I gave this a negative review on Steam in Feb 2020 and had every reason to never come back, which felt like a tragedy because I bought five copies of it so I could play with my wife and kids. The moment-to-moment gameplay was fun an engaging (which made it so we put in ~300 hours and felt like we got our money's worth), but the macro game was chaotic and genuinely soul-crushing and we just had to stop playing.
Out of sheer curiosity last week, I read some of recent the patch notes from after I quit playing, and to my utter surprise you had addressed every concern I had in my review, so I took it down and switched it to positive. There is still a long ways to go, but this last year has made me optimistic about ATLAS again.
I gave a fair amount of the new content a try-out in single player and creative mode so I could see how things would work and play out and I'll give PvP and PvE a try when I have time. I'd like to start off with some compliments about this game, so you guys can get some insight into what you're doing well and what makes ATLAS fun.
1. It's beautiful. Your modeling, design, effects, everything is just absolutely stunning. I like that it's UDK as well, because it means my kids can play it on their cheap-o computers and enjoy it with me on my monster gaming rig. I'm a bit of a sucker for pretty scenery and your game is very nearly at the top of my list for well-crafted beautiful scenery (you're honestly going toe-to-toe with Guild Wars 2 and Breath of the Wild here)
2. I love the volume of content, and honestly can't get enough of it. I love sailing out to a new zone because I need aged wood, or a different kind of fiber. If you added a dozen new types of fiber, and trees (hopefully with unique models to match), I would absolutely take my time to go exploring to find them all. It makes exploring feel emotionally rewarding, and makes me feel content with the time I spend in game. So long as I avoid the wiki, it's just impossible for that to get old.
3. The micro mechanics of the game are fun. Sailing is engaging and I feel like I'm rewarded (with lower repair costs, and combat advantages) for being good at it. Base building is fun to place and design your outpost, so much so that I am motivated to go gather more materials for it. Building and designing a ship makes me feel invested in its survival.
I mention these first because I feel like understanding what you do well helps you play to your strengths as you explore new mechanics, and code, and features.
So, here are my observations and suggestions. I've put a lot of thought into these so I hope you see this for what it is: someone who is absolutely rooting for you and your game.
1. Individual player interactions with devs is bad. I know most people in engineering are not sociable by nature, but this is currently the #1 point plaguing your steam reviews. Get a dedicated support person. They can write bug reports for you, interface with customers, make social media posts and do light advertising for you—in short, get a community manager. A good one will pay for themselves. Have someone help players feel like their concerns are being addressed because my personal interactions with your support was. . .well, lacking to say the least. Your devs are busy, resources are tight, and I get it, but I truly believe this should be your next financial investment so you can start building the community around this game, as opposed to most of us just dying off after a while of playing because we're frustrated. If you already have a community manager, then I am not seeing any real benefit from them and you should look at one with a different skill set.
2. Early access is a generally poor decision for a game in the MMO genre. Wipes are needed, frequently, in Early Access, stuff gets lost—it's part of the social contract. However, MMOs require a huge time commitment from their players, and many MMO players are ready and willing to devote entire days to playing their game of choice all in the pursuit of character development and in-game wealth. Because of this, early access will naturally poison your developer/player ecosystem.
There are few remedies to this that have worked inside of the MMO genre. You already tried the first one: wipes means newer players have a better chance at claiming an island, but it disenfranchises more established players and the ones who are genuinely desperate for their friends to convert over to their MMO of choice. The few MMOs who have tried this eventually die off. I think the best approach is two stages: public test realm, and tailored migrations.
Public Test Realms are inherently imbalanced and indifferent to exploits. Players can start with nothing, mid-game stuff, or late-game stuff whenever they want and play with other people in a public arena. This fixes some of the problems with Early Access because a player can always just hop on PTR and play the parts they enjoy without the typical MMO investment. They can kind of do this on single player, but, yeah, tether distance makes that not-viable and it limits the content they can explore. More importantly, not having it means that the only place you have to test experimental features and new ideas is your production public servers that some players have invested close to 1000 hours into.
See the problem?
