Feedback from a fellow game designer
Hi guys,
Took a bunch of notes today and wanted to dump em here. I'm a game designer myself, so I'll try to focus more on big-picture stuff rather than the small nitpicks and other problems that will work themselves out as you improve the art/sound/interaction/etc.. Overall, I think there are some very good innovations in the design, but I do have some concerns.
First off, I wanted to mention my initial impression, which was actually very negative. I was super excited for Atlas, but after watching Sean's intro video, I felt really let down. My idea of the game went from "squad-based RTS with lots of secondary objectives on the map" to "Oh, it's basically League of Legends with a splash of RTS elements". Now, I've probably played about 5k hours of Brood War and SC2, which I feel set the bar as far as RTS is concerned. I've never enjoyed MOBA-type games much, mostly because I feel that they're as wide as an ocean and shallow as a puddle... or rather because their depth comes mostly as a result of complexity rather than actual strategic depth.
I don't know how representative I am of a typical strategy game fan, but it felt crystal clear to me that your design goals were focused around appealing to the more casual MOBA fanbase, with heroes (exactly what I didn't like about Warcraft 3), and team-based games (ugh), and what I assume will be a free-to-play business model based around having a huge variety of heroes/squads available. These are all fine design choices, and I think you have a good chance of doing very well with them, but I personally felt very let down upon finding out about them.
The biggest disappointment for me personally is that there's no 1v1 at all... not because I hate playing with teammates, but because I've always felt that team-based RTS games are missing that chess-like exchange of pressure as you set up timing attacks, respond to your opponent's build choices, and restrict their options. The strategic aspects are heightened in 1v1 because you have one set of hands providing all the instructions.
Now, after playing the game, I was pleasantly surprised that it actually felt pretty good. The agario/travelling salesman mechanic of gem collection is interesting. I actually wish it were even more emphasized... i.e. 5x more gems, more frequently spawning, each worth much less. The mechanics are pretty cool. Execution is coming along well, though I could nitpick a bunch about pathfinding/sound/visual feedback/etc.. But overall I had a lot of fun.
That said, gameplay-wise, I still don't feel that it scratches that itch of a "real" RTS for me, and I doubt I would sink more than 100 hours into a game like this. The super simple tech tree and economy felt really shallow. I never once felt that I outplayed my opponents due to superior build order or upgrade choices. The game felt highly dependent on unit micro in a bad way, even worse than Warcraft 3 and much less like Brood War or SC2. No air units or cloaked units, so there's not much in the way of winning via tech. No simcity, no tower rushing, etc.. It just feels like all of the most interesting aspects of a real RTS have been removed—the aspects that provide a really high level of strategic depth. It feels more "arcadey", like one of those Brood War UMS maps where you just drag a civilian over to a beacon to select what to buy. Or maybe a bit like DOTA.
The game also feels a little "ballcraft-like" to me; I never had much of a reason to split my army to harass or multitask; in fact it mostly felt like a liability to not have all forces together at once. I never felt that my attention was a valuable resource; in fact I frequently found myself with nothing to do during the midgame in between gem spawns, especially if my guys were on cooldown. Maybe it's a good thing for many players if you don't need 200APM to play Atlas, but for me it was a strong negative—it felt boring.
Now, I'm not one of those folks who complained about multiple-building-select and control-groups-larger-than-12 in SC2. I don't necessarily think a good game is one with a lot of busywork. But even in SC2, I feel that there is constant pressure on my attention, a massive list of things to do, and I need to ration where I spend my time. I felt that the exact opposite was true in Atlas. I never had much to do, so I never felt compelled to play another game to make better use of my actions.
Hotkeys felt almost unnecessary... I hotkeyed my base so I could reset the rally point and my other structures so I could upgrade quickly, but I never felt I needed multiple control groups. Actually, the only thing I felt I needed was just a "select all units" button, as I would just stick everything in a single group and then add new units to it whenever they spawned. "Select hero" wouldn't hurt either as you frequently want your hero away from the rest of your group to play more Agario.
