Dead Man Walking trunk 5/22/2013

Download Dead Man Walking [trunk 5-22-2013]!

It's been almost a month since the last trunk release or, really, any news at all about Dead Man Walking. I've taken some time off from development and spent a while thinking about the design principles that make Dead Man Walking the game it is.

If you played the last trunk build, you'll know I had a variety of ideas to improve the longevity of the game, including the ability to upgrade equipment. The way that was implemented is not something I now consider satisfactory, so I've taken it out for the time being. If I can think of a better way to implement it into the game, I'll consider putting it back in.

Beyond that, I've removed bleeding from the game. I think it was one of the more confusing mechanics and there wasn't really a good way to explain it with the existing UI without just adding clutter. I've replaced bleeding with a new mechanic, Cursing. It is ultimately meant to serve the same purpose (high damage in a short time frame) while being less confusing overall. Being cursed will immediately remove a large chunk of your hit points and reduce your maximum hit points for a while. Enemies take a quarter of their maximum hit points as damage - that means there is a mechanical difference between how being cursed affects the player and how it affects enemies, but I think that's okay.

The next major change in the pipe is the ability to miss. I think a well-constructed to-hit system has the potential to add in "tactical" randomness and open up more interesting equipment options. Note that those equipment options aren't really present in this trunk build, but I'll be expanding on it in the very near future. The old to-hit system was based on Brogue's, but I've decided to move towards something a little more straightforward (in my mind): the player should have a 2-out-of-3 chance to hit, on average. The major exceptions here will be highly evasive monsters (there won't be many of them) and players that choose to go for bonus accuracy on equipment. Ideally players that choose to go for lots of bonus accuracy will become aware that they're doing so at a cost - sacrificing damage or block to get a more consistent hit ratio is a trade that should be interesting.

Beyond the addition of a to-hit system, I've also been working on critical strikes. I've been struggling with a way to introduce critical strikes that won't break the game. Monsters would need to either deal very little damage when not critting or I would just have to accept the fact that the player will take random, massive spikes of damage. I decided to do something a little different: critical hits get rolled just before the game checks if a hit will actually land or not. If a critical hit is successfully triggered, the hit automatically goes through, dealing bonus damage for players and just letting monsters ignore dodge for a moment. This is another example of a mechanical difference between players and enemies - some hardcore enthusiasts may see this as blasphemy, but I believe it's a justifiable solution when trying to create a balanced and fun environment.

Peace!

Game Hunter Looks at Dead Man Walking

Game Hunter has taken a look at Dead Man Walking!

Overall I'd say the overview was quite favorable. As regular readers will know, one of my major goals with Dead Man Walking was to allow players to jump into the game without a great deal of confusion. Watching Game Hunter take two inaugural runs of both the 7DRL edition and 1.1 has reaffirmed my belief that this has worked out reasonably well.

There are still a few things I'd like to work on and streamline. One of the most immediately useful things already in the next version is the ability to quickly swap out items between the inventory and equipment screens. Less menu juggling overall should be a win for everyone.

Peace!

20 Misses

If you make a game where the player has a 95% chance to hit a target, somebody is going to miss 20 times in a row. Someone else might miss 100 times in a row. Some other guy may never miss at all. This is the terrifying truth of trying to work with randomness. How can you balance around someone missing twenty times in a row? Dying in half as many turns is hardly unheard of in roguelikes. From my perspective, there are three realistic approaches a designer can take to solve this problem:

• Allow the player to mitigate randomness with time investment. This is fairly common in the major roguelikes - you start off with a relatively high miss rate, but it is reduced over time by increased attributes. It will often times end up at absolute zero. Does this solve the problem? Potentially, but players can still run into truly cruel RNG early in the game before the stat system really comes into play.

• Remove missing. This has been my personal go to solution in the past. Does this solve the problem? Yes, but it makes combat rather predictable throughout the entire game. No one will ever miss twenty times in a row in most of the current versions of my games, but it will always take three rounds to defeat the most basic enemy with little variation. This isn't an inherently bad thing, but it has the potential to make combat stale.

• Ignore it. The guy that missed twenty times in a row, the guy that missed one hundred times in a row, and the guy that never missed in a playthrough - they're all outliers. Perhaps there's no need to balance around them at all? This answer may seem like a copout, but if the rest of the game works fine, you can probably just ignore strange one-in-a-million chances.

