These days, every RPG pitch is exactly the same. A grim, dark world where death waits around every corner, every decision could be your last! Face tough moral choices in this epic fantasy adventure with four hundred hours of gameplay!
OK, all of those pitch elements are bad. Briefly:
We can tell how grimdark your setting is from your art samples, sell us on something unique.
We can tell how epicfantasy your setting is from your art samples, sell us on something unique.
Tell me what in your game supports hundreds of hours of play, because hundreds of hours of empty fetch quests is what I assume.
Death and deadly decisions mean nothing, because we can save and load pretty easily. Also, it's bad gameplay to kill the player off for choosing poorly: don't offer a choice if one of the options is just flat-out wrong.
But by far the most egregious of these claims is the idea of "tough moral choices".
This is an incredible combination of poor gameplay, poor writing, and avoiding responsibility for the content of the game you're making.
First: poor gameplay.
Choosing between a few canned options is extremely dull gameplay. This is especially true in modern games, where unique characters are expensive to make and full voice acting makes every line expensive.
In older games, you would usually be presented with "moral choices", but the game rarely judged what you did. For example, in Fallout 2 there was a string of events where you could sleep with a farmer's daughter, then be pressured into marrying her and taking her with you. The game doesn't really care what you choose to do, it simply gives you a variety of choices, many of which are based on the core game features (gunplay or stat/skill tests) rather than being strictly canned dialog.
It isn't expected that you'll have the farmer's daughter in your party. But she was very cheap to create: her graphics are (I think) identical to other NPCs, and her dialog is very limited. That was acceptable, because it was true of all the party members. If she is in your party, there's a lot of things you can do with/about her, all due to the in-world mechanics of how party members are handled.
But today, it'd be a huge deal - you've got to give her a 10,000 polygon face and 8 hours of voice acting and an interesting back story because she's a party member! You've got to make sure that every player gets her on their team or you're wasting that cash!
Basically, "tough moral choices" are poor gameplay because choices are poor gameplay. "Tough moral gameplay" is a better thought: players should be able to make their moral choices through core gameplay, or should have direct core gameplay intertwined with them. Not small loot rewards or people liking them more - that's not core gameplay.
Second: Poor writing.
When you start every scenario thinking "what's the good, neutral, and evil path?" you end up with an extremely dull, repetitive set of quests where the player always chooses the same path.
This sabotages your writing in a lot of ways. Firstly, it makes you create canned paths through every scenario. This sabotages the tension and uniqueness of the scenarios, and it's deeply wrongheaded. The biggest thing you need to understand is that the player chose to enter a scenario, and that is a big clue as to what they want to accomplish.
For example, there's a mugging. The player steps in. You immediately know you don't need the neutral or evil paths. Nobody neutral would step in, and nobody evil would step in just to make it worse. That's nonsense.
Do you hear me, Bioware: not even a Sith would step into a robbery just because they wanted to make the situation worse!
It's easy to write a situation such that you know what a player is trying to accomplish when they step into it. If the situation does have good and bad paths, rather than offering a popup choice, make the points of entry different: if the player talks to the nice people, he's obviously trying to help the nice people, and if he talks to the baddies, he's an asshole. You can even make the context command for starting the sidequest literally pop up your goal before you hit the button: "Press A: Try to intervene".
Once you have gotten the player to commit to a "want", you have a fulcrum. You know what the player wants, so you can simply offer him different ways to get it - many of which can be integrated into the core gameplay. Are you going to try to talk the robbers into leaving? Bribe them into leaving? Threaten them? Kill them? Hypnotize them? Bluff them? Grab the victim's hand and run? These are all much better role play options than "make the robbery worse because you're a really bored cartoon villain".
Of course, this thinking can also go awry. When you come up with canned paths based on gameplay, they can end up serving the same purpose and causing the same damage as your ethical choices. Every scenario must have a warrior, rogue, and wizard path? Just as bad.
And the same solution applies. If your player waltzes up to a wizardy challenge, they are choosing the wizard path before they even see a dialog box.
