I’ve been obsessed with Marathon over the server slam weekend. I don’t necessarily love the game, but it’s given me so much to think about that I wanted to get my thoughts out and onto digital paper. This post is concerned with the social layer of Marathon, or rather the absence thereof: why adding proximity chat won’t suddenly make the game social, why (I think) being a social game matters for mainstream success, and how certain design decisions came together to preclude a social layer from ever developing.
Let me start by making a few comparative statements between Marathon and Arc Raiders. Two disclaimers up front:
- I’ve played around 100h of Arc Raiders and around 9h of Marathon in this weekend’s server slam event.
- Everything subjective sounding in this article should be read with a mental “for me”. When I say “Arc Raiders’ voice chat makes other players feel like characters in the world” and you disagree, just add a “for me” in there. It’s just, like, my opinion, man.
I’m trying to make a few comparative statements up front without value judgment. These are my observations, and for now I’m not talking about which is better or worse. Okay? Here we go:
Arc Raiders has an abundance of loot. Almost all loot has utility beyond “can be sold”.
Marathon loot feels scarce outside of door key rooms. It always feels viable to share in Arc, but not in Marathon.
Through crafting, recycling, upgrading benches, and by often being an input to quests, loot in Arc Raiders has many layers of meaning.
By contrast, Marathon loot falls into two categories: “I can use this” or “this has direct economic value” (either by selling the loot or because it’s salvage, which is a category of many currencies).
Because so much of it has only economic value, loot in Marathon feels interchangeable.
Arc Raiders has a long TTK and a sound system that allows players to locate fighting reliably and at long range. Bailing from a fight is usually possible.
Marathon features a much shorter TTK and effective mid range sniping even with early game weapons. You can die to a human before you have time to register there is another player.
Fighting AI in Arc Raiders feels like fighting a very non-human threat. The AI is in a category quite distinct from you and other human players; it moves, shoots, fights, dies very differently from players.
By contrast, UESC drones in Marathon feel at least player adjacent.
Once you’ve learned them, Arc Raiders AI gives you plenty of warning and doesn’t usually kill you suddenly and unexpectedly.
I don’t know if the same will be true for the UESC drones in Marathon, but their damage profile can be quite similar to players’ damage profile. I therefore fear they’ll remain a sudden death threat.
Between organically grown community rules of behavior and Embark’s aggression-based matchmaking, Arc Raiders matches have a wide range of expected outcomes.
By contrast, in my 9 hours of Marathon I’ve seen almost exclusively highly lethal matches that ended with my death.
Arc Raider’s voice chat makes players sound like characters in the world. The city of Speranza and the common fight against the clankers predisposes at least some players to collaborative gameplay.
Marathon’s voice chat sounds similar to speaking to someone on Discord. This highlights the game-nature of the experience and primes me for thinking of winning and losing, competitiveness, the other as a player with cold risk/benefit calculus, rather than an embodied character in the world. The narrative positions players as faceless mercenaries for nefarious organizations. Everyone is a target.
Okay, so far I’ve described my sense of a few ways in which Arc Raiders and Marathon differ. Again, not inherently good or bad. A Tarkov player in fact might quite prefer Marathon’s setup. But I want to use the above points now to make an argument: Marathon does not support social play in the way Arc Raiders does. Why does this matter?
I feel that playful social moments, uncertainty about the intentions of players encountered, and an overall sense of the game being an unpredictable playground are what gave Arc Raiders the success it has enjoyed. I would point to the incredible longevity of games like GTA Online and EVE Online as well.
Let me talk about my theories of longevity. In short, games can be long lived for one of two reasons: either the content of the game itself allows for meaningful exploration and engagement or the way the community engages with the game becomes the game itself.
In category 1 we find games like Counterstrike, League of Legends, Starcraft: Brood War, but also most of the long-lived games enjoyed by the fighting games community, including various version of Smash that haven’t received any meaningful post-release support and are still enjoyed by many players. In this category we also find games like WoW, where the developer has to sprint on the content treadmill just to stand still, and games like Call of Duty or FIFA (or whatever it’s called now), where a full price annual release takes the role of a live service game’s support cycle (nevermind that all of the above have ALSO built live service game stuff on top of all that, much to their communities’ exhaustion).
In category 2 we find games like GTA Online and EVE Online; also games like Rust, Arc Raiders. But also games like Among Us, Fall Guys, the modern wave of what we call “Friendslop” with games like Peak and Lethal Company, and also games like Roblox, Fortnite, Gary’s Mod, and arguably even horror games like Five Nights at Freddy’s. These horror games are defined by social play mostly through streaming and videos. Content Creators model a way of playing these games and their community enjoys the content not merely because the gameplay is great or because they love the content creator, but because of the intersection of the two.
