Everyone here is wrong. It's embarrassing on a forum specifically for game design.

Firstly the word game is too broad. Without going into stuff like Burgun's work on improving game design theory "games" are interactive systems with an incredible array of goals. Some "games" would benefit from simplicity and some require complexity. Simplicity is probably financially vastly preferable.

Simplicity vs complexity hinges on your design goals. The whole depth thing is garbage. "Depth" just means being in the right spot on the spectrum given your design goals. The design goals for something like Chess or Go demand simplicity. Other kinds of games have design goals that demand complexity. Some things are impossible without complexity.

EVE Online requires both a complex system of rules and knowledge and quite a bit of dull activity. Its simply not possible to achieve its goal, simulating space imperialism, without those games. You can't do it. Period. WoW has no need for complexity and in fact it would probably do better to be even less complex than it currently is.

The objective truth is that some kinds of games will always be more popular regardless of their design. An completely perfect EVE style MMO is NEVER going to have more players than even an above average WoW style MMO. Never, ever, ever. There is simply not enough of a desire for what it provides regardless of the quality on offer. Similarly LoL will always have larger audiences than Starcraft. The market for MOBAs is larger than the RTS market. The quality differential would have to be HUGE for an RTS to out compete a MOBA in profit or popularity.

The market is nearly always more significant for those purposes than the quality of the game above a minimum quality threshold. Depending on marketing power that threshold can be pretty low. Furthermore a shitty MOBA will almost always bleed players to a better MOBA and not to a better RTS.

Games that objectively require more complexity to function are inherently going to be less popular and less profitable than more simplistic games. The market for simple games with a low time commitment is vastly larger than the market for games that require a large time commitment and more intellectual resources.

You told everyone they are wrong, then went on to give some very unhelpful information "genre X is just going to be more popular than genre Y." How does that help? Can you dig a little deeper? There is a reason why MOBAs are popular: They are simple to learn, but difficult to master (as someone astutely already pointed out). Also you continued to reiterate the same "incorrect" points that others had already made as correct when coming from you. Not to mention, the typical MOBA/MMO player would be considered more hard core than someone in the "general audience," which is what the OP is asking about in this post.

An awesome read on this topic is the Mid-Core Success Series by Michail Katkoff, the guy who made Clash of Clans. I also recommend subscribing to the http://www.appmasters.co/ podcast. Steve P. Young is primarily a PR/marketing guy, but he interviews all the most successful mobile game developers from some of the most successful games (Crossy Road, Angry Birds, Color Switch, etc.). His podcast is extremely informative and motivating.

I also recommend the Game Designer's Round Table podcast. Just like Steve P. Young these guys interview a lot of the top game designers, both in the digital and tabletop spaces. The discussions on tabletop games do correlate well with digital games and might get you to think about them from a different perspective than if you just think about the AAA games from big studios like Bethesda, Blizzard, etc.

Well, if you are not planning the next 500 mio $ AAA smash hit, you don't need to appeal to everyone and their dog.

Instead, you should try to find a niche, and make the best game you can to cater to that niche. That niche might be very well the ultra hardcore gamers that like overly complex, or even very hard games (see dark souls).

First and foremost you need to find an audience. Get people excitet for your game. Which is going to be hard if you try to copy others. Be original, come up with something innovative, don't overcomplicate your design, but never be afraid to try something more complex.

Just look for feedback from friends and potential players online early. You will soon get an idea if your game might be too complex or simple if you keep prototyping ideas and let other play them.

Simplicity vs complexity hinges on your design goals. The whole depth thing is garbage. "Depth" just means being in the right spot on the spectrum given your design goals. The design goals for something like Chess or Go demand simplicity. Other kinds of games have design goals that demand complexity. Some things are impossible without complexity.

In this usage, complexity refers to the factors that must be explicitly learned to play reasonably, and depth refers to the amount of "stuff" below the surface that can be discovered for greater mastery (or fun). Usually complexity means inherent complexity (a long rule book) and depth means emergent complexity (many interesting patterns derivable from the rules). Sometimes it means the surface rules needed to understand a game at all (a strike in baseball) versus information which is useful but not critical (infield fly rule).

Go players talk about "eyes", "knight jumps", "moyo's", "influence", "heavy play", "fighting spirit", etc. None of those appear in the rules. Contrast that to an argument about which MOBA character has better control skills, which might focus on cooldown times and range. Those are explicitly defined variables with specific values set by the developers. There's a different experience between deriving further and further understanding of the consequence of rules in Chess, and pouring over the relative cover bonuses of 30 tanks in a crunchy wargame. People have different preferences between the two, but they are different things.

If we want to really dig into designing appropriate complexity, the more relevant discussion there is probably things like learning curve, epiphany moments, levels of mastery, tree width, rule scope, etc. But for the question of "simple or complex", I think "depth over complexity" and "as simple as possible but no simpler" is sufficient.

I think that a good game is a mixture of both complex and simple. The user is presented with simple choices, input, requests... but the game does complex operations and interrelations based on the users choices.

So, it should be simple for the player to manipulate the game and determine what the game is requesting the player to do... but the systems and mechanics should be deeper and more complex.

-or-

Simple to play, complex to understand.

You shouldn't need to cross-reference three different charts to make an informed, useful decisions... but that chart cross-referencing should reward the player who does try to optimize at that level or detail.