matthewmatosis
patreon
Project Betelgeuse Developer Diary 004
🕑 Added 2021-09-01 02:35:13 +0000 UTCResurrection
Proxima needs a rest. I decided to leave it behind after feeling like it needed to be expanded but being unsure how. As soon as my mind was able to relax, I began having good ideas again but it still feels like Proxima needs more time to soak. I’ve moved on to Betelgeuse in the hopes that expanding my skills as a developer is a more valuable use of time than dwelling on a single idea. Best case scenario, some of the work I do on Betelgeuse will make the path forward for Proxima more clear. I’m optimistic that Proxima will rise from the ashes sooner or later, its small scope still makes it very appealing to me.
I’m glad that I never revealed Proxima’s premise since it’s the strangest of the four and will probably undergo significant changes when I return to it. At least now you won’t have to adjust your expectations. Conversely, Betelgeuse is the most straightforward and unlikely to turn out much different from the initial pitch so I’ll take this as an opportunity to be more transparent.
I want to make a simple action platformer. If you don’t know what I mean by that, picture Castlevania (the linear kind) or Ghost ‘n Goblins. Betelgeuse is a game where you walk forward and hit stuff, it’s that simple.
There’s always a temptation to wax lyrical about the fact that games predate art and stories which arguably makes them more integral to our collective psyche. Despite that, it seems as though there’s an inclination for ideas to come in the form of stories, images or other vague concepts. This is the comfortable domain of the Idea Guy. Right now I could tell you all about the world of Betelgeuse. I know what the protagonist looks like. I know the backstory. I know where it takes place. I know what I want the opening scene to be and roughly what happens at the end. Everything is subject to change but my point is that Betelgeuse was primarily conceived as a world I imagined. It crystalized when I pictured what the opening scene might look like and went from there. As usual the Idea Guy is hoping nobody asks “what’s the gameplay like?” Don’t worry, he’ll figure that out later.
While it’s true that Proxima is smaller in scope than Betelguese, this unclear picture of Betelgeuse’s mechanics is another reason why I wanted to try Proxima first. Proxima is the opposite of Betelgeuse in this sense. I still don’t know what Proxima’s protagonist might look like, I’m still unsure where it even takes place. The driving force behind Proxima was a certain mixture of gameplay ideas, the rest was secondary. I imagine this is how most of you would expect Matthewmatosis to design a game and it sounds like a good approach but the reality has been less than picturesque so far. There are a multitude of potential causes for Proxima’s (hopefully temporary) demise. Maybe it was never a good idea in the first place, maybe I lack the technical ability to do it justice or maybe I’m even misguided about abandoning it so soon. There are a million ways to fail and few to succeed. The specific cause doesn’t really matter right now though, what matters is that my mechanically driven idea hasn’t borne fruit. Unfortunate but comes with an upside. If my mechanically driven concept hasn’t worked out yet then maybe it’s not such a problem that my picture of Betelgeuse’s mechanics is still hazy.
All of this is my way of saying that Betelgeuse will probably play similarly to games you already know. This is my main source of trepidation since I don’t want to release a mediocrity but in my more optimistic moments I remind myself that a great game can obviously be made from familiar mechanics. This is probably where beginners should begin. If I’m an expert in anything, it can only be video game analysis essays and my advice to someone entering that field would not be to start with a microvideo compilation or multi-hour commentary. I would advise writing a simple review and building from there, partly because its less likely to end in regret since a straightforward review is an easier target for an unskilled analyst to hit and partly because it’s easier to gauge the quality of your own work when it falls into a category with many existing examples. Of the four projects I outlined before, Betelgeuse is the one which most matches these characteristics.
A conflict as old as architecture, Man versus Castle.