Usability-wise, the main problems I had were just not knowing where my guys were. I wish my units were a different colour on the map than my structures. I often had no idea which units were selected. The battlefield provides visual cues as to [cooldown, HP, selection] but it seems that they are in the wrong order of importance... Selection and HP should be the most important, but actually the cooldown seems the most visually emphasized. Personally I would prefer it if cooldown wasn't visually shown at all and your selection was much more emphasized. I played hydros a bunch and felt that the "selected units" panel from Starcraft was sorely missing because I couldn't heal the unit "whose portrait was most red".
I'm also concerned a little about some Yomi/RPS issues in the design. Squad selection is particularly tricky because if you end up in a bad match-up, it feels that there's not much you can do. E.g. playing Hydros vs Eris in the bottom lane felt pretty unfun, very reminiscent of orcs vs night-elf in WC3, but even worse because of the speed ability on Eris's side. Maybe it's OK if there's a balanced metagame, but 3v3 with blind picks seems like a risky choice to me if there are hard counters.
Ultimately, I think your success will depend on whether there's really an audience for this sort of game. I tend to enjoy strategy games that are as hardcore as possible and Atlas missed the mark by a lot for me. At the same time, many League fans who would prefer to just control one hero will also be put off. But I suspect that there are plenty of folks around for whom Atlas is an ideal balance of the two. It's mostly just a question of how big that middle ground is.
Anyways, that's as much as I can say for now. Overall I think you've done an excellent job and I do really like the gems—best innovation by far.
Comments
I agree with pretty much this whole thing. Especially the way units are used.
I feel like there's not a good reason to split your army at all. Any sort of army splitting seems like it would be solely based on team splitting (ex. my whole army is south, while my team-mates armies are north).
No air units and no ways to sneak behind enemy lines (without being tagged by towers) makes me feel pinholed into a single play style: Ball up army and fight other army, rinse repeat. And that might just be the nature of this game and its win condition.
Below is my take on some of the stuff you brought up. I agree with some parts, disagree with others, and there are also some general comments.
I spawned in the south location frequently playing as Eris. As a result, I felt that the early game was often packed with a very micro-heavy series of 1vs1 exchanges. Positioning, pressure, and early trading seemed to make a big difference in how quickly you could build up your economy early in the game.
I definitely don't feel the same urge to jump into another game and learn/practice more that I felt after a game of starcraft. The simplicity of the tech tree only became a problem for me because of how long the games were overall. A very simple tech tree and low unit diversity mostly started to feel bad because of 1) the emphasis on micro, and 2) in the late game you start to feel like you can autopilot most of your macro decisions, i.e. there's just not much to do.
On the other hand, I don't think the game is devoid of variety. Today I played games where tower rushing, turtling, and very actively tracking your opponents movement and punishing them with timing attacks each led to victory.
The dependence on the hero for construction/harvesting/combat prowess and the synergy between units seemed to play a big role in this. My units did a very poor job of applying or holding off enemy pressure when my hero went elsewhere to build/harvest something. Splitting my army also didn't seem very valuable in most situations. In general, you didn't have too many things pulling your attention in different directions.
Note: F1 selects your hero.
I've played a large amount of both SC2 and LoL, and while overall I appreciate SC2 more as a game, I don't think it's fair to say MOBAs are without depth. This game definitely has MOBA-like features: it's micro-intensive, there's little splitting your attention other than map awareness, and there's a lot to 'know'.
On the other hand, strategic depth comes from your ability to communicate and interact effectively with as part of a team where each member fulfills a different role. This is something these 'class-based' kinds of games have that starcraft was entirely without (no one reasonably argues that 4v4 games of starcraft were particularly 'deep'). A lot of games have started adopting this model, and it doesn't seem to be inherently bad or any less 'hardcore'. It's just 'different'.
