i've been thinking this for a while now, but...
it's really, really captivating how fighting games tend to have an ebb and flow that constantly build up and break down their own strategies!!
that is to say, in a fighting game, even a simpler one, a player has a lot of choices at any one moment. at any time, if you are free to act, you can walk forward, walk back, crouch block, dash forward, backdash, jump forward, neutral jump, jump back, one of two punches, two kicks, two heavies, throw, or use a special move... even without really getting into unique system mechanics or movelists you like have 20-30 actions you can take, and besides all that, you can always delay the timing of anything. it's enough to be overwhelming.
the thing that makes it bearable is that most of the time, there are several explicitly "wrong" things to press. if you're at disadvantage in 5K range, it's not the time to use your 30F windup special. you probably want to block, or jump, and if you just have to press something you press your fastest button. you do something safe and low-risk, because you're at disadvantage.
but the thing that rules about fighting games is that, if your attack hits your opponent, then it was successful. if you use specials to send yourself recklessly flying across the screen and it lands, then it landed! you got the combo! it's certainly a dangerous and terrible move, but it also worked. yes, if you do this predictably you'll get annihilated, but in the right situation, a typically terrible option can become great. even a niche or punishable move on your list has a purpose if your opponent has to consider it, and if they're not considering it, then it's suddenly a lot more powerful.
in this way, the lowest risk, highest reward strategies that form the majority of a fighting game match are broken down. of course you should always take a safejump. you'll get a combo if they mash and pressure if they block. and of course, if they know that, then they'll always just block the safejump, because the outcome is better. and since you know they're likely to block, you decide to try the higher risk action of feinting a safejump into a throw. and because the threat of throws is on the table, they might mash.
and so it goes!! throughout a series you and your opponent constantly shift from safe "expected" actions to risky unsafe nonsense and back again, a cyclical arms race of players being clever by being foolish. because there are so many directions in which you can shift, players stand out by their favorite evolutions. and it's constant. watching these evolutions all happen in real time, mid-round, in a matter of seconds or half-seconds, is so genuinely exciting... it's such a brilliant result of the genre.
any niche action can be valid if your opponent isn't respecting it. and by forcing them to respect it, you make all of your other actions stronger. it's totally fascinating, i love it!!
Strictly casual here, but breaking people's brains by using Ganondorf's >1 second of startup time up tilt never gets old.1 I guess it works in part 'cause we're normally really aggro with Ganondorf? So people have been waiting for the pressure to drop, and are missing why they suddenly have a chance to close in at their own pace—ironically often hesitating just enough to ensure they're caught by the hit.
-
It's not a hugely reliable choice, but the risk/reward is good enough to throw it out from time to time.