The idea for Betelgeuse is a few years old and was initially inspired more by Castlevania (the linear kind) but having replayed some of the Ghosts ‘n Goblins series in the intervening years (mostly Ghouls ‘n Ghosts for the Mega Drive and Super Ghouls ‘n Ghosts for the SNES) I started to see something in those games which distinguished them as a worthy rival to Castlevania in my mind. (I should clarify that GnG predates Castlevania and even though I played GnG on the Mega Drive many years before I first touched a Castlevania, my true appreciation of each series just happened to form the other way around.) Basically, I started to feel that the Ghosts ‘n Goblins series had important characteristics which help elevate it among the genre and this feeling has intensified immensely with the release of Ghosts ‘n Goblins Resurrection - a game I partly bought as research for Betelgeuse - which has me astonished by its quality. Recency bias might be clouding my vision but right now I’d say it’s one of the best action platformers yet made, maybe even the best full stop. I honestly didn’t expect much out of it - in fact I hesitated to press the buy button on the store page for quite a while, wondering if I’d regret the purchase - but in the end, I found myself playing a truly exceptional entry of the genre I was about to start working on. Needless to say this is inspiring but also intimidating. It feels as though the gauntlet has been thrown and I currently have little hope of turning Betelguese into something which can rival Resurrection. It’s funny that a series notorious for its difficulty has shown up and raised the bar right before I try to vault over it. As a novice I might faceplant before I even get off the ground but the challenge is strangely motivating.
To Be Fair
In my previous diary, I put forward the idea that designers need beliefs so here’s another one of my own: Action games have been converging rather than diverging. Simply by being a proper Ghosts ‘n Goblins game, Resurrection managed to set itself apart from its modern contemporaries because other games and series have followed a more-or-less similar path over the past 20 years.
Resurrection’s greatest strength compared to its predecessors might be its consistency but prior GnG games offered highlights on par with Resurrection nearly 30 years ago, meaning there was no reason we had to wait this long for Resurrection arrive. If you’ve never played Super Ghouls ‘n Ghosts, you owe it to yourself to boot up the first stage and observe what happens. Coffins appear from the ground and zombies emerge from them, in detail it goes like this:
- The coffin appears in a random horizontal position.
- The coffin can either stay on the ground or rise into the air.
- The zombie can either emerge from the coffin immediately or wait a while.
- The zombie is given a random walk speed within a certain range.
Simple but effective.
At a glance it seems simple but it actually leaves players with a lot to consider, all accomplished using nothing more than simple terrain and a single enemy type. No matter how many times you've played before, you can become totally absorbed in the process of staying safe from the very first screen. It’s worth clarifying that difficulty isn’t exactly the appeal here, at least for me. The thing I actually enjoy is being forced to foresee several possible futures in order to take the best course of action. That’s probably correlated quite strongly with difficulty but I think it's a distinction worth making. All I really know is that I find it to be a relentlessly engaging formula which Resurrection leans into even more than its predecessors.
Assuming nothing changes in the coming months, Ghost ‘n Goblins Resurrection will be my favourite game of this year. Last year, that title went to Streets of Rage 4. Now, I haven’t been playing a whole lot of recent games (I often find great things I’ve missed out on years after their initial release so I wouldn’t be surprised if Streets 4 gets retroactively dethroned at some point) but this seems strange nonetheless. Why do I feel like two old school revivals stole the spotlight 30 years after their initial relevance? I think it's because even though old action games like Streets of Rage 2 or Ghouls ‘n Ghosts were flawed in serious ways, they were also good in a way which has been fading. By faithfully adapting these games in the modern era, many flaws were ironed out while the good parts remained intact.
Atsushi Inaba gave a GDC talk a few years ago which has been on my mind ever since. I'm going to take the liberty of paraphrasing his point so forgive me if I butcher or misunderstand his intentions. Put simply it might be phrased like this: Action games are actually reaction games. This observation is then underlined by pointing out that action games are primarily passive because the player is always responding to a situation rather than creating the situation. It can be a surprisingly hard pill to swallow but the more time has passed the more I agree with him. If anything, he probably didn’t go far enough.
