Having just had a workout and taken a shower, I was eating a snack and rambling to myself (as I am wont to do) when I stumbled across an idea that sounded somewhat profound. Rather than letting it drift off into the ether, I decided to grab hold of it and expand on it here, for your entertainment and/or benefit. Strap in, my friends, for this might get weird.
The specific line of thought that crossed my mind was this: The presence of choices does not indicate the presence of choice, nor does specialization necessarily incentivize variety.
Taken in the abstract, this can apply to many things, but I was thinking of it within the specific scope of game design (even more specifically, within the scope of Dungeons & Dragons, but it's relevant enough to expand it to cover all types of games).
One of my fundamental thoughts when it comes to game design (which I don't do in any professional capacity, just to be clear) is that if a player is presented with a choice, their decision should be meaningful. It can be meaningful in an almost entirely superficial way (such as putting in a custom name in most RPGs, which rarely serves any greater purpose than adding a little personal flair to menus and dialogue in a Mad Libs fashion), or it can have a profound impact on the player's narrative and/or mechanical experience. That doesn't matter so much as the fact that the choice has SOME identifiable impact.
If the choice is something that has no impact beyond the moment in which it's made, it shouldn't be offered in the first place. At that point, it isn't an actual choice. It's just busywork. Of course, this is pretty obvious to anyone capable of even a modicum of critical thought (and I'm sure that my audience here surpasses that criteria with ease, naturally), so I'll take it a step further and say that this also applies to blatant cases of But Thou Must. In Demon's Souls, when the Monumental asks if you'll take on the quest of killing a bunch of demons in order to bring forth the Old One, it's really a pointless moment because you can't advance in the game without agreeing to it. Declining just gets you laughed at until you agree. At that point, why even ask? If it was like Shin Megami Tensei II, where declining at a But Thou Must moment would get you points towards the opposing alignment, there'd be some sense to it, but this isn't the case in Demon's Souls. It's merely a moment of bad design, and I say this as someone who enjoyed Demon's Souls immensely.
Taking this train of thought to its logical conclusion, this also means that a set of choices which has a clear and objective "best" option is really no choice at all. While part of me is tempted to turn that into a rant on why grave robbers were and still are the worst characters in Darkest Dungeon because there's never a case where there isn't a better character for the same role, I want to touch on what started me down this whole train of thought in the first place. I was watching Matt Colville's series of videos where he makes a human fighter named Duncan under various rulesets to explore how the rules changed over time. In the video using the Unearthed Arcana rules from 1985, he mentions how fighters should always opt for weapon specialization (and furthermore for double specialization) because the benefits of hitting more often, doing more damage when you hit, and getting more frequent attacks with a given weapon type outweigh the advantages of just not having a penalty to hit with one more weapon type. What exacerbates the problem here is the combination of the following facts:
Fighters have the best chance to hit in the first place
Fighters are penalized the least for using a weapon that they're not proficient with
It doesn't take long for magical weapons to provide bonuses to hit which equal if not exceed the penalty for lacking proficiency
Almost nobody used the rules for weapon speed factor or weapon vs. armor type (the latter of which was presented in a horrible way up until 2nd edition), reducing the mechanical differences between many different weapon types (like the dozen-odd different types of polearms)
You can only attack with one type of weapon in a given round of combat
So, we have a character whose role is to hit stuff with weapons, and we're saying that they have the choice of either being better at hitting stuff more often/harder or being able to use more types of weapons to hit stuff at a lesser competency level (even though they can only use one type at a time), within a context where the penalties for using other stuff that they aren't proficient in are easy to overcome. At that point, you might as well just say that fighters have only 2 weapon proficiencies instead of 4 but can choose one of them as their specialty at the start. This problem persists in both 2nd and 3rd editions as well (both of which I played many years ago, and I don't think I ever saw a fighter that didn't specialize, as the lopsided cost:benefit ratio was apparent even to brand new players). I can't speak for 4th edition as I never played it, but it's a definite step forward that 5th edition changed specialization to being a fighting style thing that fighters/paladins/rangers get as an automatic class benefit after a certain level. The offshoot of the old system in feats is still problematic (i.e. is there any reason why a ranged character WOULDN'T get the sharpshooter feat?), but at least it's an improvement over the past.
The concept of fake choices ties in nicely with the second half of my original postulation. If a player is to be granted choices, and there are meaningful differences between those choices, then it stands to reason that where choices are offered related to the development of a character's abilities, there will be a trade-off (over time, if not initially) between specialization versus generalization. The problem that arises, then, is that if the advantages of being specialized are great enough to compensate for lacking diversity (i.e. if the game is designed for a player to be able to experience all of its content regardless of their character's build), we reach a state where being a generalist is meaningless. In other words, if character development choices are offered in a game, it would be a bad design choice to make an intentional effort to allow any character build to be able to experience all of the game's content.
However, while this should be simple enough to hold to in video games, the line is blurred in tabletop games, where a player's ingenuity can compensate for the shortcomings of their character. To that end, as I'm on the cusp of running a new 5th edition Dungeons and Dragons campaign, as well as going though introducing another friend to the game through a little series of simple one-shots, I've made it a point to try following Matt Colville's idea of not handicapping the challenges that I'm presenting in order to accommodate a so-called unbalanced party. The world will present a set of challenges which is sensible within the rules and logic of how it works, and if, say, a party with no wizard happens to come across a spellbook, I'll leave it up to them to figure out how to make use of it. I will have NPCs in the world that they can ally with to cover what they lack, of course, but those will not be permanent party members ala DMPCs unless the party is able to give them reasons to stick around. I believe that'll increase verisimilitude as well as proving sensible challenges which will feel all the more rewarding for overcoming despite lacking access to the "obvious" solutions.
Minor spoiler for Critical Role: in as much as Matt Mercer caters to his players, you can ever see a bit of this there when they found a Holy Avenger sword in a treasure pile despite lacking a paladin PC.
To sum this all up, offering choices in a game is easy. Actually offering a real choice to the player is a step harder, and thinking through on the implications of offering choices in a way that presents a richer overall experience (potentially at the cost of leaving content unseen) is harder still, but I think it's something that good designers should do. Otherwise, why offer the choice in the first place?