next »

[deab]

Tacticrawl - Turn Based Dungeon Crawler
Explore a world full of interactive elements, discover the local inhabitants - then destroy them in turn based tactical combat.








Tacticrawl

I'm working on a tactical turn based game. An interesting, deep combat model is my initial priority, where using the world around you is the key to success.

Story / Background
After decades of learning and study, the Mage Lords created a powerful magic that would remove those consumed by negative emotion. Greed, anger, jealousy, hatred - all were targeted.

When the magic was released, anywhere these emotions outweighed all other was ripped from the world and banished to 'The Ever'  - an alternative dimension inaccessible to all but the most powerful magic.

Unfortunately the magic worked a little too well - whole villages, castles, caves - along with all their occupants (intended targets or not) - were trapped, each imprisoned in a separate bubbles within The Ever.


Art Style
As a developer by trade my art skills are limited, so I'm learning as I go. Aiming for a low poly, mostly flat shaded look. I like the readability of this style, plus creating assets should be quicker than a detailed, fully textured style.


About to receive a warm welcome from an Orc:



Game play / Design

Grid based exploration of either procedural or hand build maps (to be decided - maybe a mix of both). Use the environment to your advantage.

I plan on supporting multiple play styles (ranged, melee, stealth, pet/minion based).

Depending on how you play, you could be fighting many enemies at once, or carefully pulling out individuals to tackle. Some enemies will also seek help, so managing your foes will be a large part of the game-play.

I'm focusing on a single main player character, but I'm also including support for multiple player controlled characters, so summoned minions and controllable allies will be possible. Maybe a rescued character would offer to join you.

Planning to include a number of different biomes - both indoor and out.


Combat Mechanics
To allow for multiple play styles and combat options, the following features are planned:

• Field of view with ‘memory’ of previously visited areas (same for AI characters).
• Directional combat - positioning and facing direction impact abilities.
• Tile based environment which can be altered.
• Light, sounds and heat will all have an impact in the world.
• Multiple AI factions with different allegiances - not everyone / everything is out to kill you.
• Interaction with items found - light/extinguish torches, open/close doors, chests etc, pick up/push/throw objects.

Original Post - 26/2/15
8/4/15 - I have refocused this project, dropping the hexagons and moved to a new low poly art style. Playstyle is now based on the 3D world with much more detailed world (lots of interactive objects, not just floor and wall tiles.) The original idea of a playable plan view has been dropped. Here's the original post and old gifs:

Tile based map with orthographic and plan views are done, and both can be zoomed / rotated. In ortho view, walls are clipped if obscuring a pathable tile behind.

Field of view and memory/fog is in. If an enemy leaves your current view, last known position is shown.

Here’s a couple of gifs showing both views (plan view now removed). Grey tiles are out of your current field of view, and show last viewed state.
(Contains programmer art, and the blue guy is just a reused model)



The ortho view shows wall tiles capped if they are blocking your view.

[deab]

Update #1

Recently I've been working on an overlay for using abilities. On a player turn, valid tiles are highlighted either for movement (shown in white) or the selected ability (currently have abilities on keys 1-0).

A light highlight is shown for the available range, and a saturated highlight where all restrictions on the ability are met. In the example below I cycle through nothing selected (white), a melee based attack (red), a range attack (red), and finally a healing ability (blue).

The attacks and heal all require a valid target within range. Rats are of course evil so can be attacked, and the other blue chap is a valid heal target.

The overlays are simple meshes and I'm using vertex colour, so only require 1 draw call.

[erebusman]

I really love turned based combat games - there's a really rich history (X-COM, Heroes of Might and Magic,OGRE, many others).

I'll be watching to see how your game progresses!

[deab]

Update #2
Better missiles; reactions; and positional combat!

Good progress this weekend.

Missile abilities (that would cover firing a bow, throwing items and bolt type spells) now check for a hit every step of their flight path, and will even extend past their target if they fail to hit.

I say 'check for a hit', as characters now have a chance to react to an ability (currently the only option is to DODGE, but later BLOCK will be added for relevant characters/equipment). Chance to react is based on the side of the character facing the attack - highest chance for facing the attack directly, a lower chance for the two front sides of the hexagon, and no chance for the back 3 sides.

So that's the first step towards making a characters facing direction relevant. Potentially characters could also have different armour values for each side of their six sides, allowing for weak spots.

I've also extended the overlay to show the tiles a missile ability will travel through when used, handy to spot potential friendly fire issues!

[deab]

Update #3
Unity 5

Small update, I've upgraded to Unity 5 and found two issues that needed a little digging.

Firstly, my shader fell over with the error: 'Too many texture interpolators...'. Now my shader is very basic (my shader skills are minimal). I use vertex colours for all the lighting, and use the colour alpha value to switch the 'fog' mode on or off for each tile / character, avoiding an extra draw call (in fog mode, colour is desaturated).