Practically every 3D action game I can think of has an instant way to escape danger, often even while in the middle of an ostensibly committed action. Devil May Cry, Bayonetta, God of War (old or new), God Hand, The Wonderful 101, Revengeance, Sekiro, the list goes on. I’d probably include the Ninja Gaiden series too but it’s still on my enormous backlog. Pick any 3D action game and it seems likely to make the list. I’m sure I don’t need to restate my love for some of those games but as far as I’m concerned, 3D action games are not necessarily better than their 2D counterparts. There’s something about 3D perspective which seems to trick players into perceiving added depth, maybe because of the literal depth implied by the graphics. This is not to say that 3D games lack potential - quite the opposite - just that there is likely still room for meaningful improvement. As of writing, the Devil May Cry series just celebrated its 20th anniversary. Two decades of refinement may seem like a lot on a human timescale but it’s nothing for an art form.
We need a positive example so try not to laugh at me comparing Resident Evil 4 to the list above. If this seems too ridiculous, consider the following: Every one of the 3D games I mentioned allows the player to hit stuff directly with their hands (or a sword) but everyone knows the best Ghosts ‘n Goblins weapon is the dagger, therefore an action game game doesn’t necessarily need melee combat, unless you’re going to tell me that Ghosts ‘n Goblins somehow shouldn’t be considered an action game. Genre classifications rarely make for interesting discussion, my point is just that Resident Evil 4 bears at least as much in common with Devil May Cry as it does with Doom. I’d say it’s far more about choosing where to place your shots so as to control enemy patterns than it is about raw aiming skill. Who should I prioritise for a knockdown? Do I need to conserve my shotgun ammo? Is it possible someone is about to flank me? Is now a good time to reload? Can I somehow group them up for a grenade? Here we find the tactical element which seems lacking in modern action games, they might reward this kind of prediction but they rarely require it. Even prediction may not quite be the right word. When a grim reaper emerges from behind a tree, it's only a matter of seconds before he’s sprinting towards Arthur. Enemy behaviour is so clearly defined there’s almost nothing to predict, only simulate. This is what makes Resurrection so good, it’s like a bullet hell where the bullets have complex behaviour and you control a ponderous knight who can barely eke out victory through constant threat assessment.
It's not so different from Ghosts 'n Goblins when you think about it.
I suppose there are several different forms of depth that we might consider. It’s clear that many of the aforementioned games have greater depth than Resurrection when it comes to offense. You certainly won’t be seeing any GnG combo videos but ignore all that and instead imagine you were tasked with creating an AI which could complete these games without taking damage. You have access to all the relevant variables like enemy position and status, in other words you know when an attack is one frame away from hitting the player character. Despite being one of the most recent releases on the list, Devil May Cry 5 is one of the easiest to solve, especially if we’re talking about Bloody Palace. Simply attack until you’re in danger, then instantly activate Royalguard to negate damage. Others like Revengeance and Sekiro will require slightly more awareness about which attack type is coming but will ultimately be solved by pressing a certain button in response to the enemy. Bayonetta and The Wonderful 101 are complicated by a larger number of stage hazards and gimmicks which take extra consideration but many encounters could be solved in a similar fashion. Only God Hand might pose an enduring problem approaching (or exceeding) the complexity of Resurrection, largely due to Gene’s extremely limited range of movement. This isn’t some irrelevant curiosity, these defensive algorithms are running in your brain as you play.
Perhaps I’m being uncharitable about how easy some of those games would be to solve and it's worth noting that the simple solution would not necessarily be fun or look pretty. Nobody wants to play DMC5 by spamming the attack button and hitting Royalguard every single time an attack comes along. That said, think about how clever Arthur’s AI would need to be by comparison. If a grim reaper is running at him, it’s not enough to jump or throw a dagger on the last possible frame. You need to be able to think ahead and position yourself in the safest way to advance. Ghosts ‘n Goblins is simply a more complex game to survive. For me, this quality makes Resurrection extremely engaging and I assume I’m not alone.
Figure it out fast.
If that’s the case, then why did this style of design seem to decline? There are many possible reasons but I think fairness is the chief culprit. When players have an escape hatch (usually in the form of invincibility frames or a parry) they can be sure that everything they encountered was fair which must dramatically reduce frustration. Naturally, nobody wants the player to be frustrated which makes this an appealing design pattern. Resurrection takes a different approach but I would say that all (or almost all) encounters in Resurrection are entirely fair regardless. Here’s an attempt at defining fairness in this specific instance.
