Share
submit to reddit
Pin it

I don't know if I'm going to be buying a PS5 anytime soon (or ever), so I may not have an opportunity to play Miles Morales, which is Insomniac's follow-up to its smash hit 2018 game Marvel's Spider-Man. Instead, I decided to go back and play the DLC for the 2018 game, "The City That Never Sleeps", which I had bought, but never got around to playing. This got me thinking more about how Insomniac implemented the web-swinging mechanics, and what I hope they'll do to iterate and improve the mechanic in future games.

This blog post is a transcript of the video essay above.

I had previously mentioned Marvel's Spider-Man essay about open world gaming's possible recent inflection point, but I didn't go into much detail. Basically, I just threw it in as an example of a recent open world game in which the traversal of the map had a large mechanical focus, turning the map into more of a play space and less of a convoluted, time-wasting mission-select screen. I didn't spend more time talking about Spider-Man, however, because as much as I like Insomniac's game, and as fun as the web-swinging is, I still felt like the web-swinging traversal in that game was pretty simple, and the environment did not act as much of an obstacle to the level of the other games in those videos.

Besides, Insomniac's Spider-Man didn't stray very far from the boring checklist-inspired open world design that my earlier videos were railing against. Traversal doesn't consume resources other than the player's time, and the player isn't responsible for balancing Peter's heroic and personal lives (the tension between the two has always been a big part of the Spider-Man story), nor are there any other mechanics that try to pull the player towards one set of content to the exclusion of another, and so where you are on the map, where you're going, and how you chose to get there is largely meaningless. It's any other open world game you've played in the last 10 years. The web-swinging is just a much more stylish and spectacular method of moving from filler content to filler content.

Spider-Man games make for an interesting case study in open world game design.

Spider-Man games in total do represent their own interesting microcosm of the virtues of open world, sandbox game design, and also of the ways in which open world games can fall flat on their faces and fail miserably. Spider-Man could make for an interesting case study to go along with Death Stranding.

Since the landmark Spider-Man 2 movie tie-in game, mainstream Spider-Man games have mostly been open world games. The most notable exceptions being Beenox's Shattered Dimensions (which was pretty good) and Edge of Time (which was awful). All the other Spidey games that I've played have been open world games in which you web swing around a virtual Manhattan to reach story missions or to thwart ambient crimes. The quality of these games has been very hit-or-miss, but (as we'll discuss soon) many of them still have their unique merits.

Beenox's Shattered Dimensions and Edge of Time are notable Spider-Man games that are not open world games.

Web of Shadows, for instance, had lots of problems with its writing, pacing, and animation, but its novel aerial and wall-crawling combat mechanics made excellent use of the map's verticality in ways that other Spidey games (and open world games in general) rarely even approach. I won't be talking much about this game because I traded in my copy a long time ago, and I don't feel like blowing $45 of Patreon contributions on a game that I'll likely play for 10 minutes just to capture footage and refresh my memory of how the game played. So my apologies if you're a big Web of Shadows fan.

The new standard-bearer

Sony's Insomniac's Marvel's Spider-Man has been anointed by the gaming public as the "definitive Spider-Man" game in most circles. Reviews, videos, and analyses have almost exclusively praised the game for its fluid, acrobatic combat, and for having the best open-world web-swinging of any Spidey game ever (including Spider-Man 2). I cannot say that those critics are wrong. The game is great! I loved it! If you think that Insomniac's Spider-Man is the best Spidey game ever (or even the best comic book game ever), I certainly won't argue with you. It is to Spider-Man games what Arkham Asylum was to Batman games: a love letter to the source material.

Insomniac took the genre-defining web-swinging mechanic from the landmark Spider-Man 2 and modernized it for the HD generation. They did a fantastic job at that. But in simplifying the mechanics and ironing out a lot of the rough edges, they also removed a lot of challenge. By looking primarily at Spider-Man 2, they may also have missed (or ignored) opportunities to borrow and iterate upon some of the good ideas that other developers have had. In fact, if Insomniac was going to look to a past Spider-Man game for web-swinging inspiration, I actually wish that they had looked to Beenox's open world web-swinging concept for the Amazing Spider-Man 2 movie tie-in.

