When it comes to gaming, there are a lot of skills that help you succeed. There's stuff like spatial awareness, reaction speed, execution, situational decision-making skills and so on. There are some skills in games that are simply a tax, though. It's a skill that adds nothing to the game, but forces you to do it anyway.
Because I play fighting games, we should point out how rife they are with stupid mechanical stuff that you don't need. Special move execution could be greatly simplified, for instance. A lot of games do simplify special moves a lot, but the dragon punch motion has been around for literal decades and there are still people that can't get into SF because of it. P4A finally got rid of it, and I have somewhat mixed opinions on that.
There is an argument that doing things like a dragon punch or a command grab or super should take time. The idea behind a dragon punch is that it's invulnerable and beats lots of stuff, so you shouldn't be able to just bust it out quickly. It should be a planned thing. That's the theory but the truth is that even hitting one button in response to an opponent's attack is pretty hard and the risk of doing a dragon punch is already pretty high. I think that 3A+B or whatever the motion is in P4A (it's down-towards and two buttons, don't know which ones) is plenty as long as DPs still carry some risk.
But honestly those skills are tame compared to StarCraft (2, in this case). Consider how much of StarCraft could be automated. An example might be zerg queen larva injection. It's something you want to do every time perfectly and you literally never want to delay it. Larva injection is never a strategic decision. If you want more queen energy for transfuse or tumors you'll make another queen. Yet so much of zerg skill is tied up in the ability to hit backspace, select queens, V, click hatch, next hatch, every 45 seconds. Not a good skill.
Macro in StarCraft is in general kind of bad. BW is the worst. You press a F key to switch to a different screen, then click build dragoon over and over and over, every minute or however long it takes to build a dragoon. SC2 made that easier but it still emphasizes a skill that is almost never involved in strategic decision-making. If you want to stop making dragoons or stalkers or marines or whatever, you could just tell your unit production to stop making them.
Of course StarCraft players will balk, and I think the game is a good game. But it's hard, and it's ultimately hard for the wrong reasons. So much of StarCraft is jumping through hurdles that what you actually build in the game is sort of irrelevant until you're very, very good. No fighting game has special moves that are that hard OR that essential. Yes, it's harder to win in SF without knowing special moves, but it is not as impossible as winning in SC without macro.
I got into this discussion because my best friend and I are working on a game and he asked me if weapon switching should allow the player to immediately attack after the switch, even if you had used a weapon with a long cooldown (long in this case being measured in hundreds of milliseconds). It is cool to use a trick like that just like it's cool when you do macro well or pull off a combo in a fighting game. However, as designers we have 2 options. Either we balance around the trick or we ignore it and let the people who can do it have a bunch of extra power. I chose just to not have the trick. Optimal play shouldn't require players to have an extra dexterity tax.
What about hard combos in fighting games? Well they're like hard dragon punches, right? The answer is yes and no. Just like dragon punches, combos are basically a dexterity tax, but unlike dragon punches there's a bit of nuance. Doing a combo is a strategic decision. Which combo you do, more specifically, is the strategic decision.
First is the likelihood that combos will be in your game anyway even if you don't build for them. The chances are that yes, if you are making a game with hitstun, that people will find a way to link stuff even if you try very hard to keep it out. If you take control over the system, you can make combos look cool instead of lame, and to some degree control the damage levels characters get when they hit things.
When you do a combo, you're making a decision about how much damage you want to do, how much special resource you're willing to expend, and in what position is your opponent going to be after the combo is over. Some people will do the big damage combo every time, but experts will choose combos with more corner carry or have an optimized combo into an ultra or using EX moves or supers or whatever your fighting game has. There are even resets where you end your combo early (before the enemy is ready) and then perform a mixup to start your combo over!
Those things are all strategic, so we have to consider them when creating a combo system. We can't simply have "mash the punch button to do a combo," unless there are other ways to fulfill those other situations. On the other hand, we want each option to be as simple as possible. There's no reason for combo into super to be 10 times harder than the basic combo, for instance.
A lot of games have easy base combos with variation, which is great. Others have complicated combo systems but built-in helper tools to make hitting hard links much easier. In general, the execution can't be too simple or it hurts some of the complexity. Resets, for instance, basically depend on the combo system not being "just mash P for combo A, P and then mash K for combo B, mash P and then super for combo C."
Bunny hopping, various kinds of strafing that make you move faster, and all sorts of aiming bugs in FPS games are other examples of bad skills. When discussing a particular feature, just ask yourself two things:
- Would the game be better off without the skill?
- Would the game be better off if the skill was easy, trivial, and/or blended into the game's "normal" way of playing?
Even in the case of StarCraft macro, the answer to question 2 is "making it easier would make the game better at all levels of play."