Why handcrafted games have to play it safe with upgrade choices

I've talked before about Doom State - how doomed you can be without dying or knowing about it: • Design Talk: the pros ... Lately I've been thinking about how roguelikes and handcrafted games have opposite takes on whose responsibility it is to prevent it: the player's or the dev's, respectively. That lets roguelikes give you bigger and deeper choices in how you tailor your build as you level up, while handcrafted games can't afford to offer you permanent choices that might doom you. I also reference Desperation Innovation, talk here: • Dev Log: Desperation I...

[@zhoufang996]

This "it sure seems easier" to make a roguelike thing you just said sure seems like something you will REALLY REALLY REGRET at some point.

[@zhoufang996]

@@IdiocyOverdose Proc-gen isn't the same as roguelike. These days, and in the context of what Tom is talking about, roguelike comes with a connotation of a meta-progression. Either explicitly in terms of spending upgrade points at a central hub that make the game easier, or implicitly in that the player is trying to garner experience at the game so as to eventually accomplish a difficult goal. Something like Heat Signature significantly does not have that sort of meta-progression. Each run is designed as a individual experience where the player uses the tools provided to accomplish a micro-challenge. Thus, the goal of the runs on a design level is to be distinct from each other, and are kinda balanced in similar lines to what is explained (a highly skilled player should be able to complete the game without using any special tools at all). Instead, a roguelike would have to be designed in terms of "what are these elements supposed to teach the player" with the expectation that it is only after a series of iterative failures that the player can succeed. In the Slay the Spire example, the game lives or dies on how well it can teach the player the value of the cards, before the player gets annoyed and stops playing, and yet if the player can succeed just on a random selection of cards the player will also stop playing. Trying to balance that kind of meta-progression is, I suspect, going to be a real pain in the ass.

[@ragintombo]

The idea for the next game to be roguelike and done quicker is a fantastic suspicious development.

[@uheartbeast]

I love your design talks. Thank you for posting these.

[@oldsoul3539]

A couple of solutions I've seen implimented were either having a respec button, which is similar to being able to just subtract and add points but gives the feel of "now I've done it over" if adding and subtracting doesn't make sense in the context. Another is making upgrades more like an inventory system, letting you kit out your wizards before the level completely. It reminds me of how dungeons and dragons' wizards operate where their spells get forgotten after they're cast and you have to take an 8 hour rest to recharge them all but in that time you can completely swap out what spells they have memorized

[@Pentadact]

Yeah, for a while I planned to have certain upgrades in Wizards you'd always have but could switch in and out at the start of a level. It ended up not feeling like it'd solve the problem though: it'd still feel crap if you got 5 turns into a level and realised shit, I'm stuck now cos I slotted the wrong upgrade. Redoing it involves repeating yourself in a way it doesn't with roguelikes. And since they're freely switchable, they don't give you that 'my build' feeling that to me is the value of character upgrade choices. We may have some kind of respec, just for quality of life, but it doesn't really free us up to let some builds be crazy powerful and some be unviable. That makes balancing the difficulty of our levels impossible, and if a player is struggling, respec only helps if they're aware the game is not intended to be that hard, and that other upgrade choices would have paid off much better. Given they passed on those upgrade choices, it's likely they're not.

[@normalasylum]

Thank you for flying Francis Air.

[@terencebad7009]

I was glad to hear about your journey with Spire choices. I went through exactly the same thing with catalyst, limit break etc. Now I realise that I wouldn't pick a catalyst with no poison or a limit break with no strength, but the interesting decisions are instead deciding between e.g. a Well Laid Plans that will definitely be useful for most of the run, or eating my greens and picking a Dagger Throw or Sucker Punch for right now so I don't get my head splooshed by the Slime Boss or Gremlin Nob.