I could get around the error using '#pragma target 3.0' but this seemed overkill for such a simple shader. I managed to fix the error but stripping the code down to a minimum  (essentially removing a float from the structure passed between vert and surf functions). No idea why this worked, but meant I could avoid adding the 'target' line.

Secondly my ability buttons were gone (these use the new Unity GUI). These are dynamically created in code. Problem was when they are set to a parent Transform, their scale was set to 0 - easily fixed by setting localScale back to 1 after setting the parent.

So not too bad! I'm eager to play with the new lighting options, but really must focus on getting the core combat in..

[erebusman]

Nice to hear your transition to Unity 5 wasn't too horribly painful , did you get a message they were converting your scripts?

I know when I upgraded mine in beta they had made a few changes to my scripts that when I later went back to do more writing I had to glance over it and go "oh -- that's how I should do it now??" but it wasn't too scary.  What was your experience like?

[deab]

Nice to hear your transition to Unity 5 wasn't too horribly painful , did you get a message they were converting your scripts?

I know when I upgraded mine in beta they had made a few changes to my scripts that when I later went back to do more writing I had to glance over it and go "oh -- that's how I should do it now??" but it wasn't too scary.  What was your experience like?

Yes scripts were converted, and I was curious to see what changed. Very little actually!

Component short-cuts were replaced with GetComponent<>() calls, and meshRenderer.castShadows is depreciated (I may have updated that manually), that was it according to the git log.

Overall very positive, and not having Pro before lots to play with. Profiler looks very useful.

[erebusman]

Its funny  I've had pro for a while .. I used the profiler a few times to just glance at my project but I can't ever find anything useful in it .. I guess I need to do some tutorials on it because right now I get no use at all  out of it

[Cranktrain]

Looks nice, I do like turn-based games.

I also had to rewrite all my shaders for my game, on moving to Unity 5. Bit of a pain, but the new material system is undeniably better. Also had a glance at the Profiler, looked interesting, but as my game runs at 60fps on even slightly older laptops, I'm not digging into it right now.

[deab]

DizzyDoo the Cat Machine looks brilliant, will be following.

I'm nowhere near requiring the profilier, but if I ever target a mobile version will be useful (PC is only concern right now). I try hard not to over optimise every little function.. but I do love efficiency..

[deab]

Update #4
Cleave ability.

Added a new melee based ability for swinging at surrounding enemies. This required a fair bit of work upgrading existing systems:

• Required a new 'arc' range based on a pivot position, facing direction and radius.
• Abilities could only hit a single target, easy enough to improve this, but also required a refactor of the animation callback system (to provide on screen feedback timed with the animation).

This work will be useful for other abilities, so very worthwhile. I have further plans for this ability, which I'll be working on next, and will add a new mechanic.

[deab]

Update #5
Knockback!

Today I've added a 'knockback' effect. As you might have guessed, when an enemy is hit with a knockback ability, and space allows, they are pushed back 1 tile.

In future there are a couple of fun ways to expand this:

• If the tile behind is blocked, extra damage will be dealt.
• When knocked against a sharp edged tile (maybe a spike wall) even more damage is applied.
• Critical hits could extend the knockback to two tiles.
• Push enemies in to bottomless pits, lava, or other fun environmental hazards.

I also added a quick 'hammer' effect!

(Game slowed down for gif)



You can also see in this gif another addition - all characters now have a bar depicting the direction they are facing. The colour indicates if they are friendly (green) or aggressive (red).

[erebusman]

Update #4
Cleave ability.

Added a new melee based ability for swinging at surrounding enemies. This required a fair bit of work upgrading existing systems:

• Required a new 'arc' range based on a pivot position, facing direction and radius.
• Abilities could only hit a single target, easy enough to improve this, but also required a refactor of the animation callback system (to provide on screen feedback timed with the animation).

This work will be useful for other abilities, so very worthwhile. I have further plans for this ability, which I'll be working on next, and will add a new mechanic.

I really like the cleave, something about the timing, the animation and watching the hit numbers roll is pleasing to the eye .. the kind of thing that makes the player feel like they are kicking ass. Good job!

[deab]

Update #6
2 sides removed, and art!

Been awhile since last post, but I've been making massive changes. But first:

New Art Style

I'm no artist, but my Blender skills are improving. After trying various styles, I've finally found something I'm happy with, and can produce without spending weeks per asset. Would love to get feedback on this look.. So far I've got 2 floor tile types (rock and grass), plus a couple of character models and props.

Here's our hero in a grassy area, hanging out with a tree stump prop.




Rocky cavern, with a couple of giant rats, a skull and torch props.




