Scavgame, similar games, game design 2024-07-23

I don't even know what I've been doing since I last updated this website. All I remember is that for some reason the work I do for a living has been more fun, maybe it's because I managed to get in some important reworks to the thing I work on, and learned some new ways to sidestep the terrible systems my boss insists on using, so a lot of friction got removed. It's still not at all what I'd personally do, but for what it is, it's just a lot less frustrating now.

But I think more interestingly, I recently picked up a project from last year and got it to a point where I felt like making a page for it: Scavgame. I started writing a short analysis of other similar games into Scavgame's own page, but it bloated into this monstrosity so I made a diary about it instead.

The closest resemblance to Scavgame is ZERO Sievert. I have a lot of issues with that game and have done some extensive modding for it for personal use, I think of it as a very badly balanced game with critical game design flaws and disappointing missed opportunities. ZS is a lot more arcade-y, Scavgame is meant to more detailed and "realistic".

My biggest problem with ZS is the bad sense of progression. You can get one of the best weapons (although without attachments) by killing some random enemies in the first map, and then you'll never have a reason to use another weapon again. There's a lot of weapons in the game but you'll never use anything other than a few assault rifles because they're so much better in every way. They "fixed" that problem by introducing an incomprehensibly stupid "effective range" mechanic for weapons that causes your bullets to do 0 damage if the enemy is either too close or too far. Absolutely nobody seemed to like it but the developers refused to address it until recently when they announced that in the next update you can turn it off. There's many ways to fix the problem of weapon balance, many of which I'm planning to add to Scavgame (all weapon types will have noticeable advantages and drawbacks), but it seems like they just gave up after the first attempt.

Another bad addition they made was a field of view system. They always had a field of view in the sense that you can't see behind walls, but now they also have a 90 degree cone in front of you and you can't see outside of it. I've seen many games with features that have something interesting about them, but the developer doesn't quite know what that is so they design the feature poorly. As an example, the interesting part of a field of view system is that you can't see behind you, not that you can only see in front of you. Those sound like the same thing, but there's a big difference. ZS's narrow field of view seems to focus on the idea of "only seeing in front of you", and for that purpose the feature does what it should. However in Scavgame I'm focusing more on the idea of "you can't see behind you", so I made the default field of view cone wider than 180 degrees, and it only narrows when you're aiming your weapon (it narrows more depending on how far your weapon's "scope" can see).

There's many observations I made from trying to make ZS more fun, like that it's much more fun to use items as you find them rather than bringing in everything you need and slowly upgrading the same weapon. I found it significantly more fun to play the industrial map in a way where you try to stay there as long as possible and trade with the trader there, not just repeatedly running in and out with loot.

Another game very similar to ZS that I just tried out a week or two ago (2024-07) is Night Raider. On the surface it looks like a half assed ripoff of ZS, but it's quite a bit different, and surprisingly actually has superior game design in several places. It is much further from Scavgame than ZS in many ways, but also closer in other ways, I was surprised at how similar their idea of using keys and storage containers is compared to what I've been doing. I also like their idea of requiring you to find special consumable documents in order to enter higher tier maps. I find NR to be far too easy, it's only in the second-last map that I started to feel challenged in any way, and only by the events where waves of zombies charge at you.

Something I've had a very strong vision for Scavgame, ever since I saw it in Escape From Tarkov, is to add a lot of consumable key items that allow you to access more valuable locations, maps and buildings and individual objects alike. I want the world to be filled with locks that you can find keys for because I find the idea of finding a key and then using it to enter a valuable location to be very appealing. This indeed turned out to be a very appealing part of NR, but unfortunately it feels too easy to find keys and many of them just have extra generic containers behind them so it's not as satisfying to use them as it could be.

A combination of the industrial map in ZS that has traders in it, the "rare" documents in NR that are required to enter maps, and what I've heard from some Tarkov streamers, have given me an important realization about these kinds of games: they're a lot more fun when you have a reason to stay in the map longer. They can easily become boring and repetitive if you just go in and then leave as soon as your inventory is full. I think this is partly because you don't get to interact with the game's mechanics in an interesting way, you just restock in your stash and then repeat your loot run. But if you stay for longer, you have to start managing your hunger and health and armor and ammo and weapon durability IN THE GAME, you have to start looting useful consumables and using whatever you can find, prioritizing your actions based on what you're lacking. I think it's a very significant difference that turns it into a different game, and I'm hoping to capture the latter in Scavgame.

Supposedly the developer of NR is going to change the way the map documents work by adding them all to a trader and locking them behind progression. I think that might be a mistake, if you can just buy them then the feeling of wanting to make the most out of your raid goes away. Why work hard to loot convenience stores and restaurants for food and energy drinks to stay in the map longer when you can just leave and come back any time? Why try to challenge a difficult location while your resources are lacking, why go to the police station to get more ammo and open police cars for weapon bags if you can just leave and reset your stuff in your stash and come back? But who knows, maybe it's alright to progress through quests in order to unlock higher tier maps, but I'm concerned that it's the wrong decision.