[@42VS]

Thanks for the great insight. It got me thinking about Divinity Original Sin 2. All hand-crafted content, not a roguelike, and yet it comes with a ridiculous freedom, both with upgrade choices and game-choices in general. I guess the skipping content thing isn't a huge issue for them cause they made SO MUCH content. Bu they have talked about the issue of the player using that freedom to screw themselves into a corner, and they implemented something called an 'N+1' design philosophy to get around that. I believe Sven Vincke has a GDC talk about it. While I do think that design philosophy might require a decently-sized team to implement, it's interesting to think that they solved the issue at all. The consequences of solving that issue, at least for me, were immense in that game. In most games with branching choices and paths, I have the problem of wanting to explore all of them. Because I know I can, especially with saving and reloading. And I eventually just tire myself doing that. But Divinity was so rich with choices, it kinda felt like real life. Roleplaying and moving forward with choices felt okay. I think there's a tipping point here, where once your game is dense enough with player choices, it affects players in a deep psychological way. The game goes from being a puzzle that must be 'solved' to an experience the player is having. Both are valid ways of playing games, of course, I love puzzle games. But the immersion of Divinity was magical for me.

[@gloverelaxis]

i don't make games but I always love these design talks, thanks! i do UI design/dev and there's a lot of parallels in the way you need to think with empathy and communicate intention through design

[@jabberw0k812]

I think the reason it makes sense to say dying/losing is fun in Dwarf Fortress is that it's focused on simulation, not puzzles or challenge. Having your fortress fail has a high chance of producing a unique story in the process. There's also a broad failure spectrum in the main mode, since you're managing a colony, not a single character. And since the game is actually non-linear, not just randomized, starting a new fort doesn't really feel like restarting. Time continues on, and your failed fort is incorporated into the world and its history. Even in adventure mode, if my character dies wrestling a giant alligator, I've just added color to my existing world, and now another character can find an artifact depicting the time someone wrestled a giant alligator. And my next character might be on a different continent entirely, with a very different experience. In Caves of Qud, which isn't quite as open-ended, the non-linear open world means dying and restarting feels quite different from being sent back to the first floor in Spelunky.

[@gaverion]

This talk reminds me of musings I have had about Final Fantasy X. I always loved doing challenge runs where you don't level up (or other things). I noticed over time that there were a lot of conveniently located items and equipment that let you survive a specific boss mechanic, or boss specific AI that made every fight possible, even if you ignored leveling up. It makes me suspect they wanted to makes sure the game could be beaten, no matter how bad of a state your party was in, and what worse state is there than no levels and no items?

[@AzureLazuline]

i've thought about this kind of thing a lot! One adjacent dichotomy is that some players love just figuring out the specific strat they need like solving a puzzle with one real solution, but other players have a big desire for self-expression, and only really have fun if they solve it in a clever or unique way. Handcrafted games tend to favor the former and procedural games favor the latter, but it's not a hard divide by any means. I actually wrestled with this a little bit since my last game had *tons* of self-expression (lots of build variety and lots of ways to handle stuff) but now i'm making one that's way more linear/"closed", and i was worried people might bounce off if they feel like their playthrough is the same as everyone else's. It ended up not being an issue at all in my case because people were personalizing their experience in lots of ways i didn't foresee, but it seems to be a fruitful thing to think about anyway!

[@GameDevYal]

I've always been a fan of designs where there's no one intended solution, but there's a lot of potential ones in theory... that lets the self-expression crowd pick something that fits them, and the problem-solvers still need to do some digging. Bravely Default does this, a lot of bosses have a very powerful gimmick that makes going for regular fighting pretty much unviable, e.g. the evil fairy spamming the strongest elemental spells in the game over and over (if you use the skill that nullifies all elements for both sides she literally can't damage you, but you could also use the reflect buff, stack elemental immunity accessories, silence her, etc...) The gist of this approach to game design basically is saying "here's a problem, you deal with it" to the players, it can definitely be done wrong if you're not careful but if you strike that correct balance, it both makes it easier to design the game (since you don't need to come up with tons of unique solutions for puzzles) and treats the player with a respect they can feel.