This issue can apply to many concepts that crop up while discussing game balance in a roguelike. For example, if you make three items (x, y, z,) and give them an equal chance to show up, there are going to be playthroughs where a player doesn't find "x" or doesn't find "y" or doesn't find "z". A game doesn't necessarily need to have uniform structure to be an enjoyable experience though - it's well within the realm of possibility to make a fun game where missing is absent but random item distribution still exists (or vice versa).

As a closing thought, Dungeon Crawl Stone Soup is frequently lauded for containing tight mechanics and for constantly pursuing balance. It's also one of the most random roguelikes out there and yet, in my opinion, it works. Sure there's a one-in-a-million chance my assassin will never find a quick blade or I'll miss a dozen times against the gnoll, but these are issues I can work around without cheesing the system too much. Sure having the quick blade might make the game easier, but not getting it doesn't make the game impossible. And the gnoll...well, I could probably just run away.

Dead Man Walking 4/27/2013 trunk

New trunk build with a lot of new stuff! Some minor issues that need to be fixed, but it should play just fine.

Download it from the sidebar to the right or right here!

Balance in the new version is in flux. I'm trying to make damage over time effects more interesting. Poison functions as it used to, but lasts longer. Bleeds are now more akin to bleeding from Dark Souls - which is to say, it stacks up and then deals a massive amount of damage. I'm also playing with 5 flasks.

There's an abundance of weapon upgrade kits. I'm still working out how and when I want them to appear. I'm debating whether or not I want them to drop from bosses, but for now they do. I'm also playing with the idea of having the boss spawning the staircase on death, but that's not in this version. Also considering just outright removing most of the bosses and focusing more on interesting enemy types with a handful of random attributes, something familiar to those of you who have played Brogue.

1.2 is definitely shaping up to be the most lengthy version of the game and has certainly spent the most amount of time in design hell. I'm starting to pick up the pace a bit though, so it might be out sooner than one may expect. My primary focus right now is to increase the potential longevity in the game. If I just took 1.1 and made it three or four times longer, it would probably be fairly boring. If I take a similar game that has been adapted to include some "long term" goals, it may actually be an interesting game. Upgrades are the most obvious step in that direction.

Sort of an incoherent rant, but hopefully it gets the point(s) across. As always, feedback is very welcome. Anyone who has beaten 1.1 will probably find this particular version to be quite a bit easier - this isn't a long term design goal, just a result of the player gaining more power without enemies gaining any real strength.

Dead Man Walking 1.1 download & changelog

Windows download link for Dead Man Walking 1.1!

Spoilers below!

1. Several typos remedied.
2. New monster resource system.
3. Garjon's healing ability has been cut in half.
4. Yreek's Guardian now only takes four turns to die once it enrages, down from five.
5. Malevolent weapons now trigger on 1 in 6 attacks, up from 1 in 8.
6. Electric weapons now trigger on 1 in 3 attacks, down from 1 in 2. They now deal 3-6 damage, up from 2-3.
7. Reduced message log spam from damage over time effects. Poison and bleeds will now appear as part of your HUD (below the willpower bar). Enemies that are poisoned and/or bleeding now have a message appended to attack messages (if you attack them). I'm pleased with the UI update, but I'm still thinking of ways to improve the display for enemies.
8. Added in a new consumable item: the Magical Mystery Card. When used it causes a random effect. Not guaranteed to be helpful!
9. Instead of collecting healing flasks throughout a floor, you start each floor with four of them.
10. Slightly improved the AI.
11. Tweaked a few enemy stats around.
12. Adjusted the item drop rates a bit. Armor should appear slightly more frequently.
13. Yreek can no longer teleport you twice in one turn.
14. The Necromancer fight has been redesigned from the ground up.
15. Decay has gained a new ability and lost draining.
16. Standard weapon damage (from players and monsters) now rolls between -1/+1 of the actual damage number. 5 damage would deal between 4 and 6 damage. This is a substantial change from the old method, where 5 damage would deal between 1 and 5 damage.
17. Garjon's Overload ability has been redesigned.
18. Laughing Skulls can only shoot at you if you they have a direct cardinal on you (you're straight above/below/left/right of them).
19. Added in trinkets.
20. Removed the malevolent brand from weapons.
21. Added in alternative bosses.
22. Added in special floor types.
23. Restart button! But wait, that's not all: it works while dead! 
24. Added in artifact weapons.
25. Changed short swords to sabers. This is purely a name change.
26. Generally improved messages, including new hit messages for every weapon type.