These things can even be automatically generated. Can you imagine how much more interesting algorithmic quests would be if you could enter them in several different ways, each of which represented a different moral or gameplay approach?
Third: Poor authorship
Perhaps the biggest problem I have with the "moral choices" thing is that it is a cheap copout that lets you avoid putting anything of value into your writing.
When you try to brainstorm for all the things a player might do in a situation, you are throwing away your authorial intent. You can't say anything if you need to say everything.
If you want the player to be able to do anything, give them gameplay options to do that. I wasn't faced with a "steal?" popup box in Oblivion or Skyrim. Instead, I was just permitted to steal. And, since it was integrated with the game world and other mechanics, I could come up with a lot of different schemes. Wait until night, or sneak behind them, or lure them out of the room then dash back in, or put a basket over their head - lots of options. And the game didn't judge me much, aside from small mechanical considerations like not being able to sell stolen goods (even in another town?!) and sometimes having the target hate me afterwards (even if they never saw me?!)
Although neither of those games has much authorial intent in it, the ethical options are nicely integrated into the play of the game rather than limited to popup boxes.
Games that offer both freedom and authorial intent are hard to come by. Planescape: Torment is a good example, as is Grand Theft Auto Anyofthem. These are games where the developers had something they wanted to say. Maybe it was worth hearing, maybe it wasn't, but they got it across by saturating the game in character. The world is full of interesting things that have some kind of impact, not generic doodads and setpieces. The characters all have personalities and act according to those personalities.
The interactions with the world echo with the authors' intents - in Planescape you can hear thoughts about the hollowness of the world and the struggle to live here anyway. In GTA you can hear thoughts about worthless bro culture vomit.
Anyway, the point is that these games let the authorial intent shine through most of the time, and the games are renowned because of it. The "moral options" within these games are largely couched inside the gameplay and progression, and the game doesn't judge you much when you delve into them.
So what do we do?
If you want moral play in your games, I recommend putting moral play in your games.
A good, easy example of this is the new Fallout Shelter mobile game. The game allows you to control the lives of a lot of people, and you have a lot of options as to what to do with them. A lot of players create outrageous vaults full of bizarre immoral stuff, but the game doesn't judge them much. Most players probably have relatively 'normal' vault cultures, but they still make a lot of moral choices about how things need to unfold. For example, who has to stand guard, who gets to train up, who gets to have kids, and so on.
Now, no game is a blank slate. The mechanics of Fallout Shelter radically tilt the kinds of cultures players are likely to express. For example, anyone can train up to max stats in about the same time, but children do not really inherit much of the stats of their parents. This means that there's no point in breeding for stats. But, contrarily, children do inherit their parent's appearances, so there's a lot of pressure to breed for appearances.
I don't know whether Fallout Shelter's devs thought of these pressures when they designed the game, but the result is very different that an alternative would be. Say, if only children could be trained to increase stats, and children inherited the adult stats of their parents. Then there would be a situation where every generation was better than the one before it, and bloodlines would be extremely important. Would it be better? Worse? It'd be different, for sure.
Diving straight into eugenics is a powerful example, and I kind of went for the jugular. Few things are as morally sensitive as "breeding humans", but it is exactly what Fallout Shelter is about. I think a more careful game could have been made to teach the dangers of that kind of thinking - without compromising the open play. Having mechanics that emerge from eugenic practices would not interfere with the player's freedom, but it would make them think twice about exercising too much control over people's personal lives.
And, of course, Fallout Shelter has a lot of moral constraints built into it. Children and pregnant women are invulnerable to hazards. Nobody ever gets sick, or angry, or dissatisfied with their life. Nobody needs to rest. Nobody ages. And nobody dies without your permission.
The point is this: your "moral choices" are better off as "moral gameplay". By allowing your player to affect NPCs in ways that cause emergent behavior, you can allow your players a lot of freedom without compromising your authorial intent or writing. You can express yourself more clearly, because you can embed your authorial intent in the rules of the game!
You may have to let go of some of your railroad, though. Linear main quests never survive an encounter with an experimental player.
But, uh... linear main quests suck, so that's fine by me!