The problem is that by “trap choices” you can mean different things: (1) bad builds; (2) failed attempts to impose yourself on the setting with your roleplaying preferences; (3) crappy items; (4) items that are crappy relatively to other items; (5) crappy skills; (6) complicated mechanics that induce you to make bad builds, etc.
You would think no one in its right mind would present (1) and (6) as actual problems, but I remember clear Sawyer saying in an interview that was too easy to make a bad build in D&D – see also Mozg’s comment above. For me, these are only trap choices for popamole players.
Bad builds are definitely a problem.
First, building doesn't really test any skill - at least one build, the most important one, is generally done without any clue, unless you happen to already know the system and at least general tone from somewhere else or unless the choices to make are trivial.
There is no point allowing failing builds and there is no point bogging the player down with trivial choices - for example stat allocation only makes sense if there are reasons to allocate stats in many different ways that aren't, for example, summed up by the class; likewise letting player neglect particular stat only makes sense if their class or skillset isn't dependent on it - muscle wizard might make sense but a dumb* one simply doesn't and isn't meaningful in most cases (WTF is a dumb wizard?).
Second, bad builds only serve to reduce the actual depth offered by the system compared to the depth it might theoretically offer.
They are combinatorially wasteful - for example if a system offers a way to express some inherently contradictory or sucky build (like dumb wizard), it's one less useful, sensible and potentially interesting build it could have described instead of a nonsense one.
A good system should strive to only express builds that make sense in its context, fully exploiting its potential depth and only facing players with problems - and only nontrivial ones - when they can actually hope to figure them out based on available information. For example beating all sorts of problems with their build.
*)
Of course, intelligence as an attribute is problematic enough to warrant it's complete removal. Most importantly, it's not really enforceable in general case - you can stop weak character from performing any feats of strength, but you can't force dumb character to behave in a stupid manner, in fact if the difficulty is right player will have to make their character behave intelligently to survive - even a "dumb" one. Moreso, stupid protagonists make for shit heroic stories - a hero that somehow outsmarts overwhelming opposition or uses his cunning to surmount obstacles is more interesting than one that simply facerolled opponents because of stats.
Meanwhile the removal of intelligence as attribute would rid us of the most glaringly broken nonsensical trap build - the dumb wizard - and move dumb firmly into (2) as far as build choices are concerned, by making "dumb" not part of any build, remove the need for out of character behaviour and making dumb protagonists synonymous with dead ones, and dumb characters with dumb players.
(2) also seems the typical childish thing that only a Bethesda fan would require. It reminds of a causal complaining that AoD sucks because he couldn’t role-play as a two swordsman with a silver tongue. Maybe he was happy to role-play as a muscle wizard in PoE? I don’t know.
(2) is dumb if it means, for example "WAAH! I wanted to charge massed archers with my bare chested barbarian, but I die when I try" (this applies to both mechanics and story elements), but mechanics that forces heavily focused builds is similarly wasteful and actually does perpetuate (2).
First, if you have, say, ten skills or stats and need to focus on one to be successful, you effectively have ten different builds. OTOH if you have same ten stats but have to choose five you're good at, you have over 14k builds - and that's without even considering things like different levels of proficiency.
Second, if your character is good at one thing, you make your way through the game doing this one thing repeatedly until successful - which is very much imposing your vision on the world (or until you discover that it's one of the trap builds and you need to restart, which sucks for a different set of reasons - see (1)).
Of course, to really make best of such multitude of builds you'll need at least some compound solutions where multiple abilities need to be used together for effect, but on the flipside it also allows for much more robust and more interesting single skill ones, because you no longer need to ensure that every challenge on a critical path needs to be somehow beatable using, say, knitting skill (because player is guaranteed to have a number of tools at their disposal instead of just a hammer) - as long as you can think of any 6 approaches using different skill each, you're guaranteed to not trap the player; and because instead of mindlessly applying his hammer over and over to every problem he faces, player actually needs to stop and think which approaches may work that he is also capable of (and that's before involving any other mechanics) - "hmm... intimidate the king? Nope. Attack the king? Nah. What else can I do to have my way?".
