Payday 2 is a unique and ambitious co-op shooter, but often frustrating, and heavy on the grind. Given some of those keywords, I might’ve expected something more like Diablo 3: a simple game where I’d be permitted to lose focus while I levelled up or collected weapon mods. Essentially, a game for multitasking–a “podcast game”. But even though it’s a grind, it requires meticulous play and also communication, at least until each of the players understands the best practices for a particular heist and knows their division of labor without being explicitly told. That’s a mixed bag, since I don’t want the gameplay to be trivial, but I also don’t want to spend days levelling up and not even be allowed to blow through some podcasts while I do it. At least the music is pretty ballin’.
The emphasis on cooperative play also makes it troublesome to play with random amateurs in public lobbies. It’s also not great to be on the other end, feeling the guilt of screwing up a bank heist for three other people. It’s an interesting game, and it’s best for an adaptable crowd: gamers who are tolerant of mistakes and aren’t abusive to inexperienced players. Play with friends if you can. Don’t delay in disconnecting from cheaters and easily-exasperated children and you’ll generally come out alright, though it might take a while to find a game. The mechanics usually aren’t self-explanatory, so part of stepping away with a positive opinion is contingent on having someone show you the ropes.
The biggest problems are probably the crashes and faulty netcode. Sometimes you click on games for five or ten minutes before finally getting beyond the lobby on a heist you’re willing to do, only for your game to crash, or for the host to crash. Sometimes you play for an hour and the host’s internet connection drops on the last day of a three-day heist, so you get nothing. My suggestion for how this should work is to restart to the beginning of the current day, attempt to connect the other players together for a minute or so, and choose a new host–or fork it, so that each player becomes the host of separate games. It would also be nice if the host could set flags like “Stealth only” or “Loud preferred” so that people would know what they were getting into before they joined. And it would be nice if the host could specify the need for a saw, or C4, or some other specialty on a player slot before it was filled. And it would be nice if there were heist filters for joining players, so they could hit “deselect all” from a list of maps and then select the two or three heists they were willing to do. And it would be nice if someone could load a preset in two clicks from the planning phase, rather than individually replacing each spotter cam every time they failed.
There are also design issues, and these have more to do with esoterism and randomization. As for the former, the Hotline Miami crossover heist is a good example: if the player is told to trace a location in the Downtown area, they need to find a shipping crate containing cigars. To cook meth, it’s muriatic acid, then caustic soda. There are hints for these things on the walls, but they’re easily overlooked when on the clock and getting shot at in a multiplayer game. A number of heists are based around elements of memorization or wiki-lookup like this. I acknowledge that the game exists in an ambitious, uncontested territory, and that designing these heists is a trickier job than building deathmatch arenas in any other shooter, but the need for improvement is still undeniable. It’s good to be complicated, but not needlessly inimical to newer players. It isn’t just an issue with the levels, either: it’s pretty esoteric information that you can always answer four guard pagers, and difficult to go twenty minutes in stealth without forgetting whether you’ve answered two, or three. It’s harder still to keep track of how many the other players have answered. This is a good reason why I couldn’t imagine playing without the UI changes of mods like Hoxhud or Pocohud now, as it throws those numbers right on the screen, among other desperately-needed improvements.
As for the issues with randomization, it may be true that the game is most exciting when something goes wrong and players are scrambling to pick up the pieces of their action plan, rushing loot into the van as fast as possible. But on higher difficulties, stealth attempts tend to be all-or-nothing, and once the police have arrived, it’s time to hit the restart button. Some of the RNG-based heist variations are good for keeping players on their toes, like not knowing where to find a security room or keycard. But if an item is inaccessible without a perk that the team lacks, or if extra security guards arrive early, or if guards have chosen a particularly irritating patrol route and never isolate themselves, it can mean quick and obvious failure. Sometimes you just have to choose one of two possible trucks on the Election Day heist and all you have to do is peek into one crate along the way. Sometimes there are four or five trucks that could be the right one, and you have to avoid guards for fifteen minutes to figure it out (or cheese it with multiple players’ ECMs). You have to strike a good middle ground, and the game doesn’t always pull that off.
I think of Diamond Store as an example of a well-designed heist that relatively new players can learn to tackle reliably with a clear game plan, apart from a very rare variant where the camera room is only accessible from the inside of the store. Variants where there’s a camera near the outer door are fine, although someone might have to use an ECM while getting the door open. They’re also kept on their toes by the variable location of the silent alarm keypad, but not in a way that should make an attempt ridiculously harder. If players had a distaste for restarting when an error was made, loud-spec heisters could still hang out in the back alley to take care of roaming pedestrians, or use the cameras once the operator was taken out. The couple players doing the stealth work could also hopefully survive when it goes loud by using a dodge build, deferring damage to the tanking players. It’s an easy level with a low payout, but other levels could try to emulate that flexibility. It’s not perfect: for example, it’s weird that you have to lockpick the camera room’s doors from the inside, but it’s pretty tight for the most part.