- When a player gets hit there was some action(s) in the recent past which would have avoided that damage.
- The correct action(s) would have placed the player in a similarly survivable position, albeit one which might need further quick/clever responses to facilitate continued survival.
- The player had all of the necessary information to make this decision before the inflection point. In other words, given a screenshot of the critical juncture, it is possible to plot a course of action which will survive with absolute certainty.
Hopefully we’re in agreement that such a game would count as fair but this doesn’t address player frustration. Without an instant response to danger, there’s some window of time (possibly several seconds) during which the player technically hasn’t taken damage but their fate is sealed because the escape routes have closed. Note that this is also true of Resident Evil 4 and many older action games such as Castlevania. The goal of Ghosts ‘n Goblins is to avoid such pincer situations but players may not recognise where they made the mistake since there’s a delay between failure and punishment. Superficially, this can make it seem like the punishment happened for no reason thus a player might blame the game rather than thinking back to the prior moment when they should have foreseen this outcome.
I don’t know if this problem with delayed punishment has been written about before but I think it was intuitively felt by designers many years ago who then built their games to compensate. We are now possibly in a viscous cycle of “The Player is Always Right” where developers seek to mitigate frustration by providing proof that that the encounter was fair (in the form of liberal access to invincibility frames) which then causes players to become even less tolerant of perceived unfairness in other games.
Even disregarding how a hypothetical player might feel, you can see why this would be a tempting approach for designers. If you’re making a traditionally fun action game, you probably want to know it’s fair, just for your own peace of mind. In a game like Resurrection, it can be hard to tell when it crosses such a line. Unfortunately I think it does so during a particularly bad autoscroller in the second loop version of the third stage which feels more like a crapshoot than a test of skill but that means you can complete an entire playthrough without any unfair encounters. Maybe I’m being uncharitable, but I assume many players decried the game as being unfair long before they saw the credits. Who knows, I might also be wrong and eventually change my mind about the second loop too. A wide range of opinion is bound to arise because there’s no foolproof, objective way to determine fairness in Resurrection. There is no parry or invincible dodge to be pulled off at the last possible moment.
If you’ve seen just how pedantic I can be in my Devil Daggers or Return of the Obra Dinn reviews then you’ll know how uncomfortable I find this kind of fuzzy fairness. It’s simply not possible to prove - in a logically coherent way - that Resurrection is fair. The designers merely felt it was fair and so do I... for the most part. Streets of Rage 4 is relevant here again because it walks a nice line between both styles of fairness. The special moves are technically an instant escape button but they put the player at risk of losing extra health afterward. This approach allows for certainty of fairness but still encourages players to control crowds in a smart way by using throws or knockdowns effectively. It still has that old school quality of spacing and pre-empting bad outcomes.
Special attacks grant invincibility but leave the player more vulnerable for a time.
I don’t mean to imply that 3D action games are devoid of this satisfaction, you can find examples where good positioning will pay off in many of them such as a well placed Wicked Weave which catches several opponents at once. It might be worth giving special consideration to the Monster Hunter series here since positioning plays such a crucial role and invincibility frames are less abusable due to the size of most monsters. It’s true that these games require a cautious approach to combat but it’s also true that they give up a lot to achieve this result, especially compared to 2D games. Encounters are reduced to a single enemy which the player keeps in their sights by babysitting an otherwise static camera. Suffice to say this might qualify as the old school, fuzzy fairness approach but the tactical layer seems lacking compared to the multitude of overlapping enemy behaviours which need to be accounted for in a game like Resurrection.
There’s probably a dissertation’s worth of material to write about why 3D games seem to tend towards the other style of design. My feelings on this are hazy but here’s three possibilities, all of which might be true to some degree:
- Dynamic camera systems make it more difficult to precisely position yourself in 3D space.
- 3D games are still most often rendered to a 2D screen which again makes precise positioning more difficult than ideally necessary.
- 3D games tend to have larger budgets which puts pressure on designers to reduce frustration since they need more sales to break even.