Having a PTR means you can test a feature that is risky with little to no risk to your players—like having character migrations from one version to another. I hope this is a simple thing (especially since ARK already had the player transfer and dino upload, so there is a kind of migration path forward there). It's disheartening to lose your Galleon that you spent hours upon hours making, upgrading, designing, filling with crew, and became your virtual home. Really, making it so crew and ships can be stashed before a wipe would mean the world to established players, and the key feature here isn't for the players at all, but for the devs down the road. I know external storage isn't exactly conducive to UDK, but solving that is probably essential for an MMO in the long term.
The PTR also gives you the mechanism for gently removing broken game mechanics, and for offering consolation prizes to users who invested in a mechanic that ultimately doesn't work with the final game (which items often become collectables down the road). More importantly, having this would open up the user base to the more common MMO revenue streams, like paid for cosmetic skins and titles. Without having a culture and expectation of migration and data preservation, attempting to introduce either of those would likely incite a riot or boycott.
I think this two-stage approach will help fix most player complaints and problems much quicker, and to some extent would address all the other suggestions or problems listed after this.
3. The servers feel generally empty. Even when I make it to another player's base, it feels like I'm walking through a museum. Having more players and well-established companies help alleviate that, but some of this needs to be put into the feel of the game. A simple fix would be to add in freeport style NPCs as purchasable upgrades for player bases. This is a sensible thing to do with the upcoming trade system updates, you're already putting in this kind of effort, why not apply some of it more generally? It's a small thing and it may not seem like it's worth the effort, but taking the time to add in animation trees and some rudimentary patrol paths for NPC characters would make the whole game feel much more dynamic. Something like more noticeable idle animations on vendors, and if you're really feeling ambitious, having them respond to the animals nearby could help Atlas feel more like a living world of exploration, rather than an archeological dig. Even if it's just a meaningless animation, a harvester NPC from the farm would go a long way to fixing this feeling.
One suggestion I have that could also help this, which the game seems well suited for is borrowed from an ultimately failed MMO, Shadowbane. In this particular game, any skill or ability past level 20 was simply not accessible until you joined a clan. Player-made cities were the hubs of trade as a result, and everyone was intrinsically motivated to clan up. It made for some very fun social play, and since you already have player base building as a core mechanic, why not extend that into a social fleet building? I know a lot of players would moan about it, but if you think about it critically, does it really make any sense that murdering seagulls gives me the ability to craft a large speed sail? Learning skills from well-established player ports would create incentive for an entirely new kind of exploration—other people. That particular mechanic will never get old and will require a lot less dev time to develop (still needs some care and attention so it doesn't develop into a toxic community). I think the best way to implement this would be to have high-level skills (basically anything that gets upgraded in the current skill tree) come from tome drops that you need a port-specific NPC to decode to gain the skill (in addition to regular level requirements). I'd break up “Esotery of building” to individual building tome skills as well, so that individuals are inclined to clan up with strangers to get access to ports and companies are motivated to take in strangers because they have tomes/skills not found in the clan. As fun as it is to have a character that can do it all and be it all, MMOs just don't really need that. Everything should be split up into specialties and only the most dedicated players who are willing to create multiple toons for their account should be able to do that.
***This is a side note for future reference*** What killed Shadowbane, apart from bad business decisions and the market just generally moving onto the next-gen, was that clanning up was to heavily incentivized. In Shadowbane you could capture other clans cities (which took typical MMO levels of effort to build in the first place), and make them swear loyalty to you. The result was that one or two players would become the effective emperors of the MMO and it would eventually work it's way out of a PvP environment. The only solution that they could get to reliably work was to force a reset and everyone would have to start over periodically. Even putting in a taxation system only reinforced this dreaded cycle, because emperors had more money and resources and it would snowball into server-wide kingdoms. Putting constraints on it didn't work either, because they'd just have one of their long-term clan mates control a not-officially-affiliated clan and they'd just do tandem raids and city captures. Knowing their pitfalls can give you a better chance at preventing this, and some of the pirate theme could help. Reversing it so that each player-captured base requires upkeep from the conqueror seems a pretty sensible constraint, and allowing captured companies to mutiny after a certain time has passed would probably be enough. History is probably the best educator for this, and the factors that led to the collapse of the British Empire are probably your best bet (expense for making war to reclaim rebellious colonies, and just sheer distance involved in punishing those who rebelled, which you can easily emulate in the ATLAS environment) *** end note ***
4. It's been mentioned at great length, but the gold cost for building a ship just doesn't make sense. I could see justification for it at 1/10th the current rate, but in its current form, it's just nonsense. Honestly, you should hot-fix that out ASAP.