It’d also be great if a mission’s loud objectives were always alleviated when some of the objectives were completed first in stealth, as this would encourage the “Plan B” mentality of rolling with the punches. Getting the vault open first before messing up in GO Bank is an example of this. The final day of Framing Frame, where you can sneak most of the way through and then get a completely new set of objectives when things go loud, is a lot more frustrating: at best, your stealth can earn you a little money, or get the door to a server room unlocked, which is nice, but not useful enough to validate going into the assault with a garbage high-concealment silenced weapon and a suit.
Loud missions invariably involve standing your ground and shooting a few hundred people while throwing loot around or protecting a drill or circuit breaker. This is alright, as shooting people is a draw in itself for most players and is varied by skill trees, but a few more mechanics would go a long way. For example, players could man turrets, activate existing defense systems, blow up entry points (already possible with shaped charges in Hoxton Breakout), and countless other things, I’m sure. I’d rather see tweaks to what’s already in the game, but that said, it was a huge and surprising change to see driving mechanics get added in through a new heist–a big thing to add, especially 18+ months after the game was released–and direct control over an escape van could be a cool thing to do in a loud mission in the future.
But it’s the stealth half of the gameplay experience that is in more desperate need of change. For one thing, the player’s concealment level could be better as a variable represented by a UI meter: a value that decreased while crouching in shadows and increased while jumping around on the street, or having guns visible. Reworking some of the arbitrary divisions like Casing Mode, by being able to re-conceal a weapon, could be cool if they could make it work. Casing could be hybridized with the pre-mission planning, such as by placing the planning maps in the escape vehicle parked nearby, and letting players come back to the map to circle guard positions and so on.
Guards are a little too likely to blindside the player when sneaking around, given that you can only kill four of them per heist, but the game was designed as a cooperative experience, and it’s good that it’s tricky to play from the perspective of solo stealth games: it becomes all the more important that players have several means to spot guards for each other. Considering the numerous ways to do this with perks, like spotting while in casing mode, or placing laser tripwires that highlight any guards that move through them, it’s clear that this was the intent, though it could be implemented in more fun ways. A recent heist had a drone flying overhead with a camera feed that could be hijacked, and if such a drone could actually be deployed and piloted around, maybe even aiming a laser at cameras to temporarily blind them for other players, I think the idea of cooperative stealth could be even better realized.
Making deep fixes to the stealth system would require changes to each heist, but that would come near the top of the changes I’d like to see, and could be doable if some older heists could continue to use the old system in the meantime. I think it would be best with no limit to guards killed, as long as the guards are blissfully ignorant until they’re stabbed in the back. Players could hijack these guards’ coms through some new mechanic, toggling a radio to the correct guard’s frequency to say “All clear over here!” every ten minutes or so. But when players are detected (red exclamation mark overheard), guards should signal their panic buttons immediately, and if players were able to kill the guard quickly enough after that, they could call in on the guard’s pager to report a false alarm. Some guards might always keep an eye on each other or move in pairs, which would require two players to stealth-kill those guards simultaneously, or else kill one right after the other and answer only the second guard’s pager. On the second-highest difficulty level, players might get a half-second of guard alertness without having to do answer a pager, and maybe a full second on the difficulty below that. Yet with this system, four answerable pagers wouldn’t be a sure thing either–they might get just one, or none at all on the highest difficulty. It could even vary depending on how close together the pagers were flagged, like one pager every five minutes for a maximum of three.
Payday 2 has done some interesting things with its business model, and at times I’ve viewed its pay model optimistically. It’s seemingly increased the life cycle of the game, sustainably, while at the same time allowing people to access most of this new content without buying it (DLC heists can be freely accessed by joining public games, but not soloed unless purchased). There’s certainly nothing wrong with $20 cosmetic packages that let superfans feel like Kickstarter backers, at least not as long as the developers continue to meaningfully update a game that was released two years ago. And some of the new heists have been pretty cool (the aforementioned surprise driving mechanic, for instance).
But it’s also failed to live up to certain promises, such as with a recent upgrade to the infamy tree that added nothing but some masks and more experience levels for the sake of making numbers go higher: a classic behavioral trap common enough in MMORPGs. The developers have pledged to take a look at revising this system in the coming months, but it remains to be seen if that quote is worth anything.
There is a good chance that the game will never see a deep stealth overhaul or any of these other wished-for changes, despite its pioneering pay model. As countless Early Access titles and games with significant post-release updates (like Wasteland 2) have shown, updates can be fantastic, but it almost never happens that a game sees deep structural change after reaching players’ hands. To be clear, Payday 2 has already held my interest longer than any other multiplayer FPS I’ve played in several years, and certain updates (like the Fugitive skill tree) have been great. But as long as the game drains wallets on a higher level, it seems fair to hold the post-release content to a higher standard as well. Spend wisely, Overkill, and speak frankly with your audience. It would be the easiest thing to revise this review if the game saw a successful revision of its own.