When it comes to not just game development, but anything that involves competitive markets, there are two major traps a budding entrepreneur can fall into.
Number one is entering a dying market with a declining number of buyers. As my friend Geri says, it is impossible to overestimate the importance of doing some market-research before committing to a project, or even conceiving of the very idea of said project - failing to do so is the fastest route to unemployment and bankruptcy.
However, taking this advice too rigorously can result in you going to the other extreme, and falling into the second trap: attempting to enter an already oversaturated market.
In other words, the sweet spot would be to enter an undersaturated, but still very much thriving market. But how does one do that? Well, obviously, you need to be creative, but lacking that, you can always look at the most popular genres and games that make millions, and try to find the ones where a single company has monopolized the entire genre. I am going to give you three examples, where, for all intents and purposes, a single company has a monopoly on an entire genre, and is long overdue for some new blood to enter the competition.
Number 1: Bethesda Softworks - The sandboxing action-RPG
The Elder Scrolls and the Fallout series aren't the only action-RPGs out there. There are a bazillions of them, many of them far higher-quality than what Bethesda produces, when judged purely as action RPGs.
The Elder Scrolls and Fallout aren't the only sandboxing games either: after all, Minecraft exists too, and offers far more freedom as a sandboxing game.
The Elder Scrolls and Fallout aren't even the only games that combine high-octane action, RPG elements and sandboxing: Mount and Blade also exists, offering far more freedom and far more RPG elements at the same time.
So what makes Bethesda's games so unique then?
I'd say it's the Bethesda-formula. Bethesda combines these elements in a unique fashion that no one has attempted to replicate before.
Bethesda's games put the player's fully-customizable character into a wide-open sandbox full of fully voice-acted NPCs. The games give the player a lot of freedom - freedom to do sidequests, to enter any building (not just plot-important ones), to talk to any NPC (not just plot-important ones), to buy property, etc. - but not too much, where the player's actions would be completely derailing the lore (e.g. becoming a monarch of a country). The games don't railroad you like 90s JRPGs, but neither do they put you into the world with no guiding hand or leads to go off. They give you a main quest, but you are free to ignore.
Obviously, there is one more hard-to-replicate ingredient to the Bethesda-formula: moddability. The ability to mod the game is precisely what has kept the community alive literal decades after the release of each game. Hell, The Elder Scrolls III Morrowind and The Elder Scrolls IV Oblivion were released respectively in 2002 and 2006, but people are still making mods for them! Hell, some people reverse-engineered the Elder Scrolls II Daggerfall (1996), and are making mods for it.
Sadly, the Elder Scrolls and Fallout games are AAA-games, which makes than an indie-developer won't be able to just craft a TES-clone and enter the market. However, a wise studio with the budget for the voice actors, artists and writers could easily take the Bethesda formula and apply it to their own fantasy, to build a game that offers players the same mix of character customizability, modding-support, wide-open sandboxing, RPG elements and high-octane action.
Bethesda is clearly overdue for some competition.
Number 2: Creative Assembly - Grand strategy
The Total War games are unique. They offer a formula, that no one seems to have managed to replicate so far, at least, not in the way that you'd recognize.
Effectively, the Total War games consist of turn-based strategy and real-time tactics. Or rather, turn-based "ordering your armies and agents to move around your provinces", and real-time battles once your armies actually engage with the enemy. The Total War games additionally augment this with RPG elements, allowing your agents and armies' generals to have different traits (virtues and vices), skills, level up, etc.
In the first two Total War games - Shogun Total War (2000) and Medieval Total War (2002) - your armies and agents were all put onto a Risk-style map, where they just snapped from one province to another, while fighting their actual battles on a 3D battlefield. In Rome Total War (2004) and all following Total War games, the Risk-style map was replaced by a two-dimensional map where armies and agents can freely move, adding additional strategic elements into the game, e.g. hiding your armies in forests to set an ambush, building forts to block chokepoints with your armies, flanking the enemy army, etc.
Surely, the Total War games can't be the only ones that combine turn-based strategy with real-time battles, right? And you'd be right - I actually know a few others, tho most of them are almost two decades old.
However, none of them feature battles as realistic, and none of them feature the same amount of freedom to shape the world to your will (albeit, some alternatives - like Gates of Troy - compensate by having a more sophisticated economical simulation).
Creative Assembly is clearly overdue for some competition.
Number 3: Paradox Development Studio - Grand strategy
Paradox Interactive may have published several diverse kinds of games, but when people say "Paradox games", what they actually mean is a subgenre of pseudo-realtime Grand Strategy games developed by Paradox Development Studio: Europa Universalis, Victoria, Crusader Kings, Hearts of Iron, Stellaris, and last but not least - Imperator: Rome.
What do all these games have in common? Well, in all of them - except Crusader Kings, which I'll get into later (and maybe Imperator: Rome, which I never played) - involve you taking control of a nation, and controlling its armies and agents to conquer land, managing the economy of said nation, and so forth. These games also involve randomly generated events, as well as decisions the player can take to influence the course of history. Battles are just as turn-based as the game itself, and you have no direct control - it's basically all just dice rolls, where you get periodic updates on what's going on. The games are pseudo-realtime, or rather, turn-based, but you don't get to pick when do turns end: it happens at a regular interval, unless you pause the game. In all Paradox games save for Hearts of Iron, a turn represents a day, while in Hearts of Iron, a turn represents an hour.
What makes Crusader Kings different? Well, given how it's a game set in the Middle Ages, it has to simulate the medieval feudal system: nations didn't exist yet, instead, feudal lords swore fealty to one another. A king may have had 10-15 dukes as his vassals, but each duke in turn also had 2-3 counts as his vassals, and each count probably had 1-2 barons as vassals. If you were a king, you had vassals, who had vassals, who had vassals.
Which means that in Crusader Kings, most provinces - ruled by counts - are effectively their own independent actors, albeit if they are someone's vassals, they do provide levies and tax revenue.
Crusader Kings is thus, more character-driven than the other Paradox games, where even a lot of unlanded characters play vital roles and their traits matter. But otherwise, the Crusader Kings games follow the Paradox formula, and play like the rest.
Where was I going again?
Right, so, unlike the Elder Scrolls series and Total War series, the aforementioned Paradox games are not AAA games. There are no voice actors. There's a rather limited amount of low-poly 3D modeling involved, lots of 2D art and lots of scripting, but otherwise, we're talking about games that could be reasonably replicated by a handful of indie game developers, if in a form somewhat reduced in the audiovisual department.
Paradox is clearly overdue for some competition.
To clone, or not to clone?
Imitation is the highest form of flattery. Where does a game genre end, and [insert game name]-clones begin? When the genre of first-person shooters first became a thing, FPSs were called "Doom-clones" (1993), and the term "First Person Shooter" didn't become widespread until Quake came out in 1996.
At what point do start defining the Elder Scrolls games - such as Skyrim - and any potential Skyrim-clones as a new genre? Likewise, the same question goes for the Total War games and Paradox's pseudo-realtime Grand Strategy games. They clearly have many common elements.
Competition breeds innovation, and anyone with the budget should see the potential behind usurping the monopoly of the aforementioned three companies on their respective genres.
Obviously, I'm not encouraging you to wholesale copy-paste and entire game with all of its mechanics. No, quite the contrary, I am encouraging developers to innovate. What I am encouraging however, is adding some new blood into the mix.