I think an alternative to it that achieves what you were after would be a patron/loan system for a ship, where you can rent one and get a commission as a privateer. Make it the start of your power stone quests and in true pirate/privateer form make players obligated to fulfill the quests of the commission, otherwise your patron hunts you down for failing to pay your debt. Existing random-generated maps are a great first draft implementation, and you could expand it to doing trade hauls as you build out that system, that way you don't just increase your server load without increasing your player count. Formalizing this and enforcing it via mechanics could even make it possible for existing players to sponsor new players and give them a commission and a ship (which you'd want the rewards to be essential to unlock certain game mechanics to get the seasoned players onboard).
Now for the last problem on my list.
5. Persistence and Instances. There's no possible way everyone who plays ATLAS can get their own island, and even trying to find a good spot for a PvE base can be challenging. It's also a core gameplay mechanic, and a fun one. The decay timer solves your needs and alleviates this problem to some extent, but it puts a time bomb into our experience. We know we'll lose everything sooner or later, contrast that to other MMOs I've played where I come back after 3 years and I'm filled with nostalgia at my armor, gear, and items in my bank. I will never have that with your setup, and I am honestly concerned that the game will never make it out of Early Access for this reason.
I also need base of operations on land. I think armored docks are a step in the right direction, but I think the ultimate solution is instancing and zone specification. I don't see any real technical reason why H3 couldn't be a PvE zone, and H4 a PvP. It's good to have that blend so players can kinda pick and choose what kind of game they want to play that day without having to maintain two separate characters. It gives the Pkers more prey, and it enables PvE-centered players (like myself) to be careful, wary, and pick up some PvP skills along the way without fully committing to the pwning noobs cutthroat environment common to PvP games. It creates some excitement and keeps me from getting bored from staying in my comfort zone to long. Risk vs Reward becomes the player's guide, and that's something you can tune behind the scenes to ensure you have a welcoming and engaging gaming community. But, instancing could work just as well. It's a cheap and dirty way to make the world have less boundaries, and would make most players indifferent to another world-size shrink. In my head, for ATLAS, it would look something like a zone labeled “the unknown.” As soon as you sail into it, you're given some navigation options. “Use a chart” could take you to a specific player/company base (sharable amongst friends of course) charts, “Follow the stars” could take you to a random player base of someone who is currently online, “Pillage” would take you to a player base that's specifically in a PvP capture window, and “Explore” would generate a random bare island.
Of course, to save on costs, you'd want to limit the “random” islands to the ones you've already made, and so it doesn't become to comfortable make them only spawn base_ resources. That way, you could limit what you store in an economically IOP provisioned database, and map it to static assets. Long load times are acceptable for these regions, and a single island could be smaller in size so more would fit on a single host, because they will implicitly spend a long time in them and it would be something they could actively choose to go to, and you could put the servers you're running or hosting it on in a cloud autoscale group with a shared data drive that loads the instance when a player travels to it. That way your costs better match your usage.
If you're willing to take the risk, you could even have this data exclusively stored and hosted client-side (so it really only is available when a company member is online), and just put in constraints on what resources can leave the region (like, players have to register a ship and cargo manifest before it can leave to ensure nothing impossible or unlikely makes its way into the public, and the item count and value could put a throttle on how long it takes for the system to approve their manifest). That lowers your operations costs, and then it merges the three different systems you're maintaining: PvP, PvE, and single player. It really is just about which resources you have at your disposal, more Dev, or more Ops? If you have more Ops, you probably want the autoscale group with shared data drive.
Anyways, sorry for the long read, hope some of this is helpful.