What I Learned From Dead Man Walking

Hi. In the midst of working on Dead Man Walking 1.1, I thought it'd be interesting to talk about what I learned while developing it. It'd be nearly impossible for me to fairly go over just what I learned from the 7DRL given how much additional time I've invested, so this post will go over everything related to Dead Man Walking - 7DRL and 1.1 alike.

First up: accessibility. Dead Man Walking is really easy to get into - there's no character creation process and the number of funky controls is greatly reduced when compared to traditional roguelikes. Beyond that, the theme of the game ties directly into the core mechanic (Willpower) and the game wastes no time telling the player that.

Making the game very easy to get into was a conscious decision during the 7DRL stage of the game. I didn't have a lot of time to make an interesting character creation system, so the best course of action (in my mind) was to simply not have one. This seems to have worked out for the best.

Second: gradual build up in difficulty. I deliberately made the first floor quite easy. Death on the first floor requires a very serious tactical failure and should only be a possibility for the most novice of players. There's only one enemy type to face on the first floor and it's the simplest enemy type there. This allows players to learn the controls and collect some items with relative security.

When a player descends onto level two, two major changes occur: you find a new enemy type (the skelly) and you begin encountering bosses. Each floor introduces a new boss and a new enemy type (though they aren't necessarily related). This is a trend that the game begins on floor two and carries on till the end of the game (on floor five).

Third: streamlined consumables. This is a change in 1.1 wherein the player is no longer capable of collecting healing flasks. Instead, they begin with a set number of flasks on each floor and must utilize them appropriately  This means less inventory management and more interesting gameplay.

Fourth: bosses don't have to be tank & spank. This technically started in the 7DRL version with Yreek the Invincible, but has been expanded to cover every boss in 1.1. One of the major design goals in 1.1 is to ensure that no boss fight will let you easily stand in one spot and smack a boss until it dies. Generally speaking, those types of bosses are fairly uninteresting and they barely even feel like a true boss. Bosses in 1.1 utilize a variety of new mechanics to keep fights interesting and potentially deadly.

This is something that I tried and failed in @Star Wars. However, new mechanics and a better idea of what makes a boss fight fun has allowed me to make more enjoyable boss fights. This is one aspect of Dead Man Walking I'm definitely planning on putting into @Star Wars.

Fifth: no need for excess. I chose to make three uniquely different weapon types instead of making a whole bunch of different weapons with a long list of tiers and all that traditional madness. In general I believe that this works quite well and that each weapon fits a play style nicely.

On the topic of weapon types, balancing a game around the player only having access to melee is much easier. While I have no plans of removing ranged or melee weaponry from @Star Wars, only working with one type of distance for the player has allowed me to carefully construct enemies that work within that spectrum.

The most significant takeaway for me is this: a game doesn't need to be long or complex to be fun. Dead Man Walking is focusing on being short and polished. In my opinion, it's already superior to @Star Wars, though @Star Wars can only benefit from Dead Man Walking in the long run.

Peace!

Dead Man Walking 7DRL Source

I've decided to release the folder that I used to create the 7DRL edition of Dead Man Walking as is. You will need Python 2.7 and possibly libtcod 1.5.

As mentioned in the source, my work is distributed under an Attribution-NonCommercial Creative Commons license.  You are free to do whatever you want with the source code provided as long as you don't re-release it commercially. A minor credit towards me would also be appreciated.

I haven't decided if I'm going to maintain the source code releases alongside the standard releases. If the open source release is popular and/or helpful, I'll consider releasing all versions of Dead Man Walking openly.

Keep in mind that the source code is an amateur project and I had no prior experience with programming until I started @Star Wars ~2 years ago. The code is poorly commented and lacks any serious documentation. Feel free to ask questions about what does what if you get stumped by my lack of clarity.

Download!

If you have trouble accessing the source, please make sure you have Python 2.7 and consider getting libtcod 1.5 as well. Beyond that, I probably won't be able to help you a great deal. I can help with issues in the actual code, but getting into the source is not really my field nor do I intend to spend a lot of time on it.

If you want to troubleshoot me about the source or have a cool game/feature you want me to try, you can:

1. Comment on basically any blog post (on this blog, anyways).
2. Tweet me (twitter.com/roocey).
3. E-mail me (roocey at gmail dot com).