Kept having more card ideas and throwing them in, ended up with a huge pool of cards and nobody could keep track of them all. I quite liked the feeling of it, the dungeon may be a tiny grid but the possibility space of board construction is a cavernous labyrinth rich with hidden treasures. But it was too much for me to test and tune. Also other people who tested it consistently reported that they found it overwhelming to explore but I wasn't bothered by that, someone wants my terrifying death maze to feel safe and comforting they can maybe go somewhere else. Still, to practically be able to finish the game I started shifting some of the ideas that weren't working well yet off into a separate list of "possible expansion items??". I wanted the initial game to be still be big enough to create a variety of possible boards, don't want people to just play a couple of different setups and feel like they've seen it all, but it was helpful to stop feeling like I had to fit in every cool idea I had now.
So as soon as I had the game out I was right back into testing the weapons that hadn't made it. At first I just had a random mish-mash of items, I preferred to have some kind of theme to tie them together so it could feel like "here is the dwarven erotica expansion pack" or something not just "here is some more of the same stuff as before". Thought about introducing a New Mechanic (magic-the-gathering-speak for a rule that appears on multiple cards) but the things I wanted to add were mostly about filling gaps and expanding possibilities of what was already there rather than having room for a separate new thing. Then Jonah suggested that since Skeleton Festival was coming up I could tie it into that, it's not something I personally celebrate but it's big in the USA which is my main market, plus it fit several of the items I already had and some others could easily be renamed to match it. It ends up maybe not very different from the fantasy rpg dungeon setting there was already but I think it's enough for expansion identity.
Things I wanted to add were: more ways of cursing and using curses to add more variety in possible curse boards, useful effects to help survival especially in the early-game, new combo endpoints to build around, more ways to hit multiple enemies, combos to encourage using the items that weren't seeing much play (especially ranged weapons and weapons with mana costs). With more items, everything is going to see proportionately less use so there's room for some rarer situational effects. Also some new heroes, that's like adding new levels because they're different challenges to build around.
Every roguelike now gets a bunch of people saying "PUT A DAILY CHALLENGE IN IT" whether it makes sense or not. I didn't want to do this with 868-HACK because the level generation isn't determined from the start, it depends on the position of enemies left over from the previous level - this is a subtle rule you don't notice while playing but it goes a long way to making it feel right; it's important that enemies are carried between levels (both for balance and atmosphere) and it feels bad if they get stacked on top of walls (you can move them on top of walls with .PUSH etc. but that's because you're hacking things, it's not the natural order). Point of a daily mode is that everyone competes at essentially the same challenge, so it doesn't make sense if it's going to chaotically diverge. Some effects in Imbroglio give similar problems - a new gem's position must be at least 4 steps away, so if you've moved the gem with Dream Wheel then maybe the position it would have been is no longer valid (maybe it's even right under your hero). But these problems are less central so I was prepared to work around them; probably wouldn't matter too much if a few times Wheel lets you score slightly cheaper, or else it probably wouldn't matter too much if gem positions diverge once you start driving a Wheel - level layouts aren't as critical as in 868-HACK anyway. And arguably Imbroglio doesn't do enough to stop you restarting a bunch of times to get good luck, means weapons with an early-game advantage are undervalued / weapons with an early-game penalty in exchange for late-game power are overvalued, so this kind of "you just get one try" challenge might be a good fit. 868-HACK has streak scores so doesn't need it.
I tried random boards, they tend to be really really bad. Not even entertaining challenge to play around, just bad. Tried having a limited set of items to build a board from, seemed potentially interesting: removes the option of just playing the best board anyone's found, to encourage you to play interesting sub-optimal boards. Needs enough room to still make different things, I estimated it would need about twice as many cards as a full board so you could possibly make two boards with no cards in common. So I generated sets of 32 random cards (with possible duplicates), this ended up feeling too overwhelming. Tara suggested just having 8 cards and you can put up to 4 copies as usual; doesn't work with Vesuvius Bob but otherwise plausible. I worried this would be too small a set but trying it out it felt okay.
Part of the process of constructing a board is playing it out and adjusting it, so in addition to playing once with the daily seed I needed there to be an option to test-play your board. If this option wasn't there you'd still be able to test boards by constructing them in the standard game mode, it would just be tedious going back and forth. But this was a lot of extra complication, more menus, more variables to keep track of, plus you might get a good score while testing but not replicate it when you play it for real and that feels bad. So why not just let you have as many tries as you want? (Bonus: avoids any problems with having a fixed seed.)
