random musings on rpg and strategy game design

warning: text ahead

I’ve been burying myself in a lot of RPGs/Strategy games from the mid-late 90s and early 00s lately; partially out of wanting to unbacklog some old things, partially wanting to stay to older works in general, and partially looking for games that did something interesting/unique instead of trying to be The Best X That Ever Was. Sitting with the ability to read Japanese has opened up a lot of options that weren’t there whenever I was in the mood for this in the past. (see Dimguil thread. also ran the original Oreshika, remind me to say things on that game sometime! it’s super weird)

And, well, it’s probably obvious that I’m not a fan of a lot of the games in those genres that I’ve tried in recent years. Like, XCOM is a game I wish I liked, but generally find boring rather than interesting, and I’ve wanted to come to terms with why I feel this way in a more specific sense than I already had.

Mostly, in the end, I feel like it comes down towards how it approaches two related aspects of the player’s choices: The savegame method, and how to approach hidden information.

Many games developed in the lineage coming from desktop computers generally relied on save-anywhere load-anywhere as a main feature, and were largely designed around this being a main option. The old Ultima series was a big offender here, essentially making the tedium of dealing with a dead party member too much to bear for most of the game, but I think it extends to newer games like the aforementioned modern XCOM as well. But where it’s just tedium with one, it’s something more frustrating with the other.

Hidden information is generally approached as a way of giving you things to learn that are not directly the mechanics that are in front of you. For example, you can define this as what the stats and capabilities of enemies are. But also, in terms of strategy games, where they spawn, what their AI is, and so forth. Basically it’s what the game expresses to learn the game beyond the game and modify your gameplan to match.

But, far too often, this is used as a replacement for interesting gameplay. You lost! Because you didn’t know thing XYZ, and now you do, and now you can play again. But because the consequences of not knowing what just happened has had lasting effects on your game, you have a choice between reloading a savegame, or trying to play out a game that is largely designed around you reloading your savegame when something you don’t know about kills you.

These games often can be played Ironman-style, and there’s often an encouragement to do so, but it’s not really designed into the game as something you should do as a primary way of playing, and I believe it shows as, outside of really wanting to prove something, most players will value their own time consumption over the game’s willingness to throw hidden curveballs at the player’s face.

I always tended to have a stronger interest in simulation games because they tended to be less about concrete paths and goals to follow and more about trying to make do with what you have, even in the face of loss. Is it less or more stressful or challenging? I don’t know, but I wanted to answer these questions. Usually the failure state is a little fuzzier and more “you spend more time” than direct loss, until you hit the full ‘welp, starting over’ mark. I think the original X-Com falls into this camp pretty well, aside from getting your base invaded, you have a lot of flexibility and ability to rebound.

One reason I’ve been spending a lot of time with Wizardry-likes is that the original game was designed around autosaved gameplay. Back in 1981, even! That’s some absolutely amazing stuff to think about, because it’s almost unheard of even today. No reloading when things go south, you take the hand you’re dealt.

In it, there is no such thing as an unwinnable game, because status of the game’s progression is stored either globally for all characters, or individually on each character. For example, opening the second elevator in the first Wizardry is permanent and global, but the defeat of the last boss counts on a per character basis and you will need new characters if you want to re-challenge. Later games expanded on this concept (Dimguil included!), leading to the sort of strange system where every party is effectively its own timeline.

While it’s not really the perfect implementation, I also think it’s incredibly forward-thinking for its time and more games could see fit to emulate the idea of the true failure state being the player’s unwillingness to continue, rather than the game deciding that the game is over for you. At least, that’s my opinion.

this is just late night ramblings brought on by way too much thinking, don’t mind me

8 Likes

Additional thought: The approach to hidden information where you spend a lot of time learning about composite situations to deal with and reflecting about them after the fact has a lot in common with competitive games. But competitive games are usually five to thirty minute affairs, not 30-60 hour games, and savescumming reality as opposed to predictable AI isn’t really possible.

This is why I tend to favor games where finding the hidden information is itself the pleasure. Implementation/tactics/micromanagement comes with its own rewards, but in general I fall on the side of the spectrum that wants a Tim Rogers ABDN Dead Space Review “I get it” button.

I really rambled aimlessly last night, this is kind of hard to read back ahaha.

I think hidden information is a completely valid thing to have, but too often it’s tied in with the save system as a way of going “you didn’t know that? well, too bad, you failed, and now you get to reload and try again.” Which I’ve really come to detest from a game design aspect these days, I don’t enjoy it and would rather hidden information be more integrated as part of the overall design.