Category 2 games are defined mid gameplay loop by social, human to human interactions. GTA Online’s most successful servers are full on roleplay servers. EVE Online is often lovingly referred to as a somewhat interactive Excel spreadsheet; surely its sterile combat isn’t what keeps people coming back, it’s the social play between corps, the human to human relationships that result in the logistics that actually end up winning wars, and so on. For games like Among Us (or other Werewolf-like games) it’s purely a game of social deduction. It literally would not work without human players. Games like Fall Guys and the many other funny-because-physics games are social games in that purely playing them to win isn’t particularly enjoyable; you’d rather play with a bunch of friends and laugh at each others’ misfortune. Peak and Lethal Company are very similar. Theoretically cooperative experiences where “accidentally” causing your friend to fall to their death is peak entertainment. There’s an argument that part of Helldivers 2’s main loop can fall into this, with friendly fire that’s gleefully applied to everything you do, but Helldivers 2 can and often is played “straight”. Maybe they use the socially fun play as an onboarding mechanism.
Finally, games like Roblox and Fortnite and Gary’s Mod of course feature ludicrous amounts of content, but I’ll argue that that content wouldn’t be enough to keep people coming back. Roblox experiences aren’t great by themselves I don’t think. It’s in the social nature in which these experiences are discovered, enjoyed, and mastered that Roblox shines. Fortnite has a similar claim to category 1 nature, just like Helldivers 2, but I don’t know enough about the game to judge how many people play it purely competitively. I figure it’s big enough that you’ll find any kind of play you care to look for.
Okay. A lot of recent failures tried to be category 1 games because this is how we’ve made games for the longest time, and it’s the best understood way of making especially high budget games. Games are primarily well-structured competition with winners and losers and clear mastery curves.
This is how Wildgate positioned itself. It’s obviously not the biggest problem Concord had (hot take: none of their left clicks felt good is what doomed them), but they did also subscribe to the win/lose round based nature of traditional games. Highguard subscribed to this model as well. And of course I’ll argue here that so does Marathon.
But not Arc Raiders.
I’m not saying category 1 games can’t be successful. Marvel Rivals was successful. Battlefield 6 was successful. Delta Force was successful. These are all category 1 games.
Category 1 games live and die by their content. Of course they all have really big and important communities built around them, but largely how you play them and how you derive pleasure from them does not depend on the community. Sure, top players in League define the meta for players who really want to win; Overwatch’s Arena mode even has pre-made builds authored by influential esports athletes baked into the game mode. But many players do ignore this, and even those who don’t make moment to moment decisions based mostly on the gameplay the developers created. The same cannot be said for games like Arc Raiders. Devs very much did not predict that players would be as peaceful to each other as they are. They didn’t predict mobs of players hunting down the one player on the server who called fake friendly and murdered someone. And “friendly!” is just one of 6-8 default voice lines and yet it probably accounts for more than 95% of character voice line usage in game.
I don’t see that kind of behavior in games like Hunt: Showdown or Tarkov, but I admittedly don’t have too much first hand knowledge of these games. I certainly didn’t see any of this kind of behavior in my 9 hours with Marathon. I think that makes Marathon a category 1 game.
Let’s come back to my list of comparative observations from the beginning of this article and connect them together, because I was making a design-based case for why Marathon will not have the kind of social expression Arc Raiders does.
Loot: A scarcity mindset turns every other player into a drain on your expected value. (Nevermind the frankly insane UX decision to let multiple players access the same loot source at the same time. WTF, Bungie). An abundance mindset means that if two players cooperate, they will both get meaningful value. Maybe you’d get slightly more value for killing the enemy and taking what they have, but now there’s a good chance you’ll die and extract with 0 value as well. The math doesn’t math.
Add to that a subtle implication of Marathon’s lack of crafting. Outside of direct power items (weapons, mods, equipment, cores, implants, etc) you have a motley array of “vendor trash” (aka auto-sell) and “salvage” (aka auto-vault) items. Let’s focus on vendor trash items because the case is a little clearer there. These are items that have an icon and a theming purely as a cosmetic layer. Yes, rarely they’re called for in faction contracts, but most of the time they reduce down to their monetary value. Even after my relatively short time I stopped looking at the icon or the text and just scanned the money value in the top right corner after seeing the little “bag of gold” mini icon on the item that told me it’s vendor trash.
This means you’re not looking for specific stuff; you’re looking for the most value density you can achieve. And since most of these items stack, and since you won’t likely have full stacks of all of them, there’s a very good chance that no matter how shit the rest of an enemy’s player’s backpack, they’ll likely have some vendor trash items you can add to your existing stacks, thereby increasing overall value density.
This means that unlike in Arc Raiders, killing an enemy in Marathon is almost always clearly optimal. Add to that how many times you’ll find yourself running with a free kit and you’ll almost always want to kill an enemy to grab additional guns, mods, implants, ammo etc. A free kit will not kill very many AI and you’re unlikely to find a lot of what you need to kill AI in the world.