Beenox's Amazing Spider-Man 2 did some things that I wish Insomniac's Spider-Man had replicated.

Now, before you close this video in disgust or go into the comments to tell me I'm crazy for implying that Beenox's Spider-Man is better than Insomniac's Spider-Man, please hear me out and let me qualify and clarify that statement.

Amazing Spider-Man 2 is not a very good game, and doesn't even approach Insomniac's game in terms of quality or playability. Amazing 2 is very clunky, its controls and physics are imprecise, it has a sloppy and unfocused story, ugly visual design, and has generally poor mission structures. The web-swinging, specifically, is very un-polished. The controls for web-swinging are as imprecise and clunky as the rest of the game, and the physics blatantly cheats by not actually requiring webs to attach to buildings.

Beenox's first open world Amazing Spider-Man game allowed Spidey to fly.

Now, to be fair to Beenox, this sequel was a huge step up from the first Amazing Spider-Man movie tie-in from a mechanical standpoint. That first game allowed Spidey to literally web-swing from thin air, well above the skyline. He could even point-zip into open air, which was necessary for collecting many of the game's hidden collectible comic books. So credit to Beenox for making an effort to iterate and improve on its previous effort, even if the end result was still very rough around the edges.

So if the controls and physics sucked, why do I think that Beenox's web-swinging system, from critical and commercial failure of a game, should should have inspired Insomniac's game. Even thought the mechanics are under-developed and lack polish, I actually think that the concepts behind the design are solid and have a lot of potential -- given the right developer, a healthy budget, and enough time to play-test and polish. Maybe Beenox could have achieved a more satisfying experience that could have held a candle to 2004's Spider-Man 2, had they not been constrained by the strict deadline of meeting a tie-in movie's release date? Or if they had an opportunity to iterate on the ideas further? Unfortunately for Beenox, they never had that chance, and we'll never know.

Gravity and geography are no obstacle to Insomniac

Before I get into what I liked about Amazing 2's web-swinging, let's first take a look at how web-swinging works in Insomniac's game, and try to identify some of its design weaknesses and limitations.

Web-swinging in Insomniac's Marvel's Spider-Man is pretty simplistic: you hold R2 to swing. At the most reductionist level, the web-swinging can be accurately described as "hold R2 to go to place." You can get by simply holding R2 and pointing in the direction of the objective marker. It isn't pretty, but it gets you there.

In actual practice, you're probably releasing R2 or pressing X to jump, and you're rarely ever following a straight line like this, so web-swinging is actually more involved than simply holding a button. This isn't The Witcher 3 or Red Dead Redemption 2, in which you can hold the button to let the horse walk to your destination with no further input from the user. Nor is web-swinging as passive as letting an NPC literally drive you somewhere, like in Final Fantasy XV. Though if you watched my video on the inflection point of big-budget open world games, you'll know that I feel like Red Dead Redemption II and Final Fantasy XV have some other ludic considerations going for it.

Games like Red Dead Redemption II and Final Fantasy XV have completely automated travel.

Regardless, the skyscraper jungle of Manhattan island simply isn't that much of an obstacle in Insomniac's game. If you run into a building, Spidey will automatically transition into a wall-run with almost no loss of speed or other penalty.

Gravity also isn't much of an obstacle. Spidey can fall from any height without taking fall damage. He can even web-zip to the ground from free falling at critical velocity, and somehow that negates the impact...?

Once connected to an anchor, the player also has tremendous control over Spidey during the swing. He can effectively shift his weight to defy the force of gravity by sheer force of will to allow him to swing in any direction, no matter where gravity was taking him or where the web is anchored. On the one hand, you can easily make your screw-ups look like they were all part of the plan, but it also means that there's virtually no consequence for those screw-ups.