Couldn't have said it better myself.
I agree with pretty much everything you said OP.
I assume that many of the alpha testers here heard about the game from Sean, and so I'm concerned that many of us who joined this alpha test were expecting the game to be more of an RTS because of his history with the genre. At least, that's my story. This is honestly the most innovative MOBA I've seen since the genre's inception, and the final product might be amazing, but personally, I'll still see it as "amazing for a MOBA."
My point is, if they want the game to be a MOBA, for better or worse they probably have a lot of RTS-leaning playtesters right now. We might be out of their target audience, and that's okay, but personally I'd love to hear what the dev team plans for the game. If they intend on keeping this micro-heavy MOBA style, I'd at least like to know so that I can stop pining for a game which they aren't making. Otherwise, I look forward to seeing their implementation of more macro systems.
Either way, I'm super excited to keep testing it. Even if they go in a direction which isn't "my kind of game," so far there's still a lot to love in the game, and it's only going to get better from here.
I don't have much to offer other than what's already been said, but I agree with all of this.
I don't agree with it at all.
This game doesn't have any consistent versus game modes. Not just no 1v1. I agree that it would be a fun play test but calling it the "Biggest Disappointment" during the Technical Alpha seems like you're a bit quick on the draw.
I think it's in poor taste to say MOBAs have no strategic depth and this doesn't either. Granted I haven't spent weeks playing this game against real opponents but the taste that I got left me with a very different feeling. I believe there is a lot of potential for variation to the gameplay. Sure you can just ball it up and send em down the middle but you won't get very far doing that. There is the opportunity to employ a bunch of different tactics such as flanking, resource denying, kiting, sandwiching (The most delicious strategy) still exist.
On top of that there are also the Macro Elements that while not as pronounced as the "Combat" still allow for a wide range of possible strategies to employ.
If you come at this game with the mindset that you'll be playing a hardcore RTS you're going to be disappointed if you don't broaden your horizons. Judge the game for what it is and less for what it isn't.
Cheers
I think I understand where you're coming from (op). I never played Brood War, only Starcraft 2. I played macro protoss for 3-4 years, and as I got better at the game I tried to optimize my builds and strategies so that I could win versus everything my opponents would do. What I ended up with were very precise builds that I spent hours and hours practicing in build order testers. When everything lined up well and I could successfully scout, react, and defend my opponent's attack and come out ahead, it felt really good. But when I messed up and got supply blocked when I needed to warp in my key defensive units, and I got steamrolled as a result, it didn't feel so good at all. Starcraft 2 wasn't a game that I could play when tired, or when out of practice, or even on a bad day.
I haven't played much Atlas yet, but I'm not sure that that is missing. I saw that people said that Hydros was weak, but I thought that it would fit my playstyle, so I tried it. I saw that people said basically to spend all of their money on Tier 1 units at the start of the game, so I started the game by upgrading to Tier 2. And it worked fairly well in all of my games yesterday. I'm not yet sure if this is actually a "viable" strategy, I'd need to see it tested several times between "top level" players, and I'm guessing that the game itself will continue to change, but it seems to be a possibility to me. I think that it's possible that multiple "builds" for every squad do exist, and that the differences between the builds could affect the way that you play versus your opponents. It could affect the way that engagements go, the way that positioning goes, etc.
In short, I think that it's possible that Atlas has build order depth, probably not as much as Starcraft 2 does, but still a meaningful amount. But the best thing is that you don't have to worry about making sure everything is 100% on time, you can execute your build well enough when you're tired, out of practice, on a bad day, etc. And your focus is on your army positioning and control. I don't think that a build order would necessarily give you a huge advantage, perhaps a small advantage, but different build orders could drastically change the way you play the game, and combined with allowing a weaker player like myself to actually execute a build order reasonably well, I think it's very promising. I wouldn't judge the lack of builds/strategic decisions behind them quite yet.