Daily challenges are too frequent for me, I don't have a reliable daily routine, usually I play them for like two days in a row and then have a few really busy days and miss them and then I never come back because I broke my streak. Also constructing a board takes time, you want to be able to test a few different possibilities. So maybe more like a weekly challenge? A week is quite long though, hey what's a number between 1 and 7? Okay so I wanted a name for a 4-day week, if it was a science-fiction kind of setting game then like "tetrad" or something would do but here it didn't feel right. I had a look for any existing calendars using a 4 day week, found the Igbo calendar, borrowed the name "izu" from there. (This could be seen as cultural appropriation; I've used a word from a culture that is not my own and that honestly I don't know enough about. My intention was to respectfully acknowledge prior use of the concept.)
Chaos Oracle Molly
Control of random selection was originally an item effect, something like "Torch: 5/10/20/25% chance to select a target for any random effect". Nice for players who find the game too random, giving them an option to take more control. Disrupts the flow of the game so it's good to restrict it mainly to one character. Her disadvantage restores some primordial chaos to keep the cosmic balance. Matching enemy types to gates was the main breaker of board symmetry, but I figure it's okay for one character to have a more limited space of boards especially with new items expanding it anyway. (Technical detail: for Molly the "3-in-5" rule applies to both gates and enemy types.)
Simon Rainbow Jester
Changing weapon colours also started as an item. Something that changes itself, something that changes another weapon. Raised questions about whether Susannah/Ixxthl can use it, had some really strong interactions, in general was pretty wild, breaks down the careful work I've done to set the tone of red and blue weapons. Better as a character so all that weirdness is contained in one place. Super unfair that he gets the coolest disadvantage as well as the coolest rune effect, but they feel like they fit together even though they're mechanically unrelated.
More resource exchanges! Was almost in the original set but someone convinced me that an alternative way to spend runes was a special enough thing to save for later. Would be fine without the cursing, but I want more weapons that just curse a little bit even if you're not building around them, little pieces that can work together. I don't want a binary "this is curse board so it has all curse things" vs "this is not curse board so curses thing is useless", I want reasons to end up somewhere in between, maybe because you have a little bit of cursing you want something else that's stronger with curses, then maybe that makes you want something else that grants curses, enigma.
I was trying something that scored gems on levelup, very simple alternative scoring mechanism, but didn't feel right, I guess you get a lot of gems anyway so a static bonus isn't that great, something repeatable like the Wheel is more attractive. Plus the healing from the gems is often wasted; in general I moved away from giving effects on levelup because it's really hard to time them usefully, either you level up anyway and they get wasted or you're going to great lengths to avoid leveling and that ends up very inefficient. But runes we're getting already and the specific timing doesn't matter because you spend them when you choose, plus they're scarce enough that a few extra can be significant. Hopefully useful for heroes with a difficult opening.
A version of this was the second item I put in the game (after Whetstone), it attacked enemies that stepped on it. At that point when the game was mostly in my head this seemed really exciting, I could imagine all the ways it would interact with positioning, you could combine it with a Whetstone, you could block off a gate to eliminate one enemy type. But once in the game these interactions seem more obvious, fun to play out once but not so interactive. Spent a long time trying to find a version that wasn't ridiculously overpowered, didn't find one in time so it got pushed off to expansion file.
Thinking about how can I trigger a remote attack without it just always happening automatically, maybe there's a button I can press to trigger it, like maybe the button I already have right here. Nice with Runestaff and Sigil, we've got a rune sub-theme going. Really like this version, how it plays different for every character because their rune situations are different. Totally different from the old version, with that the challenge was to get a combo set up and then it would automatically deal with things, here there's a constant challenge to use it effectively during the game. Adds a lot of potential for skillful play, extra depth to decisions, you have to anticipate problems and deal with them early because the Bonesnare might not be where you need it on the turn you're about to die.
Usually I prefer every weapon to have a strict improvement at level IV but I was finding this just a little weak at 11112 and sufficiently complex that I didn't want to go for 11122 and a IV rule, but I figured that since it triggers off runes it's okay if IV just gives you another rune to activate it.
