AI on Not Alone, so I'm glad it all works in MF.  The finite state machine plugins are very handy.

Edge: That is one thing I'm getting better at, that and picking object instances.  Plus the new class system that I stole from you was amazing - Changing anything was super easy

Adam D: Glad you liked it! I did find the picking in Construct 2 confusing at first, but that's probably because I refused to read the manual.

Edge: There are a few things that you can only find out by stumbling on obscure forum posts though

Adam D: Yeah it's always the way, there is always something left out or that has been discovered by someone experimenting.

Edge: So shifting focus again here - as the weapons ended up being the main deal, do you think it worked well?  The shooting is at least OK?  Some of the weapons do feel quite fun

Adam D: The shooting is fun, the variation in weapons is fun. I think the shooting feels much better than the movement.

Edge: Yep, the movement is balls.  Guilty as charged.  It was on my list to sort out, but never got to it

James: I think the weapons work and the shooting feels fun. Maybe they have the same problem that enemy behaviour has, they just need more interesting variations.

Adam D: Some of the weapons are not useful outside of very specific circumstances like the duplicating bullets. I like a lot of them though.

Edge: Again, too many though right?  4 super interesting and polished weapons would have been better

Adam D: Yeah too many I think. I tried Proxima B as well to compare against a LD shooter that we made, and the movement in MF does feel less sluggish than that. Fewer extras, more focus on the core.

James: We didn't have time to play test properly so we couldn't really test them.  In hindsight fewer weapons would have been better

Edge: I think we could take the essence of what James said there and just re-use it after every jam: "In hindsight, fewer X would have been better"

James: At least we decided not to have the special power ups as well, that would have been crazy.

Edge: Man the powerups, most of those mechanics are still in there just commented out!

James: So we agree, fewer things and make the ones we have feel great.

Edge: Yes

Adam D: Yes, let's see if we can stick to that...

Edge: To expand on that same point: I spent a lot of time on 3 things that probably weren't needed
* Procedural Poems
* Increasing Difficulty
* Random Level Order

The feature list of this game is actually ridiculous for a game jam

James: What's this called, feature creep?

Edge: Yep, that's me

James: I think increasing difficulty isn't a waste.

Adam D: Hmm it's hard because all of those things add to the feel of the game. I like the procedural nature of the game, and the poems add to that, and the difficulty increase is important. However, yes things like procedural poems are less important than the core gameplay.  I definitely think procedural poetry falls under the category of feature creep. It was a nice touch though.

Edge: Interestingly the difficulty doesn't reset at the end of the game, it just keeps getting harder forever.  The less said about that the better.

Anyway, I am very happy with the music, wrote it a while back but one of my favourite tracks

Adam D: I like the music too, fits nicely.

James: Yes, music was excellent.

Edge: Sound effects not the best, but that was just a time thing

Adam D: I think they work fine, you just haven't had the chance to do much with them and to do some more creative stuff.

Edge: Only thing we haven't talked about is the level design...

James: It's the first time I've done it really.

Edge: It was so helpful having you help with the levels James, game literally would not have been completed without that

James: I was happy with them but I just ran out of time and I agree, they could have been more creative.

Adam D: The levels look good to me and were interesting to explore, I don't think I had any issues with them at all. As you guys have mentioned before though it'd been nice to have a tutorial island/level.

Edge: It was hard because you have to make them look nice but also work when you're trying to shoot a bunch of enemies

James: One thing I realised you can lay out a nice visual design and then when you play it you find it's rubbish.

Adam D: It definitely needs to be 'form follows function'. So making them look pretty needs to be secondary to making them work as fun levels.

Edge: It's definitely a skill all by itself, big developers have people just dedicated to it.  I think we're slowly getting better though

Adam D: I think it is more how we approach it. We're always going to be limited in jams, and have to wait for mechanics to be implemented.

James: I did some great symmetrical designs in a nice pattern and it just didn't work

Edge: So for the 17th game jam in a row - I conclude we need to spend more time on level design

James: I enjoyed doing some level design though and would like to do it again

Edge: That's good to hear

James: The system you guys set up with auto tiling really helps.

Adam D: Auto tiling is definitely great, all Edge's doing.

Edge: James can you imagine Morphling & Proxima where we had to do every tile position manually?  It was not fun

James: I've just broken out in a cold sweat thinking about it

Adam D: Oh it was painful. So painful. Maybe James' idea of making one main level and making it fun, then building out would be good. It helped on Not Alone to keep it focused to one area.

Edge: I have some ideas on that front for a platformer, but again I'm just a bit rubbish at top-down games.  Next time we do one just need to make the core mechanic simple, which might help with designing levels.  Plan to have fewer levels too.

Adam D: Yeah, it's not to say we wouldn't do more levels in a jam but maybe the approach of focusing on the core mechanic, and implementing one level (at first, more if time allows) which is fun with the core mechanic, would be worth a try? We've not really tried that much before and it might be worth an experiment. It's the sort of thing we did for From the Sea and a bit in Not Alone I suppose.

Edge: Yes that's true, those games were only ever designed to have one level

Adam D: You said to me that for Rift Knight you just wrote down "joyous movement" or something like that, I thought that was a good idea. Have the core defined, and anything that doesn't support that gets binned or pushed down the priority list.

Edge: Yes that's what I did.  I literally completely restarted Rift Knight 4 times until the movement/attacks worked and then threw in enemies and levels at the end as an afterthought

Adam D: I think you can really tell when a game is focused and knows what it is trying to be.  Games like Morphling and Beneath the Sea work well for me because they are quite streamlined.

James: So for the next jam we need to make the core mechanic work and be satisfying, keep weapons simple but enjoyable, make one level that works and is fun. Then leave time for more levels and testing.  Any extra time we have we add onto those core elements.

Edge: That's a good aim.  I was just about to say that we've covered all the main game elements from MF, so what was our conclusion on the game and our process

Adam D: Excellent, yeah and it was a nice summary. Basically we need to keep things from getting too sprawling before we've got the core sorted.

Edge: For me, it's basically what you've both said - focus on getting a small amount of the core game feeling good.  Which involves knowing what the core mechanics/elements are early on

James: We need to define the game more in the planning stage maybe? Set out what the core elements are early on.

Adam D: Yep definitely, almost a mantra for the game to keep us on target.

Edge: For MF, I like a lot about it, but it's not one of my favourites.  It was too muddled in terms of what it was supposed to be, not really a clear coherent vision and that comes across.  Still looks and sounds nice though, which is where we're pretty good when we get it right

James: I think we've all come to the same conclusion there.

Edge: Well that's a relief

Adam D: We have, we just need to put it into practice now.  This was a good little experiment, I think this chat format has worked well

Edge: Yeah this was actually fun, and nice to just say this stuff as it comes into our heads

Adam D: I've enjoyed it a lot more than writing other post-mortems.

Edge: So our conclusion that we all agree on, is that all our future games should have procedural poems

Adam D: Procedural poems are given the highest priority from now on.


James: And we need more walk cycles

Edge: Each letter of the poem needs a walkcycle

Adam D: Procedural poems about walk cycles that are all animated as walk cycles are given the highest priority from now on.

Edge: The final word. Post-mortem out (drops mic)

Add new comment