Game Design

After a lot of deliberating I've made 2 major changes.

Hexagons are out, squares are in.
There were just too many aspects of the hex grid I didn't like - moving east/west meant constant up/down movement, and wall layouts just didn't look great, especially with a low camera angle. Proper square rooms are also tricky.

I initially went with hexagons for the 6 equal directions, but with squares I'll treat diagonals as 1.5 times the cost. 8 directions for combat and travel, and a much cleaner look.

3D view only
The 2nd major change is I've decided to focus purely on the 3d view. Previously I had included a top down view for a zoomed out, more strategic mode of play. However, small details within tiles couldn't easily be seen, and this was really restricting some of the elements I want to include. I can still include a map view, but this won't show details, just for orientation.


Updates should be more frequent now that I'm over that bump!

[erebusman]

Hey,

I really like the new style its great!  Overall I'd give it two thumbs up and its a really great improvement.

The only item I would have any mixed feelings about is the main character; not because hes bad or anything -- hes also an improvement ; but just because the main character is SOOOOO important to your players relating to your game.  I think if you work on him a bit over time you can probably make a few more polishing touches that will bring him up a notch - so I would keep that in mind.

Otherwise great work I really think you've nailed a style that is attractive for your game.

[deab]

The only item I would have any mixed feelings about is the main character; not because hes bad or anything -- hes also an improvement ; but just because the main character is SOOOOO important to your players relating to your game.  I think if you work on him a bit over time you can probably make a few more polishing touches that will bring him up a notch - so I would keep that in mind.

Thanks! You are right there is room for improvement - and it's something I'll work on over time. Here's the progress in the last couple of weeks as I stumbled around for a style I could produce:



The first example was an attempt at the Warcraft hand-painted look which I've practised before, but it was very time consuming and I really lack the skills (only worked on the helmet edges and beard before moving on). In game the camera is going to be zoomed out a bit anyway so I don't need massive detail.

[deab]

Update #7
Let there be hills!

I've added height level to the world map. Adds a lot of interest, and will of course lead to game play mechanics later on.

[deab]

Update #8
Animation!

Lots of code clean up this week after recent additions. Variable tile height is working well, but still needs to be considered for pathfinding, line of sight and combat.

This week I tried my hand at animation, and created a floating eye to try it out:



First time I've really attempted animation, and pretty pleased with the results. Quite a learning curve - I've never used any of the animation features in Unity. I also created an animation for the 'nerve endings' at the back of the eye - these will animate constantly, while the eye will look around at random. Need to figure out blend states (I think) to run 2 animations at once, plus I need to explore controlling animations from code.

The floating eye will also be a test case for a range attacker - when an enemy gets too close, the eye should back off, so planning to get back the AI soon.

[Caravaggi0]

The mechanics of the game look interesting and as if they're on their way to being refined.

For the art, and this is just opinion obviously, you might try and concentrate more on whether or not you're using low poly counts to emulate a retro feel or create a more cartoony vibe.  If you're going retro then the high-resolution textures (such as in the world pieces) are clashing with the current poly count.  If you're just going for a cartoony feel then I don't think there's anything wrong with throwing more polys in.  Look at any movie with cartoonish characters, their bodies can be cylinders and cubes, spherical shapes, basic primitive shapes, but still have enough polys to hide the individual polygons.

The higher polygon count of the characters doesn't change that the head of the old man in "up" is basically a square and the head of the boy is basically a sphere.  Obviously you don't need those millions of polys, just enough to make it about the shape instead of the poly count.

http://pixar-animation.weebly.com/uploads/8/7/6/3/8763219/9575750_orig.jpg

[deab]

For the art, and this is just opinion obviously, you might try and concentrate more on whether or not you're using low poly counts to emulate a retro feel or create a more cartoony vibe.  If you're going retro then the high-resolution textures (such as in the world pieces) are clashing with the current poly count.  If you're just going for a cartoony feel then I don't think there's anything wrong with throwing more polys in.  Look at any movie with cartoonish characters, their bodies can be cylinders and cubes, spherical shapes, basic primitive shapes, but still have enough polys to hide the individual polygons.

Thank you Caravaggi0  for taking the time to comment, appreciate any and all feedback!

You are correct that I currently have two different styles - characters and props have very little texturing, while wall and floor tiles are flat with textured detail. To be honest the low poly choice is about ability (or lack of) and being realistic about the time asset creation takes.

I'm experimenting with reducing the texture detail. In this shot, grass tiles using less detail plus a new grass prop in the foreground. I keen to avoid pixel art style textures, so the actual resolution is still high.



I also built the wall section as a mesh, with minimal colouring:



In game, the directional light rotates with the camera, so the walls do pop more when moving around. Comments welcome! I'm going to stick with this for a little while and see how it works out.