Another early weapon, the first curse combo. Curse them with a wand, and then every turn after that the venom drains their life. That's too much, maybe just whenever they move, or whenever they attack. Still a lot given that you can stack up four copies of this. Eventually hit on the idea of triggering it when an enemy is stunned, turn a defensive effect into offensive. There were a lot more sources of stuns and curses back then so ordinary damage got too strong, I made it damage an enemy's highest life total to slow it down. This made it quite wordy, and it was still very powerful, plus building a curse+stun+venom combo is quite complex so I cut it from the base set.
Coming back to it for expansion, I found it could be simplified now there were fewer stuns. Letting the effect be "hit" is nice to allow combos with anything that interacts with hits. Making it only trigger once per turn removes a lot of the problems with stacking a bunch of them and instantly killing everything; I hadn't had the "once per turn" rule in the game when I'd worked on it before. Tried having it replace the stun rather than just trigger off it, so four copies will need four stuns, but went back to the simpler version. Power level might be a little low but I'd rather that than the omnipotent earlier versions; the combo still works. I haven't gotten 256 with this version but I can reliably break 200, fine by me.
Started as Ghost Sword, Blue 11222, 0: Attacks through walls, IV: Can move through walls from here. Attacking through walls turned out to be not great unless you kill them in one hit because they can't hit back so they move as usual so now they're an even number of steps away and you have to take a hit to move back next to them. But moving through walls was cool. Plus blue items stopped being explicit martial weapons and got less damage so it wasn't a sword anymore. Then I was looking for more things to stick mana costs on and this one made sense, it was a bit strong for free anyway.
Ended up killing it because it had bad interactions with Bob and Masina. With Bob it's ambiguous whether you're swiping towards a wall to break it or to move through it, there were situations where you'd definitely prefer one or the other; it might seem like moving is better than breaking (because you break it to move through it anyway) but sometimes you want to break just for the wait to get in phase with enemies. Plus things can be triggered by spending mana or runes and you want to know which one you're going to get. (Eventually I made it trigger both effects because that's logical and nobody can complain they didn't get the effect they were asking for, even if it's not really the outcome they wanted.) With Masina there's the question of whether you can get into the inaccessible square; obviously it's not a desirable place to be because then you're stuck, but it feels bad to not be able to use your key anyway. And there was a high chance you'd include it when building a board for either character; Masina because she gets level layouts with more walls and it might seem like a way to deal with that, and Bob simply because he includes so many weapons. These interactions are still not great but i figure it's safer to put in an expansion because with more weapons and more characters you're just less likely to meet them, and people have more experience with the game so they can maybe deal.
Spent a while fumbling for an IV effect. Had 11112 damage for a while which was okay but an ability would be cooler. There's not a very clear rule behind which blue weapons get 2 damage, it's basically if they need a bit more strength or if they need an IV upgrade and I couldn't find a good one, but also based on what just feels right. I had a weird effect where it opens up portals in the edge walls (hey if we have bad interactions with some heroes why not bother Johnny too) but stacking all this extra mobility got too strong. Leon is constantly pressuring for more "telefrag" effects (he's responsible for the update to Hourglass) so I gave in and went with that. It's very situational so probably you won't prioritise leveling it.
Red version of "Ghost Sword"; adding range fixes the parity problems when attacking through walls. Was almost in the original set but I realised I wasn't going to be able to design many different ranged weapons so I held it back. Fits here in keeping with the "skeleton" theme.
Gravestone & Bog Hands
See Ixxthl's Ring; I wrote about finding that triggered effects tended to be preferable to rewriting weapons. I had a few weapons that attached extra rules to other weapons, like "Barrel of Tar: when this levels up, a random weapon gets 'when this hits an enemy, 50% chance to stun them'", and I wanted to find simpler versions. Here's a solution: effects that trigger when an enemy is there. Analogous to grafting an effect onto adjacent weapons. These also give a bit of a boost to ranged weapons, more things that can reach the target tile.