I’ve been grappling with this with roguelikes a lot lately. A really big fan of Shiren’s simplistic enemy and map design meaning you can figure out what things do very quickly and adapt your gameplan as you go, even if sometimes you are better off having planned for it 20 floors ago you can still make do usually.

Well, I’m speaking of “hidden information” in a much vaguer sense that applies, to, say, adventure games, which are 100% about hidden information (i.e. divining the designer’s intent) and have basically no implementation requirements or barriers.

But it also applies to like Souls games. I’m the guy who enjoys the fact that Demon’s and Dark had opaque stat systems represented by nearly-identical icons and bizarre upgrade weapon upgrade systems that would almost certainly result in a lot of wasted effort if you didn’t FAQ them - because the thing I value about those games is the mystery they evoke. It certainly helps that there is no permanent failure, you can always play more and grind more and get more stuff if you really waste something.

Is that analogous to save/reload? Sort of I guess. But if you’re militant enough about “any loss is recoverable without repeating content” you end up with Prince of Persia unfailable jumps, right? Like you’re always talking about some point in the middle of a spectrum here.

The ideal here is Crusader Kings 2 right? Its got the right balance of hidden info and recoverable game states. As far as I can tell, even with no knowledge of the systems it is ridiculously difficult to actually get a game over, and consequently the smaller failures as you play the game are actually entertaining and memorable. Like, I dont actually find losing at modern xcom to be a really engaging experience (and we must always remember one of the earliest rules of rpg design in this discussion: roleplaying happens when we are no longer playing to win)

Paradox games are great at this, sure, but their strengths aren’t readily transferrable to other genres. Part of the reason there’s no real fail state is that there’s no real win state, which is true of many other simulations (think, I dunno, SC2K, or MS Flight/Space Simulator). I mean, how would you apply CK2 methodology to like a dungeon crawler? I think if you could answer that question on any but the vaguest levels you’d be an award-winning game designer.

Like I stress, hidden information is completely okay! It’s how it interacts with savegames that gets to me. Obviously something like a puzzle game is going to be entirely about attempting to hide the solution from the player, something I’m well acquainted with.

I think the point I’m trying to get across here is that I’ve come to believe that manual savegames are a bad idea to design your game around. They can exist, but by having them you’re essentially leaving the pacing and flow of the game up to the player. Which isn’t wrong but, I think, kind of shallow. And I think a lot of games stemming from desktop computer lineage are designed heavily around the idea that you are not going to know things and we’re going to mercilessly kill you with them and you know that last several hours you spent not saving because everything was trivial? … Yeah, that’s not the kind of design I have much patience for anymore.

I don’t mind fail states, personally speaking. But there’s a big difference in a fail state for an hour-long game and something you’re expected to spend 30 or more hours in, and this also applies to soft-failures where you haven’t technically lost but there’s no coming back from the resource deficit spiral. It’s more about how much of the player’s time you are wasting with your failure states, to me. Is this going to cost you ten hours of your time or more for one mistake? I don’t think that’s terribly compelling, and just encourages grind/repetition instead of spreading more interesting decisions around.

Paradox games are interesting but I can’t get into them because they’re too far on the simulation side of things. It’s grand machinery, beyond ability to play any but a bit part in. And that’s okay, but it’s hard for me to grapple with.

also hi i played competitive games and roguelikes for years and now i kind of want everything to play like them

1 Like

what real time strategy games are actually worth playing

I own anno 2070

rise of nations

I’ll take up this challenge just because I am perhaps uniquely suited to answering that question. What would you consider to be the CK2 Methodology?

mega-lo mania

I think modern game designers realize the overwhelming importance of checkpoint design to games’ pacing, but in the 80s and 90s it seemed to just come as a side effect of platform capabilities/limitations. Consoles had slow+small SRAMs, whereas PCs had fast+huge hard drives as well as available keyboard keys to tie quicksave to. Then the game designs flowed naturally out of that (and I agree with you that the more constrained saves encouraged better game designs).

For example, one reason Halo was a positive step forward for single-player FPSes is that it replaces quicksave with autosave, I’m guessing mainly as a concession to the lack of space for quicksave key on the controller. Then, it took until Destiny for checkpointing to truly be reified as a top-level concept in the game design (via “Darkness zones”).

I definitely agree that it’s hella frustrating to lose literal hours of progress based on a single mistake made hours ago, but it’s hard for me to identify any game where that’s the case. I guess the closest would be 4X games? Which do have problems with snowballing and comeback mechanics, but also work based on showing you the cumulative effects of dozens or hundreds of decisions you make over the course of the game (rather than dooming you because of a single mistake, unless by “single mistake” you mean an aggregate strategic mistake that actually incorporates many individual game-interactions, like gunning for one victory condition when you should have been gunning for a different one).