I’ll leave my speculation there for now. The important point is that a notable difference exists in how some games choose to craft fairness, whether 2D or 3D. As you can probably tell, this idea of fuzzy fairness relates back to Betelgeuse which is why it’s important for me right now. The way I see it, in order for Betelgeuse to be the best game it can possibly be, I need to let go of any temptation to craft perfect fairness. Of course I want it to be fair and will strive to ensure it is but I won’t be able to guarantee it the way some games can. The vague picture I have of Betelgeuse is one where you cautiously progress forward and assess each threat coming your way. Given my love for fast-paced action games and my fixation on fairness, I think this will be a difficult mentality to adopt. It also probably doesn’t bode well for critical or commercial success. This post has more or less taken for granted that Resurrection is something to aspire towards but it sits at a 75 on Metacritic (bordering on abysmal if we adjust for how games are typically scored) and unfortunately it seems not to have found much of an audience either.
Creativity is strange. I’ve given up on Proxima for the time being because I couldn’t quite sculpt it into something I was happy with. Despite that, there is some blurry, “perfect” version of it in my head still, something I feel could plausibly exist even if I can’t quite make the pieces fall into place yet. The same is true for Betelgeuse, Procyon and Sirius. All other things being equal, I think my ideal versions of Proxima and Procyon would be better received than my ideal versions of Betelgeuse and Sirius. I’m not convinced anyone actually wants this game but I want to make it so I’ll try to do it justice.
Comments
gibbdude
Great take
wondermagenta
it's funny you single out ninja gaiden as probably very similar to those other games, because it actually stands out quite a bit. you hold the left trigger to enter your guard stance, tilting the left stick during will perform a quick step (basically a short distance i-frame dodge.) however, you can't actually truly cancel the recovery of your attacks with either of these like in most other action games. the one safety net is that, if you hold L2 while in recovery, ryu will enter the guard state the exact instance he's attacked. i wouldn't say it's really comparable to ghosts n goblins, but blocking always carries the risk of having your guard broken or being thrown, so it does end up feeling more committal than other action games.
Interesting update, Matthew. I am frustrated by the negative reception of games that don't provide some sort of free pass for all of a player's bad decisions. Out of boredom, I've stopped playing games that give ample access to invincibility frames.
I want this game.
Blooper
Great update, I appreciate a writeup like this a lot. I hadn't played many older beat em ups before I decided to play Streets of Rage 1 and 2(hey, maybe 3 someday) before 4 came out. Coming from modern action games, this focus on prediction and positioning felt completely unintuitive. I usually imagine fights(in games naturally) playing out like fighting games - you get up in the other guy's face, you block or dodge his moves and swing back with whatever you've got. Fighting games are mostly 1 on 1, but that goes for stuff like Dark Souls or DMC too. Punch-out is an old game that works like that and was completely intuitive in that sense. Meanwhile, with Streets of Rage, I really did have the slow realization of enemies being "bullets" like the metaphor about GnG being a bullet hell. It works fine as gameplay, but it doesn't exactly capture the Idea of how I think a martial arts fight is gonna work intuitively, and I much prefer enemies that act like you're "actually" fighting them - at least when they're shaped like humans. The patterns the enemies in SoR 1 and 2 took often seemed completely arbitrary to me - for instance there's nothing about a Donovan that says he'll anti-air you. You just learn it by being anti-aired. You're also very limited in your approach as a player. You can throw in different ways and all, but there's not a lot in there to experiment with and play with like in many modern action games, like you mention. And the approach to enemies is really counter-intuitive, at least to me. It made no sense that the right way to fight many of these goons was to abuse the "3d" space to walk into them from above and below them since they have few attacks that hit vertically. However, Streets of Rage 4 worked a lot better for me. The controls, on the whole, felt better. There was a larger variety of moves at your disposal, so you weren't seeing the same attacks again and again for ages, and could make a lot of fun little combos. You could save your game(without any savestates) and continue without starting over from the start. The health vanishing when you used special moves meant I rarely used them in SoR2. But in SoR4, because you regain green health when you hit an enemy, I used it a lot more liberally since it felt more fair. Even poison only takes away health you can regain, so for once I didn't just loathe poison in a game. It's less punishing by a mile. There were a lot of quality of life changes like that