Used to be "ghosts are your friends". Back then I had more ways of creating lots of ghosts so this was even stronger, had to balance it with stuff like 0 damage at IV, maybe feeding restrictions. I didn't mind that it wasn't stackable (you could still use a second copy to convert a few waves) but actually it is nice that the current version does stack; more masks means more friends. There were problems with having too many ghosts, maybe you could get 15 of them and then the screen is full and no new enemies can enter, limiting the number of ghost friends deals with that while keeping the fun situations. Also moving into a friendly ghost used to swap places with them, now they're still treated as an enemy in most respects so you can still attack them (or sometimes, must still attack them), sort of a disadvantage but it lets you still use them to feed your weapons, in keeping with sinister BLACK SWORD BLACK SWORD.
Basic utility item to avoid being stuck in bad corners. I like the efficiency of getting a move and attack in one action. Maybe a little close to Blink Dagger but it has the stun and more predictability.
Problem with Ixxthl's Ring in enabling a mana-spending board is that it's blue and all the things that spend mana are also blue, so you're pretty limited in how many you can have given that you want some red weapons too. Here's a red mana-spender to take some of that pressure off. It compares closely to Rimeclaws; has more damage at IV but that's in keeping with red weapons generally having more damage. Possibility of combination with on-hit effects. This might not be very clear in the picture but I tried to make it a little bit SPOOKY by the drum being made from a PUMPKIN and the drumsticks being BONES.
It made sense to me to call Molly an oracle but she doesn't actually predict anything, her power is about manipulating the outcomes which is kind of like predicting them only better. But how about a tool to predict things as well, something that ties in with her rules. Unconstrained enemy spawns can actually be pretty hard to deal with, getting some information about them helps. Then a random effect to tie in with the other half of the hero. Still useful for other heroes. Quite close in strength to Cleaver, it's like 11123 except the 3rd damage at IV is sent somewhere else (sometimes better, sometimes worse).
Really interested in what else I could make as a vanilla weapon. Cleaver doesn't leave much room for anything else to reach 3 damage, Brazier is canonical for diminishing damage, what about fluctuating damage? It's good because it starts at 2, the only other thing that does that is Broadsword and that has an xp penalty. It's good because it gets to 3, hardly anything else gets to 3. Maybe you want to stop there, and for that 3 is better than 4 because you can get Minotaurs and Cubes down to 1 and then finish them with something else.
I thought I wasn't going to do a "blue shield" after Dragons got that effect. But when I was testing Wheel versions, percentage chance of extra damage was working really well and I wanted to use it for something, spent a while trying to find a nice simple effect that would go nicely with it without feeling out of place but without synergising too directly with the damage (which was already strong). Came back to this, decided it was fine as long as I acknowledged the connection with Dragons. More mana is good anyway to encourage using the weapons that spend it.
My early versions of a "blue shield" tried to distinguish themselves more from red shield by having stun/curse effects rather than just damage. But still the method of damage is different and that creates some asymmetry between them. They have similar average damage at IV but Skull's unreliability makes it slightly worse, in keeping with the principle of red weapons having more damage. Also with less enemies targeting mana it's probably less significant overall.
Used to be a rule-grafting item, giving other weapons the ability to hit multiple cursed enemies. When cutting back on those and needing more basic curse sources I rebranded it as obligatory walking curse box. Weird. Tried a bunch of different curse punishments to go with it (when cursed enemies are hit here, when this kills a cursed enemy), ended up going back to the idea of hitting multiple cursed enemies at once.
Thinking through alternative ways to deal with boss enemies, also more tools for Ixxthl to survive, okay the mind control thing is a fairly standard game effect for turning an enemy's strength against them. Already had the Mask converting ghosts, it'd be simple to just reuse that effect, switch their team. But what if you actually directly control them? Can't lose the roguelike single-character controls so it has to be controlling them instead, that's pretty interesting because it leaves your hero vulnerable. So I was walking an enemy around, I moved them onto the gem, it felt weird that it didn't pick it up, well maybe it should pick it up - otherwise you're stuck not scoring for a while, it might be efficient for killing enemies but it's going to leave you behind in score. Suddenly it's an alt-scoring tool. But then that was kind of bad because the enemy you're controlling would keep getting healed up by the gems you're collecting while your hero withers away in the corner. So I switch the gem healing to still target the hero even though the enemy picks it up, now there's a decent time limit on controlling any one enemy. When the controlled enemy kills other enemies, that should count as the Spoon indirectly killing them, so it levels up. When the controlled enemy dies, that probably shouldn't count as a kill for the Spoon because it was the enemies hits that did it. Needs an IV effect, how can we make controlling an enemy better, +damage ok.