Okay I saved the stuff that you unlock later for a separate post so nobody gets worried about spoil-o. If you believe in spoilers, don't read this unless you've unlocked all characters and maybe gotten a couple hundred points.
Ixxthl, fairly obvious counterpart to Susannah. I thought for a while this was too much of a restriction, written in my working notes is "only-blue hero is not viable, find replacement". But I realised the heroes didn't need to be balanced against each other, plus I got better at the game and learnt to play her. Rune effect needed to be huge and damaging enemies was too close to Susannah. Was playing with polymorph effects (because 868-HACK's .POLY was top), had a wand that randomly transformed whatever it hit and it wasn't going great, I realised this might be better as a hero power so you have more control when to use it. It wasn't quite enough as a hero power, too often you would cast it and get stuck with bad enemies - this works differently to 868.POLY because casting it takes a turn so you can't repeatedly reroll, runes are a more limited resource than credits/energy were, and there's more variance in how bad specific enemy types can situationally be - in Imbroglio even one enemy can kill you if you're fighting it with the wrong colour. So I made it transform everything into a new enemy type that's vulnerable to all attacks (and then came Scythe etc.).
I've seen people commenting that Ixxthl's weapons are pretty different from everything else. I don't feel that way because I had everything as one complete set before I arranged it to unlock, but I can see why they feel it. I put the unlocks in order of complexity, the cards with longer text and more advanced combos come later so beginners aren't thrown by them, but this means the four most complex weapons all come at once in the last block; it would maybe have been smarter to spread the complexity around more.
I wrote about this already but there's a few more things to say in context of the character. I was interested in foreshadowing unlocks, so since Ixxthl was probably wearing Magic Ring I renamed it after her, thinking you might be able to guess her rules from this and from noticing which weapon slots aren't filled yet. I tried various ghost-based effects with it so it would connect with both halves of the character but the main effect is complex enough that it ended up better to keep a simple version. For a while it had "when this kills an enemy, a random cursed enemy becomes a ghost" but when gems got the basic rule of curse-ghosting it seemed superfluous to have other effects replicating it. (Plus it wasn't such a good connection with Ixxthl because she makes ghosts already, would have been better to have something that triggered off ghosts.)
The stun multiplier effect started on a red weapon: "Echo Hammer, II: when this hits an enemy, hit a random enemy, IV: when an enemy is stunned, stun them again". Was going with a general theme of doubling things but apart from that the effects didn't make much sense together and both were very strong on their own. At the time there were a lot more cheap stuns - Blink Dagger stunning all enemies when it teleports, for example - so multiplying stuns got very broken very quickly. I really wanted some kind of stun combo though so I made it only have a 50% chance. Later I realised all those mass stun sources were a problem even before any combo so I cut them down a lot and I thought I could let Rope go to 100%, but now because of the dominance of Wheel/Caltrops/Rope combos I've dropped it back to 50% in an update.
An alternate rune-source seemed like something the game should logically have, to enable builds that rely on character powers to survive further into late-game. Tried several possibilities. This used to have something like "I: when this kills a cursed enemy, gain a rune, IV: when this kills any enemy, gain a rune"; like Vampiric Spear, giving extra for curses while still having something to offer without them. This ended up far too efficient; getting extra runes early made surviving to late-game a lot easier, and then getting double runes late meant you could not just slow the inevitable dwindling of your resources but actually turn it around to get more back than you were spending, and possibly survive forever. Went with just a late rune effect and tied it to an unrelated but useful early effect that I'd been struggling to find a good home for.
Definitely wanted some kind of alt-scoring, it's a classic part of constructed-deck games, give an unusual goal to build towards and the value of everything else shifts around it. Usually you score by taking steps towards the gem, the wheel scores by making the gem step towards you; a simple inversion.
First version I tried was based on 868-HACK's .PULL, the gem was dragged directly towards the player even through walls. This was simpler to implement and I thought it wasn't going to make too much difference. It turned out very powerful, I tried adding extra costs, they weren't enough, gave up on it, deleted it from the source code and swore never to use it. Later when testing Ixxthl boards I felt a need for another way to deal with wasps so I was testing a blue weapon with extra damage - percentage chance of doing a red hit as well, red damage against cursed enemies, chance of just doing more blue damage, etc. I wanted an IV effect to go with it, eventually came back to thinking about the gem pulling effect and wondered "what if I use the really ridiculously much weaker version that doesn't go through walls? i know it seems really bad but the other one was too strong so maybe..". Tried it and it worked great, much fairer than the earlier one because it was always an incremental movement of the gem, rather than instantly making the path much shorter (or longer), now pulling the gem to you always takes the same amount of time as to walk to it. But when I tried building a board specifically around it, it revealed itself to still be very strong so I started cutting back on the extra damage, maybe it's only like a percentage chance of doing 1 blue damage to a cursed enemy or something, stack on the limitations. Messy and complicated so I simplified it out by just letting the IV effect stand alone - seemed okay. Was a bit sad to lose extra interactions that might make you want to include it even when you're not specifically using it for scoring but that's turned out okay because there's still different ways to build around the scoring effect.
