A while ago I wrote a post about some of the usability problems in Escape from Tarkov’s menus. This post is about the gameplay, the match, the raid, the actual game part of the game. The game’s developers have put a lot more effort in the in-game experience than in the menus, but it’s still far from perfect. This article points out some of the largest UX problems in the game, with some concrete suggestion on how they could be fixed.
Tarkov has been developed for almost ten years, and on public beta for almost four years as I write this. While game development is hard, that should be more than enough to fix some of the fundamental UX problems in the game. A large majority of the issues are for new players, but veteran players should be able to benefit from the fixes as well. This article is my own personal point of view; I have zero relation to the creators of Tarkov, Battlestate Games.
- The near-obsessed approach to realism with awful onboarding kills immersion. Some helpers, such as hinting the exit locations, should be utilized.
- The game is made to be played with Wiki, which is against the experience vision and makes completing tasks tedious. All the vital information should be provided in the game itself.
- In-game objects should communicate better how they can be used with their visual appearance (affordance). Players should never use trial and error to see which doors for instance are only decorative.
- Maps needs some random elements to support the experience of players staying on their toes.
Player Experience Vision
Any user experience improvements should be guided by the game vision; that’s the heart and soul of the player experience after all. Tarkov has quite strong, even unusual take on the player experience.
The developers describe the game on its website with following words:
The Protagonist finds himself in modern Russian the city sunk into anarchy: only the fittest will solve the mysteries of Tarkov and get out alive. This combat simulator in dangerous & hazardous environments uses multitude of system modules for full gameplay immersion and sense of reality.
Survival, mystery, immersion and realism. That pretty much sums it up.
In Tarkov you have to fear for your life. In Tarkov you jump for every crack of wood you hear behind you. In Tarkov you fail, die and lose everything. And that’s exactly why the occasional small victories taste so sweet.
Much of the tension in the game is caused by its semi-permadeath approach to your gear. If you die during the raid you lose all (well, almost all) of the precious gear you brought with you. This means you are basically investing some in-game currency for every game you play, making the risks really matter. A single stray bullet can usually end your raid, and the constant fear of death raises the stakes even higher.
The fear factor in the game is implemented just brilliantly. There are just enough sounds and movement in the game world to make you paranoid about every corner and tree, and the scav mechanism levels up the whole concept. You never know if those steps you heard are coming from a less dangerous AI scavenger or the ultimate danger – another human player.
This leads to an exciting cycle of sound control. Heavy gear and and looting containers leads to more noise, which in turn attracts unneeded attention. The gameplay is a fine dance where the player switches between roles of being “rat” (looting, camping and staying away from trouble) or “chad” (aggressively attacking other players). In general, the use of sound is done rather well in the game and supports overall experience.
The game has a pretty solid player base and most of the players have already gotten used to some of the oddities you might find in it. Many would argue changing anything in the game, “mainstreaming it”, would ruin the whole experience, but there’s a strong survivor bias in those words, in my opinion. With proper UX design the game would let you enjoy the game rather than avoiding the problems in it.
The Problem of Immersion
I could bet words “just like in real life” were mentioned several times in the game design meetings, and it does show in the result. Many of the conventional helpers – status messages, player names, directions – have been left out of the game to make it feel more immersive. The logic is that the less "gamey" elements and the more aspects of real life combat can be included in the game, the more capturing the game experience will be.
The wrong type of immersion kills immersion.
That’s not the case tho. Many things that are completely natural in real life, like checking a gun’s fire mode or positioning yourself in terrain, can be awfully awkward in a computer game; you are after all controlling a whole human being with just mouse and keyboard. The player character is a tough as nails ex-combatant, surviving in environments others wouldn’t dare stepping into, but currently this feeling simply doesn’t transfer to the player who is struggling to figure out where they hell they are or what elements in the game world can be interacted with. For a good amount of time, you feel like a brain damaged donkey rather than the warrior you’re supposed to be.
In most games important information is displayed in the UI, often in HUD, as a sort of projection of the player character’s thoughts, but Tarkov leaves it all to the player. It’s really hard to get immersed into the fantasy of being a seasoned soldier when you’ve got no clue of what’s going on and you find yourself “playing the UI” rather than the game. This inflexible idea of immersion just doesn’t feel believable or realistic, leaving the player often thinking how they can do something rather than actual tactics. In other words, the wrong type of immersion kills immersion.
Finding the Goldilocks spot between HUD elements and diegetic UI can be very hard in games, but Tarkov in general just offers too little information, which is painful especially for green players. At least extraction zones should be communicated more clearly, as currently everyone just plays with a fan-made map open in their browser window.