As for quicksave/quickload spam I don’t necessarily agree that game design has responded with a lot of sensitivity to the new prevalence of checkpointing. Would HL2 be significantly different with checkpoints? It already has built-in lulls between action sequences. All quicksave spamming does is permit you to force your way through the game at a higher difficulty level than otherwise possible, which is a dumb decision for a player to make more than a design flaw.

Uhhhhhh? Let’s see.

Maintenance of multiple data-heavy systems that interrelate at many points, meaning a shift in one variable has a large ripple effect among many other variables. Perfect information clarity. A large number of other actors who have the exact same capabilities and follow the exact same rules, which contributes to the “the world doesn’t revolve around me” feeling. Multifarious effects of actions encouraging player-defined success and failure. Redundant systems to prevent total failure on the one end or runaway dominance on the other.

4x games are probably the biggest culprit, yes. They’re a genre I like but they’re just really something where you get it right because you don’t get many more chances.

And yeah, checkpointing hasn’t really changed much other than giving you Designated Save Locations. You still kind of hurl yourself at situations until you go “okay, I get it” and move forward. It’s not very dynamic. Even then, I still prefer having designed checkpoints/save points to letting the player choose. (IWBTG’s shootable save boxes are entertaining)

That’s kind of what I mean though, because especially the RPG genre is based on letting you reload whenever, because Unintended Consequences usually lead to pretty swift death and they’re not designed to handle it gracefully more often than not.

I don’t know where I’m going with this, but I do know I’d rather the designer decide how the save/reload/checkpoint/whatever loops work than for me to choose myself, and to build the game around knowing that I won’t be save scumming constantly.

Thinking more I don’t think 4X games are really guilty of this. Firstly because while one playthrough is certainly longer than, like, a fighting game match or whatever, you’re talking about a handful of hours. 4Xs are designed to be played multiple times under multiple start conditions. You’re supposed to analyze the effects of your decisions, learn from your failures, and start again stronger - the game is explicitly built this way. You play Azetcs on Monarch on Continents and then play English on Prince on Pangaea, or whatever. And, like I said, it’s never a single mechanical decision (I choose this tech instead of this tech) that dooms you, it’s the cumulative effect of lots of mechanical decisions in the service of what turns out to be a bad strategic decision (I am resolved to war with this weak AI on my border instead of the stronger runaway AI). Losing in this way is fun, not frustrating, at least if you’re going into the 4X game with the proper mindset. They’re not classic singleplayer games with storylines that you play once through and that’s it.

I think your criticism applies more acutely to western RPGs of the Baldur’s Gate variety, as you say. Lots of old manuals had that “keep multiple saves!” admonition. Because they do present a single complex story, and are extremely long, restarting is losing progress, unlike in a 4X where restarting is progress. I’m trying to think of my favorite WRPGs - Fallout, say - and figure out if you could really bone yourself to the extent you’d have to restart the game if you were only keeping a single save. I guess it’d be pretty simple. But how would you propose improving Fallout, specifically, with a different save system? What other systems would you have to alter to accommodate a more designer-controlled save, and would that make for a superior game/experience?

Hmm… Save after every conversation with an npc and after every battle. It’s important that these saves are after, so you have to live with the consequences of what happens. No consequences are so harsh in these games that they lessen your ability to eventually win.

I don’t think anything else will change. Just, preventing the ability to save-scum an optimal path through a dialogue tree does wonders at rehabilitating the dialogue tree format. Similar to how the dialogs in planescape torment were so refreshing because they were so baroque, convoluted and interconnected that it was near impossible to map out an optimal path on a first playthrough anyway.

If Planescape is the pinnacle of the Black Isle dialogue tree, Alpha Protocol is the pinnacle of the Bioware dialogue wheel. Not only does the game permasave when you conversate, the conversations happen under time pressure and offer you only attitudes, not your actual words. You’re forced to think on your toes and live with the consequences, it’s great.

3 Likes

meant to respond a little faster but the last couple days have left me devoid of sufficient brains

I think 4x games are at least moderately guilty of it, but yeah it’s more integrated with the design since you are supposed to play them over and over again. There are ones that are pretty bad about it; the Heroes of Might and Magic series comes to mind here. it’s like the proto-moba except much more drawn out. Like I said earlier, simulations tend to be softer and more forgiving of short-term failure.

I’ve kind of said what I wanted to say here though. Need to put more consideration about how to put things into practice…