[@victorpellen3254]

I look forward to spending the better part of the 2020s looking forward to your new upcoming roguelike, whatever it is

[@LaughingThesaurus]

If you do a roguelike, I must strongly, strongly encourage you to study Hades before you do. Hades has a system in place to sort of ensure that you always get some sort of synergy in a run. If you pick a boon from a god, that god will continue to show up throughout the run, and boons from the same god tend to synergize well with each other. Each god sorrrrta has built-in synergies with other gods, but some work better than others. And, well, some builds don't work well with certain weapons, but that's sort of on you to spot. It's basically the only roguelike I've ever really loved, because it doesn't let you screw yourself over, and yet, also gives you a ton of build variety. Please, please study Hades at some point if you haven't already.

[@gloverelaxis]

that's something Slay The Spire gets badly wrong; your options for designing your own synergies are incredibly, incredibly restricted and far too much of your success literally depends upon the luck of the draw

[@gloverelaxis]

ah, literally as I was typing this comment, Tom mentioned Slay The Spire in the video! hah, great minds

[@yubbo3]

it's funny, the strategy I saw some youtuber use for slay the spire is, at every decision ask "does this choice improve my chance of winning?"

[@LionsPhil]

One of things I think nu-roguelikes have largely missed, but old guards like Nethack had (and FTL, as a newer one) is that *the risk you will get in an unwinnable situation is inversely proportional to the time invested*. There's very little you can do against hitting a gnome with a wand of death on level 1, but also you aren't yet invested in that progress or character, so who cares, reroll. But by the time you've spend some hours on it, and got into all the things you've achieved that run, it's not just about having more hitpoints or higher DPS---it's about having a wider array of tools and protections to reduce the chance that you get stuck in a doomed situation. Breadth, not depth.

[@Thattheretim]

This isn't Heat Signature 2! ... ... just kidding. Love you really :D

[@seussusamongoose2905]

Great video. Would love if you raised the volume during or post-recording for future videos, because even at max volume it was very tough to hear you while wearing headphones on the train (normally I can hear podcasts and videos clearly on the train at around half volume)

[@KyriosHeptagrammaton]

I've been thinking exactly the opposite. That you can't achieve true greatness without allowing for the player to fail. It's a risk, and maybe you lose 5% of unlucky players every run, but if they're willing to come back for the next one it can be that much more glorious and interesting as they stretch themselves to overcome.

[@LaughingThesaurus]

I also have to say I just.. really don't like any decision that screws me over for hte rest of the playthrough, especially in permadeath games. I actually find it more painful in roguelikes to fail than I do in most games, because I can't just retry the mission or the boss fight. It'll set me back 20 minutes of progress, and I just get totally fixated on the one specific fight I failed, or the one bad decision I made, with no way to try it again in the short-term. Roguelikes ARE popular, I'm sure they're popular for a reason, but they really don't gel with me for that reason. It MUST be masterfully well-designed and well-balanced at all skill levels in order for me to really enjoy it.

[@pisstit]

so fucking excited for this game

[@mario_actually]

Tom in front of a white wall? Time to click. I’m really that for Rogue Command we can let players run into bad builds. I feel like having an ascension system is really conducive to getting people to a comfortable but challenging difficulty level. I feel like in an RTS plus roguelike there is also so insanely many layers of improving as a player.

[@Psytrese]

Not the point but I 100% would choose catalyst if it offered it as the first card. Later ascensions have an unremovable curse card in your deck from the get go and give more curses more often, and honestly a few dead cards aren't that bad. If I'm going down the route of poison I'd rather not skip possibly my only catalyst. I'd rather lose the run.

[@CodyEthanJordan]

One thing I've never liked about roguelikes is how some of them seem more like a slot machine once you've reached a decent level of skill. Like once you can play competently it seems like there's not as much room for skill expression, instead I'm just hoping to roll some God level synergy which determines if the run works or not. And a lot of those synergies make the game very easy, so there's no challenge on good runs and not much opportunity to bring things back on bad runs.

[@CodyEthanJordan]

That said I love roguelikes, but it seems a hard balance to strike.

[@blake5430]

Interesting

[@SebastienDelfino]

Interesting points but, please, for our sake, check your audio.