which added up to make it a lot more fun and a lot more fair without fundamentally changing it from the original Streets of Rage games, so while it hardly captures the Idea of fighting any more than those old ones, I like it much more. I think it could be better - a more interesting story and context(maybe some bad guys that weren't so untreathening) and different routes to take so a campaign playthrough had more variety. If you won't let me run as every character in a fight, at least let me run inbetween encounter so I don't have to snail my way over there. Most importantly, I really wish enemies had animations that telegraphed their attacks better. I don't think an enemy blinking and heading your way accurately conveys that it wants to grab you, compared to a model in a 3d action game coming at you with its arms out. The Darkroot Garden plant people make it very clear what their intentions are compared to the green cops or biker ladies in SoR4. But still, it's a vast improvement that I want to see even further iteration on. I dunno, writing this out, much of it feels like I just wrote "same" to a lot of what you said in a longwinded way, only I don't appreciate the old position-based approach as much. But at least I want to say, even if you make a game in that style, I think you'll avoid a lot of frustration just by coming at it from a modern perspective. Telegraph the hazards and attacks well, keep the punishments reasonable, all that stuff. I dunno if you're married to a fixed jump arc like some of these action platformers, but being able to adjust it more freely on the fly would probably help a lot too. It might be more "reactive", but when Mario or Alucard for that matter could do it it's hard to call it a modern trend. But regardless, as long as you land more on the SoR4 side than the SoR2 side, it's all good as far as I'm concerned. And I appreciate the opportunity to be able to write about SoR4 for a bit, lol. I bet it's seen a lot more success compared to Resurrection but it's still not a game I get to talk to others about much.
SteveReen
As an aside, I kind of dismissed Resurrection when I initially saw it but now I'm feeling encouraged to check it out. I've never gotten into the series before, but even if just to support such a game being made it seems worth my time/money haha. The whole section here about delayed punishment did a great job articulating a sensation I have at times. I tend to get caught up in that sort of frustration when those situations occur; in the moment I think "there didn't seem to be any way out of that?!" and sometimes a minute later or so I'll think back on it like "maybe I could have done something differently before then" but as you mention it's never easy to come to a definite conclusion. Thinking about this analysis might help me discipline myself and be more patient when those situations happen in the future.
I'm glad you mentioned delayed punishment in games. I recall playing GnG Resurrection and getting frustrated when I was cornered and killed. To me it seemed unfair because it made me fell helpless, like I had no control over my own fate. I guess I didn't realize what kind of game it was. I still enjoy it quite a bit either way.
My thoughts exactly. It certainly gave me something to think about!
Viridian Maridian
The “The Player is Always Right” section really resonated with me. I tried to get my buddy into Dark Souls and near the beginning he kept getting frustrated that his character was rolling backwards randomly. He thought it was a bug. Turned out he had a curved sword and was "kicking" accidentally. I told him to go and find a safe space, practice all of his moves and figure it out. Instead he just kept complaining at me that this "wasn't fun" for him and that I needed to tell him the answer so he can have fun again. It was just mind boggling how conditioned he was by other games to think that everything should be given to him on a silver platter. Never mind that The Asylum teaches you the "kick" move anyway, but how is it not player second nature to just press all the buttons and try to figure out your moveset? This line of thinking extended to a lot of the game, and it was clear he just wasn't going to put in the energy to figure it out by himself. Whether it's finding the next objective or overcoming a tough enemy encounter, we're seeing games increasingly focus on giving the player a million helping hands so they feel like they did something special, and fewer genuine challenges to carefully consider and conquer.
SteveReen
There is basically enough thoughtful commentary here to fill one of your videos, haha. Really enjoyed reading this. The recent update videos have been really cool too, and I've also enjoyed seeing the occasional streams on the side. Keep it up!
Incaptivity
Stellar update, given me lots to think about in regards to how some games are made/designed while also explaining yer own games. Look forward to all future endeavors
Matthewmatosis
I think the recent state machine videos show how the project is going so far. Still have a few more features to bolt onto it but I'm hopeful once it's finished I can repurpose it for enemies as well. If nothing else it gives me a lot of finely detailed control over the player character.