Put simply, the web-swinging (while easy to understand, exceptionally well-animated, and fun to play) is not particularly challenging in Insomniac's Spider-Man. Oh, sure, there are optional challenges that test your web-swinging mastery, and it can be quite tough to get the high scores. But I've found them to be more a test of memorizing the locations of the target gates than of actually executing the maneuvers, and those skill challenges require a lot more zipping and point-launching than actual swinging.

Insonmiac borrowed the combat mechanics of Batman: Arkham Asylum, but made it much more challenging to play.

Overall the web-swinging system is relatively easy and exceptionally forgiving. This is especially true when you compare the web-swinging and locomotion against the combat mechanics. Insomniac took the basic ideas introduced by Batman: Arkham Asylum, gave it a Spider-Man skin, and made everything much tougher to successfully execute. Even with Spidey jumping and web-zipping around the combat arenas, timing and positioning is much more important than in the Batman games. Even though it would look a lot more believable with Spider-Man, Spidey isn't warp-jumping across the arena to jump kick thugs in the opposite corner of the room like Batman often does.

GTA not only requires driving around buildings,
but you must also avoid traffic and pedestrians.

Compare navigating Manhattan in Marvel's Spider-Man against navigating the map of something like Grand Theft Auto, which not only requires that you steer around buildings, but also requires that you avoid hitting other cars and pedestrians. And it certainly isn't as trecherous as Death Stranding, which requires you to make very delicate and deliberate inputs, often while holding multiple buttons, in order to keep the character from tripping over slippery moss and dropping his Amazon delivery.

I personally believe that all open world games should -- to some degree -- put a mechanical emphasis on traversing the environment. Insomniac's web-swinging does a passable job in this regard, but I think it falls a bit short of true greatness. The skill floor is very low, and the skill ceiling is not particularly high.

Credit to Beenox

So what did Beenox do -- or try to do -- that impressed me so much?

Well, most importantly, Beenox's attempt requires more active participation from the player. The core conceit of Beenox's web-swinging mechanic in Amazing Spider-Man 2 is that you can shoot webs from Spider-Man's left or right hand using the left or right trigger, respectively. You have to press a button for each and every swing. In principle, this gives the player more control over where you place a web line, which should mean that the game's mechanics should require more careful web placement from the player.

If you can look past the poor physics, robotic animation, and clumsy controls, the idea here is sound. It has the potential of increasing the skill ceiling of web-swinging for those who might want a more challenging traversal experience. In principle, this design should require the player to think more about the physics of each swing, require the player to make more deliberate inputs, allow the player more precision in where you place your webs, and could create more challenging environmental obstacles for the player to creatively overcome. For example, if you're coming to an intersection, and you want to turn left, then you would need to shoot a web to an anchor on the left and let it swing you around the corner.

Beenox's Amazing Spider-Man 2 mapped Spidey's left and right web shooters
to the left and right triggers, respectively

Unfortunately, that isn't the way it worked out in Beenox's final game. The poor physics and loose controls completely gutted the whole concept right out of the gate. In practice, Beenox's game gives Spider-Man even more ability to change direction mid-air than Insomniac's, while still somehow managing to provide less control and precision than Insomniac's game.

How web-swinging could work

I think that Insomniac should have taken this 2-handed approach that Beenox tried and then polish and fine-tune that concept, instead of simply porting and updating the 2004 Spider-Man 2 game. I think that Insomniac should have used the increased power of the newer console to make the landscape around Spidey more relevant, and to make the placement of web lines more strongly influence where you go.

Let's look at a specific situation to see how this design would change how you play the game. Think about the example of swinging along the waterfront. There are only buildings on one side of the street, and the other side is open air above open water, with some relatively low lamp-posts and trees. In every Spider-Man game that I've ever played, this location can be easily traversed by simply shooting webs to the buildings on the left, and holding the left stick a little to the right so that Spidey swings and jumps out away from the buildings.