There's a few core themes I see here
Also, you write that
This is false. We began by asking "what is the biggest frustration that RTS players have told us about?" One of the most frequent comments was the "one battle decides it all" experience that leads to passivity, fear, and lack of interaction. So, we explored respawn and interaction mechanics that led us to where we are today! As far as I'm concerned, our goal now is to try to add interesting complexity to the game!
Right now a lot of people I've talked to have been under the impression that Atlas feels half way between a MOBA and an RTS, and that one of them will probably become the more dominant feeling as development continues. My impression of your last points is that you're intending to bring Atlas in a more RTS-ish direction? I'd personally love to see that direction being taken, and I think that a focus on implementation of increased multitask would be brilliant. Right now I get the vibe of a MOBA in an RTS framework and I feel as if multitask is a big part of that.
To put that last point in Day[9] Format™
Issue: Game doesn't have enough happening to keep me interested in the long term.
How it made me feel: Like I'm not being pushed
Potential solution: More multitask to differentiate Atlas from conventional MOBAS,and to keep the energy high.
Cheers for your reply @Artillery.Day[9]. Some follow-ups:
Definitely happy to hear this.
Precisely how you should interpret my comments. Not really trying to "judge it for what it is" but rather give you my honest impressions. Truthfully, I feel that my tastes are pretty far from the average so I wouldn't suggest you give my opinion much weight.
These mechanics are all great additions. My comment about the game seeming to intend to appeal to a more casual audience was more rooted in the lack of multitasking, lowered APM, simplified strategic decisions, lack of 1v1, and small army sizes. I had a moment of thinking... "hmmm, I really think this isn't the game for me" because it felt like most of the aspects of RTS that I found interesting and challenging had been simplified.
Whether or not this was your intention, the word that came immediately to mind was a "casualized" RTS.
For a player like me, it was a very negative first impression. I only saw the game more positively after playing a few rounds and seeing the benefits of some of the innovations (gem-collecting interaction, etc.). In a way this is kinda bad, because the aspects I found out before playing the game were the ones that made me not want to try it, and aspects I liked were mostly those I learned about after trying.
As somebody who really doesn't enjoy Dota/League/Heroes/etc. and finds the current state of RTS to be rather unfortunate, I was really excited about Atlas from what I had heard (and I've been following the project since I first heard about Artillery's Series A round back in 2012). So I was genuinely hoping that Atlas would be "the RTS of the future"... a game that finally breathes some life into the genre.
And well... it's not so much that I had high hopes that were dashed, but rather that I feel (as expected) that you're probably doing the right thing for your product by making something that eschews the more intimidating aspects of RTS, even if that choice doesn't align with my personal killer/spike/achiever/strategizer/hardcore preferences.
It's hard to find the right places to look for market cues. With SC2, Blizz realized far too late that Big Game Hunters and UMS needed to be front and center as multiplayer modes, because the emphasis on competitive multiplayer ultimately diluted the scene (though it made sense at the time as a calculated gamble motivated by the growth of e-sports). But note also that even the casuals in Brood War played all their games on FASTEST speed, and that the vast majority of online chess is blitz.
As long as you have a really defined vision of what you're trying to create, test the crap out of that vision (including market tests, e.g. we do tests with fake Google and Facebook ads) and communicate/execute your vision clearly, I think you will do well.
We learned a lot of really surprising things in player surveys too... our game is a competitive turn-based/RTS/card game hybrid, but e.g. our fans like Dark Souls a lot more than they like League or Civ. And they REALLY like the idea of raids (team PvE).
One comment on vision... particularly the emphasis on "removing frustration"... have you seen that case study where they removed natural disasters from Civ (they were the #1 most complained-about aspect of the game) and players complained less, but the average rating out of 10 dropped from an 8 to a 7 in a big A/B focus test?
