I have thoughts on the upcoming additions. My thesis is that performance in combat should affect the development of the player’s economy, and that developing the player’s economy should develop his combat forces. This relationship between the economy and the army should be such that there are a bunch of little interlocks between the two so that the player is required to develop both in order to increase his power.
Floor System
The floor system looks promising, especially the part about magical effects producing different outcomes depending on who stands on them. It’s the first instance of strategic decision making at the macro level. I’d like to see floor types influence the specific abilities that minions can learn.
Experience system
It’s good that you tempered the experience bonuses. I wonder if the upgraded version of furniture could just recruit more competent minions (grave stones, work benches, and such could all be upgraded). This means that the base experience level of your (non-legendary) minions would reflect the stuff in your dungeon and this experience level could be augmented by 50-100% through training or successful combat.
Why do it this way? Such a system would serve to buffer the player’s power curve. It also makes your economy more relevant, so you can’t just one-minion your way through an entire campaign. (Of course, enemies have to gradually get stronger as the game progresses, otherwise it’s unfair).
resources
I’ll also test whether it makes sense to add an analogous method of iron production, such that you’ll melt iron ore in furnaces to create iron plates,
Factorio, the Roguelike? (Well, I hope not :p)
But I have a couple of suggestions regarding resources.
1. Torches should also require a constant input of wood. Since wood production is dependent on the number of imps you have, the player has to grow his imp population to grow his base. (He also has to place torches smartly, which makes torch placement have SOME consequence, rather than being completely cosmetic).
2. Get rid of the Tree Spirits that spawn from cutting down wood. They don’t really do anything from a gameplay perspective (other than provide free XP because they attack so slowly).
3. Iron ore requires smelting into iron plates, as you suggest, and smelting should require wood directly proportional to the amount of iron smelted.
4. Maybe require iron picks for mining granite, and then granite for producing steel. This would form sort of a progression for resource mining (wood -> iron -> granite -> steel).
Lastly, keep gold the way it is: to increase supply, or for purchasing legendary items from merchants (of which we need more, by the way). The way I see it, it should only be acquired as the spoils of conquest or discovered in remote squares, because it should be representative of how many enemies that the players has vanquished and as an indicator of overall game progress.
Tesselated production: the mechanic to save base-building
I think this one idea that could make for interesting geometry puzzles that synergize with the “griddiness” of the game. Best of all, it doesn’t try to rework the game, but it combines the base-building mechanic and the permanence of facility construction with the way workers randomly hop around manufactory tiles. Here’s how it works:
– Items no longer spawn completely randomly, but according to the last n tiles that the worker walked on.
– For example, for a sword to be built, the worker has to walk on a forge-tile, then a workshop-tile, and then another forge-tile (and then roll a dice check to build the sword). On the other hand, to build an axe, the worker needs to hop on a workshop-tile, then a forge-tile, then a workshop-tile (and then roll a dice check to build the axe)
– Thus, by altering the PATTERN of factory tiles, you alter the types of items that you produce. The idea relies heavily on the feature that resources are finite, and facility costs cannot be refunded
– Because manufactory types affect the recruitment of unit classes, introduce several sub-types of each manufactory. Instead of alternating workshop and forge tiles, for example, alternate WorkshopA (W_a) and WorkshopB (W_b) tiles in a checkerboard pattern. Manufactory sub-types are upgraded from the manufactory base type on a tile-by-tile basis. In the beginning of the game, for example, you would only have access to the base workshop (W_o) tiles (researchable at the cost of mana). By conquering some minor foe, for example, you might gain the knowledge of W_a construction. It’s important that the world map somehow restricts which tiles you can produce. It’s no fun if you can just make the exact same pattern every game.
– This idea supplants the simple “production efficiency”
That’s the basic idea. How can we play around with this further?
– This idea requires the existence of rooms, which are continuous areas of one tile type (and all subtypes). For example, one W_a tile touching one W_b tile is a 2×1 workshop room, but a forge touching a workshop tile is two 1×1 rooms. Workers will now need to be assigned to rooms instead of just tasks like “laboratory.”
– It’s not limited to building things. It can also be applied to training and rituals. For example, an orc could learn a multitude of different skills depending on the patterns of tiles in the training room, instead of getting a generic and comparatively boring “experience level.” Personally, I don’t like experience levels for minor allies like orcs, ogres, or anything else that isn’t “legendary.” I’d be much happier if these sorts of creatures were defined by which abilities they could learn. But that’s just me.
– Instead of upgrading tiles into subtypes, perhaps they can be (reversibly) modified by dropping items atop them. For example, a laboratory tile with a dead rat on top of it functions differently than a vanilla laboratory tile.
– Alternatively, the item itself could be modified; for example, if a sword is dropped on top of a jeweller’s bench, then the sword itself might be modified if the jewler happens to land on top of it at the end of a particular movement sequence.
Idea: Get rid of digging out tunnels for mining.
Instead, the player gets minerals from nearby zones that are under his control. The player needs to allocate imps for this, and each imp can only bring back so many minerals per day (possibly at diminishing returns). The benefit of doing this is that a) the player doesn’t have the tedious task of mining, and b) the game designer can more easily tune the number of resources gathered per imp per time period. Mountain zones could have minerals, wood zones could have wood.
This is specifically about the construction of bases and how it could require more cerebral input than it currently does, because right now, base-building is just a chore. All you need to do is to dig a serpentine path into the mountain and place a few boulder traps at the corners. From there on, whatever design you create doesn’t really matter, so long as it’s well lit.
I suspect that a lot of people would play this game for the zen-like experience of designing their own dungeons, which I get, but it really does make the game a lot easier. Maybe there could be two keeper modes, one “competitive” and one “zen” mode? Other games that feature some base-building aspect have this (OpenTTD and Rymdkapsel, to name two that I’ve played).