NR's weather system is much better than ZS's. You can use items to survive them or hide and wait it out in a house, ZS's toxic rain just forces you to exit the map if your armor is insufficient. The special weather conditions in NR are much too rare and lasts much too short of a time though, it only becomes an issue if it starts right when you're leaving a building. I saw old patch notes where they shortened the duration of weathers, and that too was probably a mistake.

I found it interesting how NR has an absolute metric boatload of different items. When you think you've surely seen them all, there's still a bunch more. Almost 100% of them have no use beyond being sold, but I found it surprisingly appealing to research what items are valuable. If there's one idea I got out of NR (NR doesn't do this, I just had the idea), it's that what if you could edit item descriptions? You could just type your own notes to items, like how many you have, what places you've found it from, what it's value is, what crafting recipes use it, how many you need for a quest, etc. It would be much more convenient than writing notes on a notepad or whatever.

One thing that I consider the biggest difference between those games and Scavgame is that I want Scavgame to be more "realistic". I don't exactly mean realism, an example of what I mean is that I plan my weather system to have various states that slowly transition from one to another. In both ZS and NR, the weather just turns on and off at random. Similarly, I want items to have more immersive use cases, for example if you happen to find a rare blowtorch and fuel, you can use them to break some containers that otherwise would require a key. A large explosive can be used to either kill enemies or to blow up walls. You can cook crappy meat on a campfire but better foods require you to find a stove somewhere. Gunshots may attract enemies from far away which makes silencers valuable. That kind of things, It's realistic in the sense of it there being more detail or thought involved, I'm not sure how to describe it in words. Realism? Groundedness? Immersion? Those words come to mind when I try to describe it.

Stalker Anomaly is similar in a different way, although it's an FPS. Anomaly is seeping with that kind of grounded immersive feeling that I really love, but it's very hard to pinpoint what makes it that way. It really pulls you into the world and makes you care about existing in it. I have more problems with Anomaly than I can count and I've also done extensive modding for it, but it's too hard to do anything that I really want. I have 2 critical problems with Anomaly. First is that the progression is totally messed up; the beginning is an agonizing exercise of masochism, but as soon as you get a gun with a scope you'll start dropping hordes of highly geared enemies with ease. The second problem is the lack of quests, I really liked doing the various quests in the official Stalker games, but Anomaly has almost nothing of the sort, just a trail of "go here" non-quests that basically just tell you to go north.

Stalker isn't procedurally generated in any way except slightly changing loot stashes. I'm considering treating Scavgame as 2 separate games; a story mode with static maps and very specific quests, and a procedurally generated sandbox mode. I'm still not 100% set on which one I want, but most likely I'll do procedurally generated maps. My goal is to design a map by hand, and then try to get all the key details through procedural generation. For example, say, a factory and a windmill that are connected by a long winding road, and wheat fields with a crashed plane along the road, and random enemies in a house nearby are aiming their guns towards the plane. As long as you have the hand-made details of the map you want, you could then think of how to generate it procedurally so it'll always be the same map but differently made.

As a side note, one of the things that disappointed me the most about both ZS and NR is that they pretend to be procedurally generated games, but then all the buildings use a static template. I really wish they were properly procedurally generated so there was more variation in room layouts, it's boring when you learn the layout and just go straight to the thing you want instead of having to find it. There's no sense of discovery in the world, only in what's inside the containers.

Escape From Tarkov is basically the game that invented the "extraction shooter" genre. While I haven't played it myself, I watch some Tarkov streamers as background noise so I've absorbed many ideas from it before I even knew about the other aforementioned games. Tarkov is also very realistic and has a crazy amount of immersive detail to it, although I get the feeling it goes too far in some places.

The most notable comparison to Scavgame and the biggest inspiration that I drew directly from Tarkov is the weapon modding. Both ZS and NR have gun modding, but they're very static and not very deep, you just swap to a better part when you find one. I'm planning to make my guns a lot more dynamic, similar to how in Tarkov you can put attachments onto attachments, except not nearly as detailed and not based on real guns. I want to try to distill the system to only the interesting parts and not the needlessly complex parts, and I want the different parts to have noticeable advantages and drawbacks so there's no obvious best part to use. For example if you can make a short or a long version of a particular gun, both should have significant advantages over the other. NR actually does this decently well because the recoil is designed in a way where your aim goes off the rails if you use automatic firing mode for too long, so low recoil and high "stability" can actually be very powerful because they enable you to fire accurately faster and for longer periods, which is often even more useful than simply increasing damage. I made some balancing decisions that I had never made in ZS.

I have a bunch more ideas that can balance different weapon types, one of which is to modify things like accuracy and reload speed based on your current movement speed. NR has a little bit of that too, I was delighted to see that the "ergonomics" stat increases your accuracy while you move, but in practice it's not particularly useful because your accuracy is basically a 45 degree cone if you're not holding right mouse to aim your gun, and you move at 0.0001 meters per minute while aiming but 10000 km/h while running, so it's always better to just run and then stop to shoot rather than shooting while moving.