I gave this a negative review on Steam in Feb 2020 and had every reason to never come back, which felt like a tragedy because I bought five copies of it so I could play with my wife and kids. The moment-to-moment gameplay was fun an engaging (which made it so we put in ~300 hours and felt like we got our money's worth), but the macro game was chaotic and genuinely soul-crushing and we just had to stop playing. Out of sheer curiosity last week, I read some of recent the patch notes from after I quit playing, and to my utter surprise you had addressed every concern I had in my review, so I took it down and switched it to positive. There is still a long ways to go, but this last year has made me optimistic about ATLAS again. I gave a fair amount of the new content a try-out in single player and creative mode so I could see how things would work and play out and I'll give PvP and PvE a try when I have time. I'd like to start off with some compliments about this game, so you guys can get some insight into what you're doing well and what makes ATLAS fun. 1. It's beautiful. Your modeling, design, effects, everything is just absolutely stunning. I like that it's UDK as well, because it means my kids can play it on their cheap-o computers and enjoy it with me on my monster gaming rig. I'm a bit of a sucker for pretty scenery and your game is very nearly at the top of my list for well-crafted beautiful scenery (you're honestly going toe-to-toe with Guild Wars 2 and Breath of the Wild here) 2. I love the volume of content, and honestly can't get enough of it. I love sailing out to a new zone because I need aged wood, or a different kind of fiber. If you added a dozen new types of fiber, and trees (hopefully with unique models to match), I would absolutely take my time to go exploring to find them all. It makes exploring feel emotionally rewarding, and makes me feel content with the time I spend in game. So long as I avoid the wiki, it's just impossible for that to get old. 3. The micro mechanics of the game are fun. Sailing is engaging and I feel like I'm rewarded (with lower repair costs, and combat advantages) for being good at it. Base building is fun to place and design your outpost, so much so that I am motivated to go gather more materials for it. Building and designing a ship makes me feel invested in its survival. I mention these first because I feel like understanding what you do well helps you play to your strengths as you explore new mechanics, and code, and features. So, here are my observations and suggestions. I've put a lot of thought into these so I hope you see this for what it is: someone who is absolutely rooting for you and your game. 1. Individual player interactions with devs is bad. I know most people in engineering are not sociable by nature, but this is currently the #1 point plaguing your steam reviews. Get a dedicated support person. They can write bug reports for you, interface with customers, make social media posts and do light advertising for you—in short, get a community manager. A good one will pay for themselves. Have someone help players feel like their concerns are being addressed because my personal interactions with your support was. . .well, lacking to say the least. Your devs are busy, resources are tight, and I get it, but I truly believe this should be your next financial investment so you can start building the community around this game, as opposed to most of us just dying off after a while of playing because we're frustrated. If you already have a community manager, then I am not seeing any real benefit from them and you should look at one with a different skill set. 2. Early access is a generally poor decision for a game in the MMO genre. Wipes are needed, frequently, in Early Access, stuff gets lost—it's part of the social contract. However, MMOs require a huge time commitment from their players, and many MMO players are ready and willing to devote entire days to playing their game of choice all in the pursuit of character development and in-game wealth. Because of this, early access will naturally poison your developer/player ecosystem. There are few remedies to this that have worked inside of the MMO genre. You already tried the first one: wipes means newer players have a better chance at claiming an island, but it disenfranchises more established players and the ones who are genuinely desperate for their friends to convert over to their MMO of choice. The few MMOs who have tried this eventually die off. I think the best approach is two stages: public test realm, and tailored migrations. Public Test Realms are inherently imbalanced and indifferent to exploits. Players can start with nothing, mid-game stuff, or late-game stuff whenever they want and play with other people in a public arena. This fixes some of the problems with Early Access because a player can always just hop on PTR and play the parts they enjoy without the typical MMO investment. They can kind of do this on single player, but, yeah, tether distance makes that not-viable and it limits the content they can explore. More importantly, not having it means that the only place you have to test experimental features and new ideas is your production public servers that some players have invested close to 1000 hours into. See the problem? Having a PTR means you can test a feature that is risky with little to no risk to your players—like having character migrations from one version to another. I