Grievances:
– I really hate traps. They let you hold pretty much any attack. There are so many situations in which you’d die without them, but you’re perfectly safe if you have them. And they’re easy to get, on top of that. Personally, I think it’s a bad idea to allow the player to create any defensive funnels. I get that people like to dig impressive dungeons, but from a strategic perspective, when you let the player make any formation that he likes, it’s incredibly boring, because the player will always find a way to exploit the AI.
– I don’t get the point of placing torches. They don’t cost anything; you literally just put them there to make your facilities run at full efficiency. This limits the size of a facility that you can create, but torches might as well place themselves automatically. Maybe they could cost wood to burn? Maybe they could present a fire hazard if gas builds up?
– The actual act of mining is rather pointless. I guess in the old days, you had to travel far out on the map to get minerals (well, not really), but now, all of the minerals are just in your home square.
Aspects of the game that I like, and which could be exploited further:
– In terms of strategic building, the game forces you to spend resources on facilties. You can’t refund this purchase, either, so whatever you build, you’re stuck with. That’s a decent starting point, I think.
– Facilities become more efficient as they grow larger because the fraction of border tiles to total tiles decreases. Combined with the fact that you can’t get a refund for anything that you’ve placed, this becomes a potentially interesting game mechanic. In Rymdkapsel, for example, rooms and corridors are built out of tetris pieces, and once a corridor is placed, you can’t refund it. Because certain types of rooms are more efficient when placed in proximity to other types of rooms, the game is a big geometric puzzle. (KeeperRL has a lot more stuff going on than Rymdkapsel, of course, so it’s not really appropriate to make it such a complicated geometric puzzle, but it’s a neat idea for a construction game.)
– The type of facility you build influences the type of unit that is recruited. This is potentially quite interesting because it gives facilities the dual purpose of making stuff AND recruiting different units.
I worry that a full-control mode might slow down the action too much. (It might not; sometimes the action in single-character roguelikes can be very slow if there is a complicated situation.)
Great!
Instead of just allowing any unit to lead any combination of any other units, there are restrictions in place.
For example, maybe a wolf cannot control undead units. Or, the number of greenskin units that a greenskin can command is directly proportional to its experience level.
This was meant to accompany my other idea about “leadership spells,” which are essentially spells that some units can cast onto other units that function as orders. Possibly, a unit could have different abilities or statistics depending on who is leading it.
The idea is motivated by the fact that this game doesn’t have much in the way of party dynamics. The only two playstyles that I can see are getting a bunch of greenskins and training them up (which I used to do all the time before campaign mode came around), or finding a legendary ally and training him up, and letting him solo carry the game (which is my favorite strategy since campaign mode came out, since I can win the game in a few thousand turns that way).
Admittedly, it’s a vague outline of an idea, but it would force the player to look at the relationships between his units and deciding what he could take on a mission, instead of just making the biggest deathball possible.
I get this bug too. Can’t get rid of zombies.
Congratulations! This is quite an achievement. And congratulations again for being invited on to Roguelike radio. You have a nice voice (and a good microphone too; seriously, I can’t understand half the people they invite on that show :p)
I’ve played a lot of Alpha18 so far, almost 8 hours. I made a few observations
and I documented a method that can beat the game quickly, fairly reliably.
In general, the game seems to work well without crashes. I also agree with you when you mention in the video that exiting a combat area needs some rethinking, because it’s quite abusable the way that it is. The game is still quite straightforward to beat. Actually, I don’t think it’s changed much strategically since Alpha 17. It’s just been broken down into bite-sized chunks 🙂
Got it down to 2675 😡
Beat alpha 18 in 3150 turns. I’m getting closer >:)
Keeper vs keeper =
Indeed.
Actually, since you mentioned keeper vs keeper, it got me thinking about Brogue (an awesome game which I’ve recently downloaded and started playing) and how the ally system in that game compares to the one in KeeperRL. I have to say, the team battles in Brogue are quite fun, a lot more so than those in KeeperRL. Obviously, Brogue has been around a lot longer and there are far more interesting tactical options in that game because there are more systems in play, but I think that one of the aspects that makes Brogue more enjoyable is that combat takes place in digestible chunks, whereas you have these gigantic battles in KeeperRL.
This situation reminds me of the difference between new XCOM and DOS XCOM (XCOM Apocalypse, especially). In the DOS game, you could bring a platoon of soldiers to missions whereas the remake limits you to six soldiers, even though it both games you still fight roughly the same number of aliens permission. I think the reason that they did this is because it starts feeling like more of an RTS-style-game-of-numbers and less of a squad based tactics game once your squad increases beyond a half-dozen members. Taking on five pods of four enemies with a single fireteam is more interesting than taking on a gang of twenty enemies at once with a dozen soldiers, at least in games with complex inventories.
In KeeperRL, I still think that the gigantic battles can be fun, but their outcome is mostly decided by macro-strategy. Most of the time, you’ve either won or lost before the battle starts, and if you can’t win, then the best you can do is to run away. It would be nice to have a few more missions that enforce a smaller team limit, which might even be a single creature, so that inventory management becomes a more dominant factor. Just some sort of mission where you can nibble away at a numerically-superior force would be a nice touch.
Okay. I still like the Fog-of-War, and I don’t think you should be able to see everything immediately. The vision range should be just enough to see clearly into the immediate future, and rather hazily after that.
Brainstorming some stuff:
Even if there is a Fog-of-War on the world map, I think that the player should still be able to see everything in a one- or two-tile radius.
Oh right, another question: where do defending units start when an invasion happens?
By the way, did you ever play Jagged Alliance 2? KeeperRL is shaping up to be very similar to that game, heh.