That right there is why game design is hard.... ^_^
Also, not sure why it's not tagging your name right... seems it thinks I'm tagging a user named @Artillery.Day[9 because the final bracket isn't highlighted when I type @Artillery.Day[9]
Hahah I am untaggable! I've done it! Anonymity achieved :P.
I personally believe it's possible to address many concerns of the "casual" and the "hardcore" within the same game. For instance, I think I'm content with the things we've done to address the "casual" concerns with respawn & squads. Respawn removes the frustration of "one battle decides it all" and "I don't want to lose hope!" Squads removes the ambiguity of "what are units that work well together?" Now, I want to add tons of additional choices for players via economic impact and tech tree choices/upgrades etc. In this way, we hope to have a game that
Oh man I cannot agree more. I actually wrote an article about how games are fun because they are frustrating and inconvenient. There was a critical time where we had to choose which frustrations we were addressing and which ones we wanted to leave in/add. As noted above, I think we've address the core frustrations we want to, and it's only complexification from here!
Also, Prismata is a fascinating game and I'm thrilled to hear how the development continues. As you say, game design is unbelievably hard, so I hope you guys continue to crush it with your clever design :D
Yeah, that sounds great.
Something I've noticed in myself is that the initial experience of, say, binging tons of games repeatedly in a strategy game... it's usually motivated by finding flaws in my own decision-making and wanting to try again to correct those flaws (possibly by exploring a new branch of the strategy space, but not always). To really get that potato chip effect going, I need to feel like I made a wrong decision (and bad execution doesn't count).
Adding more strategy space shouldn't be hard (and already there is lots of "exploration space" in that sense with multiple heroes). But negative feedback could be a challenge. One of the feelings that I didn't have in Atlas was "oh, my build sucked". At least for me personally, that feeling makes a huge difference in whether I stick around for the first 50 games. And given that you can't get supply blocked and don't have to build production structures in Atlas, I think it will be a bit of a challenge to find ways of creating that feeling. The fact that nothing costs both gold AND gems also removes the type of negative feedback you'd get in SC from, e.g. getting too much gas, or too little. Upgrades are particularly weak at creating negative feedback, especially in non-mirror situations.
It reminds me of a lesson I learned early on when introducing individual people to Prismata... we learned that if they're a spike/hardcore gamer, they'll like the game way more if we 1v1 against them and try to really smash them in their first few games after they learn the rules. They always wanna try again if we do that. ^_^
Prismata to me is a super interesting case because it focuses so intensely on one specific thing that people like about RTS/board games/etc. and somehow that leads to people playing 1000 or even 10,000+ games of it in some cases. I still don't feel I fully understand why...
Gonna jump in here instead of making my own thread since I'm a few days late on the simple analysis, but I think there are some good things to focus on in this thread. For background, I played starcraft 2 from launch until HotS came out, pretty seriously, until frustration with the design and balance drove me to cut back. I picked up DotA about 7 months ago and have a fundamentally different love/hate relationship with it after almost a thousand hours of play. I originally thought I disagreed with Elyot's analysis but upon closer inspection I actually agree with most points, just not the opening few about the game being disappointing or about MOBAs being casual. I think it's hard to understand what makes DotA interesting without investing an absurd amount of time into the mechanics; I remember watching it a few times and thinking nothing of it, and then suddenly becoming interested.
After playing PvP I thought that the best RTS comparison for Atlas was Warcraft 3. The micro is incredibly important (not necessarily difficult or complex, yet, but still mandatory), and the economy is pretty much background. I guess expansions are needed unlike warcraft, but gold bases here do really seem to be a set-and-forget type of thing. There was not much interaction around them, and it was hard to tell in game if I needed more bases. I agree with the ballcraft/deathball situation and we had a brief conversation about the potential for army splitting on chat... In my games and in what I got from chat, it seemed that having your whole army grouped and your hero either with them or roaming was pretty typical, and it never seemed to be an advantage to further subdivide them.