hope this is a simple thing (especially since ARK already had the player transfer and dino upload, so there is a kind of migration path forward there). It's disheartening to lose your Galleon that you spent hours upon hours making, upgrading, designing, filling with crew, and became your virtual home. Really, making it so crew and ships can be stashed before a wipe would mean the world to established players, and the key feature here isn't for the players at all, but for the devs down the road. I know external storage isn't exactly conducive to UDK, but solving that is probably essential for an MMO in the long term. The PTR also gives you the mechanism for gently removing broken game mechanics, and for offering consolation prizes to users who invested in a mechanic that ultimately doesn't work with the final game (which items often become collectables down the road). More importantly, having this would open up the user base to the more common MMO revenue streams, like paid for cosmetic skins and titles. Without having a culture and expectation of migration and data preservation, attempting to introduce either of those would likely incite a riot or boycott. I think this two-stage approach will help fix most player complaints and problems much quicker, and to some extent would address all the other suggestions or problems listed after this. 3. The servers feel generally empty. Even when I make it to another player's base, it feels like I'm walking through a museum. Having more players and well-established companies help alleviate that, but some of this needs to be put into the feel of the game. A simple fix would be to add in freeport style NPCs as purchasable upgrades for player bases. This is a sensible thing to do with the upcoming trade system updates, you're already putting in this kind of effort, why not apply some of it more generally? It's a small thing and it may not seem like it's worth the effort, but taking the time to add in animation trees and some rudimentary patrol paths for NPC characters would make the whole game feel much more dynamic. Something like more noticeable idle animations on vendors, and if you're really feeling ambitious, having them respond to the animals nearby could help Atlas feel more like a living world of exploration, rather than an archeological dig. Even if it's just a meaningless animation, a harvester NPC from the farm would go a long way to fixing this feeling. One suggestion I have that could also help this, which the game seems well suited for is borrowed from an ultimately failed MMO, Shadowbane. In this particular game, any skill or ability past level 20 was simply not accessible until you joined a clan. Player-made cities were the hubs of trade as a result, and everyone was intrinsically motivated to clan up. It made for some very fun social play, and since you already have player base building as a core mechanic, why not extend that into a social fleet building? I know a lot of players would moan about it, but if you think about it critically, does it really make any sense that murdering seagulls gives me the ability to craft a large speed sail? Learning skills from well-established player ports would create incentive for an entirely new kind of exploration—other people. That particular mechanic will never get old and will require a lot less dev time to develop (still needs some care and attention so it doesn't develop into a toxic community). I think the best way to implement this would be to have high-level skills (basically anything that gets upgraded in the current skill tree) come from tome drops that you need a port-specific NPC to decode to gain the skill (in addition to regular level requirements). I'd break up “Esotery of building” to individual building tome skills as well, so that individuals are inclined to clan up with strangers to get access to ports and companies are motivated to take in strangers because they have tomes/skills not found in the clan. As fun as it is to have a character that can do it all and be it all, MMOs just don't really need that. Everything should be split up into specialties and only the most dedicated players who are willing to create multiple toons for their account should be able to do that. ***This is a side note for future reference*** What killed Shadowbane, apart from bad business decisions and the market just generally moving onto the next-gen, was that clanning up was to heavily incentivized. In Shadowbane you could capture other clans cities (which took typical MMO levels of effort to build in the first place), and make them swear loyalty to you. The result was that one or two players would become the effective emperors of the MMO and it would eventually work it's way out of a PvP environment. The only solution that they could get to reliably work was to force a reset and everyone would have to start over periodically. Even putting in a taxation system only reinforced this dreaded cycle, because emperors had more money and resources and it would snowball into server-wide kingdoms. Putting constraints on it didn't work either, because they'd just have one of their long-term clan mates control a not-officially-affiliated clan and they'd just do tandem raids and city captures. Knowing their pitfalls can give you a better chance at preventing this, and some of the pirate theme could help. Reversing it so that each player-captured base requires upkeep from the conqueror seems a pretty sensible