Consider (3). In the interview I quoted earlier, Sawyer complaints about how players are forced in some cRPGs to wear padded armor at the game's opening, saying that they are mechanically awful and they'll gladly ditch it as soon as anything else becomes available. Now, you can interpret this in two ways: (a) players should start the game with good things; (b) we should have good items of each type and padded armors shouldn’t be excluded. Now, (a) doesn’t make any sense because it’s an example of (4). You start the game with a shitty padded armor because you are a newbie that will get better items as the game progress. Padded armors in this case are awful just in comparison to better items that you will earn later. Saying that every item should be good it is utterly meaningless in this context, because we only have good items in comparison to bad ones. If every item is good, no item is good.
How about (b)? If indeed is the case that some types of items are always bad (and for me it’s not clear that this is the case, see BG2 for instance), then this is just a matter of personal preference. Far from being a proof of what cRPGs has being doing wrong, it is just a personal preference. Sawyer likes padded armors and is upset because he thinks they usually receive a shitty treatment.
Another problem with (3) is that it ignores that items are not just gamey stuff to kill things. They attempt to emulate the setting in some way. If you are a poor apprentice, you are expected to start with a staff with a cloak, if you are a shitty warrior, you get a padded armor, etc. These common places are not arbitrary; they are there for a reason. Thus, Sawyer didn’t present a compelling case in favor of (3).
If you have problems with some items being unconditionally superior/inferior, then you're either not describing them with enough variables or are doing a complete botch job using those variables in actual gameplay.
Shit-tier items automatically have following important advantages:
- They are easily obtained.
- They are easily replaced (see above).
- They are easily maintained (repaired if possible, see above otherwise).
Items don't only exist for the player to swing at the enemy (or to protect the player from being swung at) - you might just as well use player's items against the player by effectively taking them hostage - which makes their value and rarity as big drawbacks as you want, turning them into resource sinks - which is a compelling reason to have a durability mechanics as long as you can ensure it consumes resources that are in some way proportional to the value of maintained items, or using them against the player in some other way (for example brandishing a legendary artifact makes you instantly identifiable - which can be problematic not only if you are doing something nefarious, but if you have any sort of antagonistic forces at work you can't just bash over the head).
How do you take an item hostage? Simple. Put player in an environment or against enemies that are actively harmful to given types of items - for example some metal-dissolving goop or rust monsters and such. Want to see your legendary suit of plate armour that was originally worn by the Emperor Idontgiveafuckabout during the battle of Whatever turn into a rusty puddle? No? Pack something you won't cry after. Put player against some hazard or potential solution to an obstacle that requires parting with some equipment - want to swim through the moat or scale that wall or avoid the guards? Take off that pretty shiny plate please. Want audience with local lord? No weapons allowed. Want to carry all that pretty loot out? While already loaded with all this gear? Pffft.
Of course this kind of approach needs a few things to work:
- You need to be able to take player's unattended stuff - the lord might not be entirely noble (if you excuse the pun), but might be a connoisseur of ancient artifcats of power, prestige, historical significance and, above all else, eye-pleasing shininess, such artifacts left unattended somewhere in the bushes will probably also find a lucky finder. You probably shouldn't be a complete asshole and don't remove irreplaceable items from the world completely if you can help it (read - item has not turned into a rusty puddle), but if they resurface don't make it an obvious "reclaim your stolen gear" quest - let the player work to find them, without any sort of guarantee regarding where and when they might turn up. Randomize as much as possible. Be mildly sadistic, just not a total ass.
- Player needs inventory limitations, and severe ones too. You can't meaningfully expect player to drop their gear if they can just hammerspace it for latter use.
- Player needs some place to store their stuff safely. Otherwise you simply won't get them to part with cool gear because they will either be using or losing it, and they will know.
- Damned Registrations must stay the fuck out of this thread or else he might launch into an inane and delirious rant about how abstract mechanics with no connection to how things would actually work in your world is almost invariably the best choice and realism always makes for bad gameplay.
Now, consider (5). Despite first appearances, this is a very trick subject. First, we need to consider just as items, skills are also tied to the setting. They help to engage the player to the game world. That it is why skills such as hardass, demolition or lockpick in W2, because these are the types of abilities we would expect from rangers living on the wasteland.
(5) is a subset of (1), thankfully easily remediated. It shouldn't be too hard to merge weak, thematically related skills or split strong ones until relatively balanced utility is achieved. It also helps to avoid making your skills one trick ponies.
Avoiding allowing ultra-focused builds also helps make situational skills useful and worthwhile.