Thus:
Loot scarcity and loot fungibility strongly incentivize PvP.
Short TTK, reliable weapon performance at mid range, presence of early game sniper-like weapons (“precision rifles”) and scarcity of shields all conspire to make Marathon a game in which encounters are resolved in fractions of a second. There is always pressure on the game designer to make TTK shorter; all else being equal, a short TTK feels more satisfying and “fun” in the moment. This is a classic sugar vs vegetables situation though, where the long term negative impacts of a short TTK can take tens of hours of gameplay to manifest. Also, once you’ve shipped a short TTK it’s almost impossible to get the community to accept a longer one down the line.
Short TTK means you cannot afford to leave other players alive. You cannot try to have a conversation or just look at an enemy to gauge their behavior. It’s kill or be killed. Thus:
Short TTK penalizes hesitation or mercy.
Arc Raider’s unique AI alongside its excellent embodied voice chat and narrative positioning of Speranza vs The Clankers creates a feeling of “us vs them”. AI is designed to be fought by groups of players and feels nothing like fighting humans. Fighting the UESC in Marathon feels like fighting human players with moderate brain damage but highly inflated stats. Everything in Marathon’s narrative is fucked. It’s a hypercapitalist hellscape of disposable bodies and organizations at open war with each other. In fact, being a team-based game first and foremost makes no sense narratively. Why would these three random Runners work together? They don’t even have missions from the same orgs necessarily. (The reason is purely gameplay; high lethality in a one-life mode is more acceptable with a team to bail you out, and character design can more easily highlight strengths and weaknesses if there’s a reasonable expectation of having other different characters on the team).
Thus:
Competitive, cut-throat positioning primes players for conflict
There was a strong push for proximity chat after Marathon’s last round of public beta testing. I think there was an expectation that if you just add proximity chat, interesting social situations would arise. Playing Arc Raiders cemented that expectation in my mind as well, but I now think that adding proximity chat was never going to move the needle on a game that so firmly positions itself as a cut-throat, competitive, zero sum game. I’m not saying that’s a mistake; that’s the game Bungie wanted to make and it’s the game they ended up making. In fact, it now makes a lot more sense to me why Bungie were so reluctant to add voice chat. I don’t think they were primarily concerned with things like moderation but more so asking themselves “what’s the point?” They knew the game they were making was a game where you communicated with other players not with words, but with bullets.
And of course this whole post must read like I’m predicting Marathon will fail, but I’m trying very hard not to do that. First of all, because that’s a god damn ghoulish thing to do, especially for someone also working in this accursed industry. I wish my friends at Bungie nothing but success.
It’s also a really damn easy prediction. Tell me that a big budget multiplayer game that isn’t a sequel to an establish, long-running and super successful franchise will come out in 2026, and I could predict its failure and be correct 95% of the time. That’s super depressing and not really interesting as a prediction.
I will say this though: category 1 games, games that live by their gameplay alone, have had a much harder time of it recently. Even games like Battlefield 6, which had an incredible release, have struggled to remain relevant mere months post-release. BF6 has lost 90% of its CCU in only 3 months. And the problem is, of course, content. Not enough maps, maps that aren’t big enough, not enough new weapons, a BR mode that the community resoundingly rejected, nothing but mis-steps on the post-release management of the player generated content side of the game. By comparison, Arc Raiders hasn’t exactly showered its playerbase with content either, and there are similar grumbles you can see in the subreddit; but player number are still incredibly healthy. GTA Online, an 11 year old game, is still pulling in north of 110k daily CCU, and it’s not exactly been a garden hose of content. There are two updates a year, and it’s been that way for a long time. By comparison, League of Legends patches every other week and has been doing so with very few exceptions since September 2009. Not every patch has a huge amount of content, but there’s always some gameplay changes to entice people to play a game or two.
I don’t know how easy or hard it is to make a category 2 game, but if your game is in a genre that has successful category 2 games I think it would at least be worth the consideration.
In spite of losing 95% of my runs, I’ve had a lot of fun with Marathon; I think the build crafting and the weapon mods are way more interesting than what exists in this space in Arc Raiders, and I’ll definitely buy the game and play it a bit. I’m a sucker for category 1 games and at a stage in my life where it’s getting harder and harder to corral friends for a round or two of a category 2 game. But that said, I benefit from Arc Raiders’ category 2 nature even without friends. I have whacky, fun experiences nearly every time I sit down to play it, and even when there isn’t a train of recorder playing weirdos or someone roleplaying a bridge troll, there at the very least is a very juicy uncertainty of “what will this person do”, and when they turn out to be super nice and helpful, that can feel very good, especially because I realize they could just as well have shot me and taken my stuff.
I think I’ll enjoy playing some more Marathon; I know I won’t stop playing Arc Raiders for the foreseeable future.