constraint, and allowing captured companies to mutiny after a certain time has passed would probably be enough. History is probably the best educator for this, and the factors that led to the collapse of the British Empire are probably your best bet (expense for making war to reclaim rebellious colonies, and just sheer distance involved in punishing those who rebelled, which you can easily emulate in the ATLAS environment) *** end note *** 4. It's been mentioned at great length, but the gold cost for building a ship just doesn't make sense. I could see justification for it at 1/10th the current rate, but in its current form, it's just nonsense. Honestly, you should hot-fix that out ASAP. I think an alternative to it that achieves what you were after would be a patron/loan system for a ship, where you can rent one and get a commission as a privateer. Make it the start of your power stone quests and in true pirate/privateer form make players obligated to fulfill the quests of the commission, otherwise your patron hunts you down for failing to pay your debt. Existing random-generated maps are a great first draft implementation, and you could expand it to doing trade hauls as you build out that system, that way you don't just increase your server load without increasing your player count. Formalizing this and enforcing it via mechanics could even make it possible for existing players to sponsor new players and give them a commission and a ship (which you'd want the rewards to be essential to unlock certain game mechanics to get the seasoned players onboard). Now for the last problem on my list. 5. Persistence and Instances. There's no possible way everyone who plays ATLAS can get their own island, and even trying to find a good spot for a PvE base can be challenging. It's also a core gameplay mechanic, and a fun one. The decay timer solves your needs and alleviates this problem to some extent, but it puts a time bomb into our experience. We know we'll lose everything sooner or later, contrast that to other MMOs I've played where I come back after 3 years and I'm filled with nostalgia at my armor, gear, and items in my bank. I will never have that with your setup, and I am honestly concerned that the game will never make it out of Early Access for this reason. I also need base of operations on land. I think armored docks are a step in the right direction, but I think the ultimate solution is instancing and zone specification. I don't see any real technical reason why H3 couldn't be a PvE zone, and H4 a PvP. It's good to have that blend so players can kinda pick and choose what kind of game they want to play that day without having to maintain two separate characters. It gives the Pkers more prey, and it enables PvE-centered players (like myself) to be careful, wary, and pick up some PvP skills along the way without fully committing to the pwning noobs cutthroat environment common to PvP games. It creates some excitement and keeps me from getting bored from staying in my comfort zone to long. Risk vs Reward becomes the player's guide, and that's something you can tune behind the scenes to ensure you have a welcoming and engaging gaming community. But, instancing could work just as well. It's a cheap and dirty way to make the world have less boundaries, and would make most players indifferent to another world-size shrink. In my head, for ATLAS, it would look something like a zone labeled “the unknown.” As soon as you sail into it, you're given some navigation options. “Use a chart” could take you to a specific player/company base (sharable amongst friends of course) charts, “Follow the stars” could take you to a random player base of someone who is currently online, “Pillage” would take you to a player base that's specifically in a PvP capture window, and “Explore” would generate a random bare island. Of course, to save on costs, you'd want to limit the “random” islands to the ones you've already made, and so it doesn't become to comfortable make them only spawn base_ resources. That way, you could limit what you store in an economically IOP provisioned database, and map it to static assets. Long load times are acceptable for these regions, and a single island could be smaller in size so more would fit on a single host, because they will implicitly spend a long time in them and it would be something they could actively choose to go to, and you could put the servers you're running or hosting it on in a cloud autoscale group with a shared data drive that loads the instance when a player travels to it. That way your costs better match your usage. If you're willing to take the risk, you could even have this data exclusively stored and hosted client-side (so it really only is available when a company member is online), and just put in constraints on what resources can leave the region (like, players have to register a ship and cargo manifest before it can leave to ensure nothing impossible or unlikely makes its way into the public, and the item count and value could put a throttle on how long it takes for the system to approve their manifest). That lowers your operations costs, and then it merges the three different systems you're maintaining: PvP, PvE, and single player. It really is just about which resources you have at your disposal, more Dev, or more Ops? If you have more Ops, you probably want the autoscale group with shared data drive. Anyways, sorry for the long read, hope some of this is helpful.