Having to rely on external resources for information is probably the largest immersion killer in the game. A large part of the players are just playing the game with an online map opened on secondary screen, and it’s a common practice to have current tasks’ objectives (with map and screenshots) available from Wiki while you’re in the raid.
It’s naïve to think that players wouldn’t find a way to get information they need online, so you might just as well provide that info in the game – that way the developers would have control over how it is presented, making the that information part of the game world rather than some ad-infested web site. This issue is revisited in many other areas of this article.
Die, memorize, repeat
Long-term player journey of the game could be roughly divided into three parts.
You start off as a freshman, obviously. Everything is new and very painful (see the next section on learning curve). You learn the basics and you die. You die a lot. At this point you learn to either love or hate the core idea of the game.
The middle part of the game opens with access to flea market. By now you’ve learned the most common exit points (from online resources) and you’re living the life of a rat, slowly building your base, scoring lucky kills here and there. This part of the game, in my opinion, is the real Tarkov experience. The mystery and exploration is still there, your equipment is still mainly improvised and your approach to combat is realistic; there’s a very healthy dose of fear of death during each of your raids leading to extreme concentration.
Adding some random elements would bring back the feeling of mystery.
The late game starts with bitcoins. You’ve got more money than you can spend, and you pretty much remember all the loot locations by now. There’s nothing to explore and each raid seems to repeat itself: everyone in the game rushes for the same high-quality loot – always in the same exact location – using the same stupid jump and zig-zag movement tactics you only see in other hero shooter games. At this point there’s not much else to do than go full-on chad, essentially transforming the game into Call of Duty, or just wait for the next wipe.
The mid-to-late part is one of the least rewarding parts of the game. You’ve started to really enjoy the game, but feel the repetition slowly corroding the parts you enjoy the most. One main reason for this is the lack of random elements. Scavs always patrol the same areas, loot appears precisely on the same locations, doors always reset to their original states and even other players always lurk at the same spots. The gameplay becomes an optimized Groundhog Day simulator. Now, I know some people actually enjoy memorizing the location of every single bush and the state each door spawns, but I would argue it is against the game experience vision of mystery and immersion.
Adding a bit of random elements into the maps would make the players stay more on their toes and add some joy of discovery even into the end game. A good start would be slightly changing location of loot for each match, while shuffling the location of small bushes and even room layouts would be a good next step. Going further, adding a bit of randomness to the mission locations would make them have the sense of mystery instead of sense of Wiki (see more about this on the missions section).
The Deep End of the Pool
Onboarding on Tarkov is an area that is clearly an afterthought. The player hideout and flea market are baked well enough into the player progression, but during the raid the player is dropped right into the deep end of the pool.
If you’ve ever played Tarkov, you know what I mean about its learning curve. For the first ten or so hours you basically just wander aimlessly, getting killed by enemies you never even noticed. The situation is made even worse by the heavy smurfing among the players, which means the first human being you might meet has played the game for 6,000 hours and one-shots you in the eyes with a pistol while jumping down the roof (true story, happened to me).
The offline mode is just plain lazy design.
Currently there are two tools for onboarding: external resources (already mentioned in last section) and offline mode. Offline mode in general is just lazy design, as its function clearly is to just let the player memorize the maps – and even practice the fetch missions! – so that proper in-game navigation tools can be ignored.
The onboarding could be improved a lot by using only the beginner friendly maps for newbies or dividing the maps into smaller parts; player would start with a small portion of the map and proceed to the full version once they have successfully extracted.
Second large improvement would be to give the player some info about directions and exit locations. The game already has a compass in it, but irrationally you don’t get it until you’ve familiarized yourself with the huge Woods map; how about just giving it as starting gear, when it is needed? Exit locations could be improved without breaking immersion by using some Geiger counter type of device that could give some clue about the exit’s general location.
Another option would be some type of intel that you would be able to gather or purchase from the game traders - or even a local AI scav that you could hire as a guide. The Sherpa program (peer support by community) is something that currently sort of works like this, but it would be really sweet to have the feature in the game instead of counting on the goodwill of the game fans.
A third problem areas is that almost every action is assigned to a keyboard button of its own. There are almost one hundred actions you can perform with various keyboard commands and it takes hours and hours for the player to learn the most important ones. And I’m not even talking about the more advanced actions such as “unload bullet from chamber” or holding C with mouse wheel to change your stance. Yes, all the actions are important, but finding the key binding can be quite slow, especially in the middle of a raid.