Map areas like this are far more trivial to traverse than they should be.

In reality, however, if Spider-Man were to try to web-swing down this street by repeatedly shooting webs to the buildings on the left, the force of gravity would constantly pull him down, while the web that is anchored to the building will pull him closer to the building. With enough starting distance from the buildings, and enough forward speed, he could maybe get off 2 or 3 short swings before gravity and the web lines would pull him into the buildings. No amount of shifting his weight mid-swing would stop this. It is inevitable.

In the games, however, Spidey is able to resist the force of gravity by sheer force of will. You can push right on the stick and Spidey will levitate away from the buildings without needing to push off or anything.

But if the game had stricter gravity and a 2-handed web-swinging system, this section of New York waterfront might suddenly become much more challenging to traverse, and you might have to get creative to successfully navigate this street. Perhaps you just wall-run, then jump away from the wall, and swing again? Or maybe you alternate leaping from a wall to an opposing lamppost and then jump off to do a single swing before starting over on the wall of the next building? Or maybe you piggy-back on a car or truck or other vehicle? Any of these methods will likely slow you down a bit, but that's the point! Trying to navigate an unfavorable landscape should slow your traversal down, or force you to use an alternative method of locomotion. Trying to, say, chase a fleeing criminal or super villain through such a location might get a bit tricky and require some quick thinking on the player's part.

Spidey should want to divert action away from parks because they put him at a
physical disadvantage and also put civilians at risk.

Or maybe the level design shifts such that instead of simply swinging to keep up with the villain, you can jump out ahead of them and have to find a way to guide or divert the villain away from civilians or away from areas that put Spidey at a physical disadvantage. That way, you can more easily keep pace, and maybe even get a better outcome that allows you to catch the villain sooner or limit the amount of collateral damage that the chase causes. Knowing the layout of the map, and taking the time to practice moving around it would thus give the player a strategic advantage in play.

Video games have largely neglected
2-handed, over-handed swinging.

Even on a normal street, with buildings on both sides, repeatedly shooting webs to one side only would eventually result in Spidey being pulled by the force of gravity into the buildings. Even if this automatically transfers you into a wall-run, the distance you can wall-run should be limited. If you want to maintain max speed in a straight line, you should have to alternate your swings from one side of the street to the other. The open-world video games have, in fact, completely neglected the alternating 2-handed web-swinging technique that Spider-Man employs all the time in movies, cartoons, and (of course) the comics.

As an aside, I also think that wall-running in Insomniac's game is too easy. As far as I can tell, you can wall-run indefinitely, and you can even sprint straight up the side of a building because, as mentioned before, gravity is no obstacle in this game. Previous games have included wall-running as well, but the distance you can run on the wall was sometimes limited. In Treyarch's Ultimate Spider-Man, Spidey could only wall run for a second. If you don't jump off the building in that time, you'd slow to a crawl (literally). Beenox, alternatively avoided the issue by including an animation of Spider-Man pulling him up the building with short web-pulls (no player-input required). Insomniac could have made it such that you can wall-run for a certain distance (based on how much momentum you have when you start the wall-run) before you have to slow to a literal crawl. If you want to maintain speed, you would have to jump or web-zip.

Outside of web-swinging, I also thought that Beenox's Amazing Spider-Man games had pretty good wall-crawling controls and camera. Wall-crawling is probably the single most under-developed concept in Insomniac's game, which I hope will get further attention in any sequels.

Accessibility options

I'm not asking for a flight sim or Kerbal Space Program level of detail in modeling how a comic book super hero in bright red and blue tights moves about the environment... That is, unless someone wants to make a Spidey VR tech demo, in which the web-swinging is the whole game.That could have sim physics, and I'd play it! You know, for all of 60 seconds before I get nauseous.