I'm really happy with how Zugzwang comes into play when you're running a Wheel engine. If you could just stand on the Wheel and hit everything coming at you you're happy, especially if you have its damage upgraded and some nice protective effects to keep you alive. But parity sometimes forces you to spend runes (limited) or move off it and maybe take hits, and that's what keeps the combo non-trivial. Sometimes you must collect a gem off-wheel and then if the wall generation is unkind you can find yourself with a long journey back home. Everything has to have holes in it and Zugzwang/parity creates some good ones.
I like to think of it as giving out fractional points, it's as though it says "when this kills an enemy, get a quarter of a gem".
This was the other ranged blue weapon. Same structure as Malign Wand; costs mana so you don't get blue range for free, gets something extra at IV just because. Delaying enemies kind of made sense at range because it's not like you're hitting them a bunch of times anyway. Just turned out really strong with Ring, you could stand there and timewarp through a whole string of enemies; slow and frustrating but very safe (similar bad feelings to heavy stuns). IV effect was something complicated like "when this kills an enemy, curse all enemies currently being sent to the future" but that was weird and too much work to set up. It's very hard to level anyway because most hits send the enemy away and then probably you kill them with a different weapon, so there was no reason for the IV to be so minor and fiddly. I was mostly avoiding the simple "when this hits an enemy, curse them" because reliable free cursing tended to throw off the balance, but it's okay here because it's so costly to level. I tried the "16 turns" effect scaling with level, "4 turns per level" or whatever, but it didn't make much sense because it often wasn't clear whether more turns was better or worse.
I felt it was being a little underused (actually most weapons with mana costs are - since I weakened Ring very late in development maybe everything that depended on it ended up a little low - though of course I wanted them to work without Ring too) so for the update Leon suggested Johnny-style telefrags as a way it could sometimes eliminate threats rather than just postponing them. I think it adds a lot more nuance.
Minotaurs started at 8/8 but that was too much, it demanded either having a 3- or 4- damage weapon or being able to get the first hit with a 2-damage one. Especially tricky for a hero like Ixxthl without access to the high-damage red weapons. You could build a combo that deals with them but depending on the luck of what you'd been able to level up maybe it wouldn't be active yet, I didn't want the rule to be "you lose if your combo isn't up by turn 100" or whatever, obviously you want a combo running as soon as possible but putting a hard limit on it removes any interesting play around dealing with things going less than ideally, it would mean you might as well simply resign if you don't hit a certain milestone. So I dropped them to 4/4, anyone can kill them but they provide a little extra mid-game challenge.
Then when I started working out what happens in the super-late-game, once you've had time to almost certainly level everything you want to, I came back to the idea of an 8/8 enemy. I had combos that could deal effectively with any number of small enemies, and weren't too bothered by minotaurs either, so how was I going to manage to end the game? I didn't want it to run forever. So I threw in the 8/8 monolith and tried my combo on it and quite easily killed that too. I've toned down some of the weapons I was using there, it wasn't always clear in development whether to make the enemies stronger or the player weaker, but still it's not hard to deal 8 damage if you need to, that's just two hits from a Broadsword or Brazier. So I took it to 16.
The 16/16 endboss worked pretty well, there was a good chance you'd die to it in a satisfactorily glorious manner and get a nice score, and if you killed it you were often weakened enough for the now-vast hordes of standard enemies to take you down. It was hard to decide whether it was red or blue so I picked both; split it into two versions each with a weakness: 16/12 and 12/16. I wanted there to be ways of killing them, just "ok this spawned so now you die" feels bad. One of the ways of dealing with them indirectly was to use a Scythe or something to transform them into a ghost, but when I gave curses the extra effect of ghosting this became much too easy to bypass the bosses rather than feeling like a feat of cunning trickery; now it could happen without you even trying and it felt very anticlimactic so I reluctantly added a rule preventing them from ghosting.
It seemed good to get some kind of reward for killing a boss, a damage increase to the weapon used felt reasonable - opened up some strategic options letting you still improve a weapon even once everything is fully leveled. Both bosses had this effect for a while but it seemed more interesting for them to be different. Damage increase is a red effect so I needed something for the Dragon. Something that also applied to the weapon used would have been nice for symmetry but there were no simple options; you could level it up but at this point of the game it's probably already there, making it ranged isn't strictly advantageous because it restricts your movement (increasing damage isn't /strictly/ advantageous either because occasionally you want to hit again so as to wait in place but this is rare enough to overlook) plus it would have broken Harry's disdvantage, modifying the weapon's rules is possible but overly complex. Tried having it give out treasure; runes and gems;
nice to get some extra runes at this point of the game when they're running short (unless you're Orbing). Four runes and one gem seemed okay, except the time that I slew a dragon from a nice safe position with walls on 3 sides and the gem took all those walls away leaving me to die (my Whetstone'd Spear would have happily dealt with everything one by one). Then someone suggested the max-blue increase and it just seemed right. It's simple, succinct, makes sense in terms of existing effects. It has symmetry with the Unicorn - upgrading the weapon, upgrading the hero. I had been thinking a "blue shield" would be an obvious item for an expansion set but it seemed nicer to have the blue version exist in a very different form and avoid the straightforward symmetry.
Favourite thing about these is people's reactions when they first see them.