A simple solution to battle both button overload and lack of awareness of match status would be some sort of beginner’s help view that can be opened easily. Instead of navigating around all the menus for button mappings and info on what the player can do, a fresh player could bring in the help overlay for quickly checking some of the info. The best part of this is that it wouldn’t hurt the immersion of advanced players and doesn’t add anything into the current in-game HUD.
The best implementation for a help layer like this would be with press-and-hold functionality. For instance, the layer would be visible while the player keeps H button pressed down, so that it acts purely as a reminder element. This type of UI element could be used for other onboarding items as well (such as tips and mission status) but I’ll just keep it simple for now.
One of the most important elements of player progression are the game missions, AKA tasks. They are given to the player in the game menus by traders and often include assignments such as gathering supplies or finding unique items.
The missions serve two important purposes: teaching gameplay mechanics and opening the game storyline. Especially the teaching part is done well; early missions take place in maps that are the most beginner friendly, and they introduce some basic mechanisms such as looting and/or crafting some basic items. One could of course ask why the most complex maps are even available for complete beginners, but that’s something I already mentioned on the onboarding section.
Missions are impossible to complete without external resources.
Missions also hint about some useful combos, such as crippling an enemy before aiming for the head and using flashbangs to enable close combat kills – the later one especially is something that players will naturally experiment with.
Missions have one huge problem tho: they are impossible to complete without using a Wiki. For instance, there’s a quest for finding three unique chemical containers that only appear on the game once the missions has started (i.e. the player will have no idea what they look like). For the note, these containers really look like ordinary metal trash you see everywhere in the game.
Okay detectives, here’s the location for the mission objectives: “These containers were seen at the resort, and one pal of mine claims to have seen them on EMERCOM people at the Interchange”. In case you’re not familiar with the maps, this resort mentioned is a huge 4 story complex that has hundreds of rooms. Trying to find that needle from the stack of haystacks – in poor lighting and in the middle of people who are trying to get you killed – without external help might take a hundred hours!
Other examples of this poor instruction include getting an item from a person who “existed somewhere” and finding two unique medical instruments from “streets or buildings”. A typical Tarkov map is some 16 square kilometers in size, so you do the math.
On the same topic, many missions include sadistic pixel hunting. The interaction distance for the player character is about their arm’s length, making “scanning” of items extremely hard. Combine this with the fact that some unique items are literally just a couple of pixels in size in the game world and you’ll understand why players just open the Wiki page as soon as a new mission is available.
In addition to being a major UX problem, this undermines the game immersion. You no longer concentrate on the game but run it along a secondary screen with third party maps and screenshots, making the whole thing feel like filing taxes. This also leads into a habit of just skipping all text and jump to Wiki straight away, essentially eliminating all the fun of the journey and missing the interesting lore; missions become just XP farms with no relation to game experience vision.
The missions are obviously designed to have a sense of mystery and exploration, so any UX changes in the system should be carefully considered. Exact instructions or a “modern” beacon system that tells you exactly where you go would be against the experience vision.
A lore-friendly solution would be traders drawing some markings on the player’s map with a rough location of the item’s expected whereabouts. This could be supported by some type of inaccurate tracking device that the player is able to activate, possibly combining it with the compass and/or map. If the mission objective locations would be randomized a bit, while still being scannable with the tracking device, the feelings of mystery would magically find their way back into the game.
The concept could be further enhanced by something that already exists in the game to a degree: mission hints. You could acquire some minor items or even pay the traders for additional markings on your map to make the hunt easier.
Game world affordance
Affordance, to put it simple, is a term in user-centered design that means objects communicating their functionality with their physical form and appearance. In video game setting this means that if something looks like it can be climbed on, it should be; if something looks destructible, it should be; if something looks like you can shoot through it, you should be able to.
Tarkov’s bullet penetration system is a perfect example of this: any fabrics and wooden doors really look like you should be able to shoot through them – and indeed you can. There’s no need to explain to players which materials can be penetrated as they can naturally figure it our from the objects’ appearance.
If it looks like you can jump on it, you should be able to do so.
Much of the level design has good affordance for how areas can be accessed. Boxes can be jumped on, stairs can be climbed and small gaps can be jumped over. However, some maps contain objects that look like you could jump on them, but they are in fact just a bit too high. The player can no longer naturally “read” the game world, and the object’s affordance is broken.
The same broken affordance can be found on ladders (you can’t climb them), shallow water (you can’t walk there) and open windows (you can’t climb in through them). In general, this causes a frustrating trial-error loop, as the player naturally feels an object should work in certain way, but it simply doesn’t.