In any case, some suspension of disbelief is necessary. A Spider-Man game should not expect the average player to be solving trigonometry problems in real-time in order to be a competent web-swinger, just like I don't think that a football game should necessarily require an armchair quarterback to have to scan the field at ground level and understand pass route progressions in order to be a competent virtual quarterback. Though, having that knowledge should certainly make you a better virtual quarterback.

Spider-Man should be pulled down and towards the building by gravity.

There should definitely be assist and ease-of-use features, and I'm not asking Insomniac to take those out of the game. Spider-Man 2 lacked such features, and it's web-swinging mechanics suffered for it. Web lines would often be too long, causing Spidey to bottom out and crash into the street below. Even Treyarch recognized this as a problem, which is why Ultimate Spider-Man's webs pull Spidey up off the ground slightly so that bottoming out on the ground doesn't happen. Insomniac pulls similar tricks by slightly shifting Spider-Man's position, or the length of his web lines in order to keep the player in the air and limit the frequency with which you smack into walls or the street. Those are welcome additions. Please don't take that stuff away.

I'm only asking for the traversal to be just a little bit more technically demanding than what was offered in 2018's games. Mark Brown over at Game Maker's Toolkit made a similar argument and compares Insomniac's Spider-Man to games like Bionic Commando, Titanfall, and Overwatch. But perhaps the best comparison that he presents is the myriad mechanics that a Tony Hawk player uses to keep a combo going. Tony Hawk isn't a simulation of skateboarding; it has a very arcade feel. And you can skate across an arena by simply pointing in the direction and going. But if you want to get a high score, and look damn cool while you're doing it, you need to really focus, have a plan, and make deliberate inputs to execute on that plan. That's kind of what I'm looking for from a Spider-Man game.

Mark Brown at Game Maker's Toolkit made a similar argument on his YouTube Channel.

Mark Brown points out in his video that Insomniac explicitly did not want players to be constantly stopped while web-swinging. They were making a game based off an older, more mature Peter Parker who has this whole Spider-Man thing down pat, and having him face-planting into the sides of buildings did not really work with their concept. Of course, that didn't stop them from making the combat relatively tough and leading to this mature Spider-Man taking rockets to the face because of player error, but whatever. It's a valid design consideration from Insomniac, so I can't fault them. So I think a good way for a developer like Insomniac to have its cake and eat it too would be to simply have some difficulty options for web-swinging and traversal that is independent of combat or puzzle difficulty.

Other games have done this in the past. Silent Hill 2 and 3, for instance, had separate difficulty settings for the combat and for the puzzles If you want the mental challenge, but not the stick-skills challenge, there is an option for you. Sports games like Madden allow you to tweak sliders that adjust everything from quarterback throwing accuracy to blocking to the distance that your punter can kick -- for both the human user and the CPU, thereby allowing the user to customize the challenge of the game to their personal playstyle or liking. The Last of Us Part II included an extensive suite of difficulty and accessibility options, allowing players to customize how much damage the player character can take, how much damage enemies can take, how likely enemies are to see you in stealth, and even the aggressiveness of your NPC companions. And you can even turn on several different color-blind modes that highlight the player, NPCs, and enemies in bright primary colors.

Heck, even Spider-Man 2 had an "Easy Swinging" option all the way back in 2004!

Heck, even Spider-Man 2 had a toggle for "easy web-swinging" all the way back in 2004! When you boot up the game for the first time, it asks whether you want to start out on "easy" mode, and makes it clear that you can switch between the two modes at any time. Easy mode allowed the player to hold R2, and Spidey would automatically jump at the optimal point in the swing and automatically shoot the next web so that he maintains a relatively constant speed. All the player had to do was steer.

Insomniac could add similar difficulty settings or optional assist features to its future Spider-Man games. Even with a two-handed swinging system like what I've suggested, an "automatic swinging" mode could be included that allows a younger or novice player to simply hold either trigger button to have Spidey automatically swing. If the game requires player inputs to trigger wall runs or make other context-sensitive moves, then the player could have the option to toggle automatic assist features for those controls.