I also agree about the simplicity of the tech tree and army decision making being an issue long term. My impression of Atlas's goals was to hybridize RTS and MOBA with the interesting aspects from each, namely, the mechanical execution-focused early game of RTSes, combined with the tactically focused MOBA lategame, as everyone ends up in mostly the same place with the same types of pieces. Converging on what appeared to be a very robust strategy for the first 5 minutes and being able to execute it reliably within the span of 4 PvP games on Sunday was a lot faster than e.g. sc2, where I would still have significant execution variability after 2500 games. That might be affect by having all that sc2 experience before though, but it still felt easier to be able to do what I wanted than in other games of this type. I know you've stated that you're against letting players make the wrong decision, and I support a ban on permanent wrong decisions (i.e. I got stargate and he went DTs, but oracles don't exist yet so I just lose), but I think despite this there should be opportunity for making a wrong choice that requires a corrective choice to be made, such as going to another part of the map, farming neutrals instead of engaging the enemy, or maybe turtling and re-teching.
I think that the key aspects of delivering an execution focused early game are the need for: (1) learning a series of steps, like a build order; (2) a mechanical component to the economy, which requires some amount of attention; (3) information about your opponent's choices that affects your own. Some build orders have already come up in a thread on the forums, but overall in game I felt that only a few objectives mattered: taking the first gold base, taking gems during the first two waves, and clearing a camp of neutrals after that. The problem for making this "easy to learn, hard to master" or at least introducing a significant skill gradient to it, is that each of the main objectives appeared sequentially in Atlas. I could make my army during the spawn time, then I could take a base before gems spawned, and then I could jockey for gems until they were gone, and then I could attack neutrals. Trying to do it differently wasn't really an option, as not bringing the army to the gem spawn meant you would lose all of them, and the army without a hero (I played Celesta 4x) was not capable of taking the neutrals (something some races could do in warcraft, for instance).
In SC2, you always have to interleave these steps: track your build order while making probes and pylons, while scouting and deciding what alterations to make from that information. In DotA the early game is very role dependent and can be a bit more subtle, but the most significant mechanical component is last hitting creeps for money, which makes it surprisingly hard to watch the map due to the timing precision required. The scouting and deciding aspect typically lasts longer, but there are often conflicting objectives for most heroes. This comes together to produce a game where the skill spectrum is huge, and people can simultaneously experience: awe at players better than themselves, disgust at players worse than themselves, and self-confidence that they're doing a pretty good job of things in even matches.
All that said I actually really liked several of the mechanics that are somewhat unique to Atlas, and I wanted to talk about why. We talked about this briefly in chat, but I really like how tactically focused the late game is. Once my army was fully populated all I had to do was think about where to be and how to prod my opponents out of position. It can get a bit stalemate-ish if someone is able to establish defenses, like TvT in 2011, so perhaps more mechanics to force action to begin are needed. The next aspect I really like is that you don't make a misstep early in the game and then spend 5 minutes getting pounded before finally losing. At some point, a team wipe can probably lead to a gg push, but there was generally a long midgame section where mistakes did not cause an immediate punishment, and due to the respawning mechanic they were not permanent. Several times we would wipe or get wiped and be able to re-establish, giving us another shot. SC2 was incredibly unforgiving in this way, with large battles only rarely not letting one side gain an advantage, which is often demoralizing considering the amount of time and effort put into getting to the 10 minute mark with a flawless, technical execution.
Finally, I agree about some player frustration being good for motivation, but I think sc2 illustrates how it can easily go too far, so it's a difficult line to balance. DotA would suffer from the same thing, but patches are frequent and radical, removing frustration over losing to the 1-1-1 for 20 games in a row while constantly introducing new things to learn.
TLDR key thoughts: more time-overlapping objectives, more tech/army-specialization choices with opportunity for mistakes and adjustments, keep the free respawns, avoid difficulty just for difficulty's sake