A simple UX fix is to make the world for the player. Any decorations that hint about functionality should simply be removed or styled so that it doesn’t look usable; in case of ladders, just cut off the bottom part of them. Any objects should be scaled so that they can either be jumped on or they clearly look too high for that purpose. Yeah, this is not “realistic”, but we’re talking about a game here, and players being able to play it is more important than everything looking hyper realistic.
One good example of this problem category is the the nemesis of all games UX design – doors. There are lots of doors in the game and they come in many forms, but their form has nothing to do with their function. There’s no way for the player to tells if a door can be opened, unlocked or breached until they walk right to it and try. Even worse, some doors are purely decorative and the only way to find out is, you guessed it, trial and error.
This type of lack of visual communication doesn’t have any gameplay advantage, while it forces the players to just memorize the maps. A good fix would be to visually code the doors, so that the player could just look at them and identify what type of door it is: normal wooden doors can be breached while reinforced ones cannot, and those that are completely barred can’t be opened at all. Yes, I know that’s not how “all doors work in real life”, but once again we’re talking about a game here.
Another huge problem with affordance is loot on map. It can be very hard to figure out which objects in the game can be picked up and which are just decoration. This is very problematic especially for new players as they don’t know each item’s shape by heart; you simply have to point at everything to test if you can pick it up.
The problem is even worse as multiple locations on the maps contain consumed provisions (food and drink) as decoration. Distinguishing unused provisions from used one is almost impossible unless they happen to be sitting next to each other. The levels are designed so that most of the time pickable items are on shelves or tables, not on ground, but this isn’t followed consistently.
For provisions the obvious solution would be to make the used provisions look really used with color and shape. Placing used provisions in the game world is good for immersion, so I wouldn’t go as far as removing them.
For other items this is a bit more problematic. The junk decorations exist both for artistic reasons and to make loot not stand out in an unnatural way, so removing them would probably cause more harm than good.
Currently many loot items can be distinguished from decorations by their more detailed 3D models if you look very closely, and this effect is more visible if there’s hard light in the room. Going a bit further with this type of highlighting would be a good solution for the problem; the loot doesn’t need to shine like AK’s muzzle flash, but enough to make it stand out while walking. A good compromise would be to increase the effect the closer you get, as player being close to an item usually indicates an intention of looking for loot.
The general hit area for picking items and opening containers should also be increased, as it usually takes unnecessary amount of time to find the right spot especially in containers (and of course all containers can’t be even opened).
One thing that really deserves a position mention is Tarkov’s button mapping. Not only are you able to remap the actions to different keys, but you’re allowed to specify the type of press as well: press of button, double-press of button or press-and-hold. This allows you to, say use press-and-hold G for throwing a grenade in case you have butter fingers (yes, you do get killed by your own ’nades).
There are two problems with the ergonomics of button actions tho.
First one is the something that might be classified as a bug: free look while aiming down the sights. This is a widely known exploit that enables the player to look around with the zoom level of equipped scope by aiming to their feet and then using the free look action to move the player character’s head up. The required button press is right-clicking the mouse and then dragging the middle mouse button. Considering the players pretty much need to use this exploit in current meta, it would be best for everyone to just remove/fix it.
The aiming down the sights + free look exploit in action.
Another one is press-and-hold action for performing certain actions such as hiding mission items on specific locations. This requires the player to keep the action button (F by default) constantly pressed for almost 30 seconds which can be a motoric challenge even for a person with no disabilities. Actions that take more than 4 seconds should be replaced with a toggle action – one button to start the action and another to cancel it, if needed. Making the action less physically demanding doesn’t remove the stress element of the design but it will make it less painful, literally.
Escape from Tarkov has so far managed to keep itself somewhat fresh by ruthlessly wiping (completely reseting) all player accounts every six-or-so months. While this allows the players experience the missions and early parts of the game again, it’s obvious the developers can’t make this a part off their long-term strategy. Whether the game will transform into a game-as-a-service type platform or a DLC-funded experience with multiple characters, addressing the numerous UX issues is something that simply needs to be done.
As stated in the beginning, a large majority of the players are veterans who have already learned workarounds for most of the quirks in the game, to a degree that some even take pride in building the skills needed to overcome the UX problems. Any large changes would probably be met with resistance, after all improving the UX would mean changing aspects of the game the pro players have already gotten used to, but that would only account the loudest 10% of the player base who have memorized every single bush and door in the maps.
Some of the improvements suggested in this article might pose technical problems, but that’s how game development in general works. So far the developer’s focus has been on content and technical stability, so who knows if UX will soon get the resources it needs, making Tarkov the gem is was destined to be.