Having difficulty options for web-swinging and traversal mechanics would allow for casual players to be able to pick up the game, play it competently, and get that satisfying "Spider-Man experience". The players who have been reading Spider-Man comics their entire lives, have fantasies about swinging from building to building, got teary-eyed when they saw that first Spider-Man movie trailer, and that want a more demanding experience, however, would be able to get an experience that requires more thought and skill.

To be clear, I'm not trying to sound gate-keeping or elitist. I may have talked a lot about challenge and difficulty in this video, but really this argument is not about difficulty; it's about control. When I push a button to shoot a web line in a Spider-Man game, I want to know exactly where that web is going, and and also to have a consistently reliable idea of how the game's physics, the character's momentum, and my inputs are going to effect the character's path through the environment in a believable way. Right now, Marvel's Spider-Man is pretty good in all those regards, but it could be better. I'd say it's about 60 or 70% there in terms of Spider-Man putting his web-lines where I expect and want them to go, and with how the physics follows-through. To put that in context, Amazing Spider-Man 2 is probably 80% or better at putting the web line where I want it, but only like 50% (at best) with regard to the physics following through.

Many triple-A games (such as The Last of Us 2) are adding extensive difficulty and accessibility options.

Further, by recommending a variety of difficulty settings and accessibility options, I'm not simply asking for the game to be harder; I'm asking that it be accessible for players on both ends of the skill spectrum. I want everybody to be able to play and enjoy this game, with a level of control and challenge that is appropriate to their age, physical capacity, and skill level.

Thanks for watching, true believer!

Contribute Comment

avatar


We'll incarnate your avatar from the services below.
PlayStation Network Steam Xbox LIVE Facebook MySpace Pinterest Twitter YouTube deviantART LiveJournal



biuquote
  • Comment
  • Preview


Grid Clock Widget
12      60
11      55
10      50
09      45
08      40
07      35
06      30
05      25
04      20
03      15
02      10
01      05
Grid Clock provided by trowaSoft.

A gamer's thoughts

Welcome to Mega Bears Fan's blog, and thanks for visiting! This blog is mostly dedicated to game reviews, strategies, and analysis of my favorite games. I also talk about my other interests, like football, science and technology, movies, and so on. Feel free to read more about the blog.

Follow me on Twitter at: twitter.com/MegaBearsFan

Patreon

If you enjoy my content, please consider Supporting me on Patreon:
Patreon.com/MegaBearsFan

FTC guidelines require me to disclose that as an Amazon Associate, I earn a small commission from qualifying purchases made by clicking on Amazon product links on this site. All Amazon Associate links are for products relevant to the given blog post.

Without Gravity

And check out my colleague, David Pax's novel Without Gravity on his website!

Featured Post

How Madden Fails to Simulate Football: Quarter LengthHow Madden Fails to Simulate Football: Quarter Length06/05/2020 This series of blog post will also presented as a video essay on YouTube. My first foray into long-form video analysis was a cathartic, hour-long, breakdown of how EA and Tiburon's design philosophy causes its Madden NFL video game series to feel disappointing and stagnant. That video was mostly about how EA's insistance on...

Random Post

Konami announces Silent Hill, Metal Gear Solid, and Zone of the Enders HD collections at E3 2011Konami announces Silent Hill, Metal Gear Solid, and Zone of the Enders HD collections at E3 201106/06/2011 Konami let slip last week (and confirmed this week at E3) that it will be porting HD remixes of the PS2 editions of some of it's most popular franchises, including Metal Gear Solid, Zone of the Enders, and Silent Hill for the PlayStation 3 and Xbox 360. UPDATE June 8, 2011 (3:46 PM PDT): Just read the following article from...

Month List

RecentComments

Comment RSS