Okay, let's go through this in chunks
So it turns out I actually didn't go through my line of reasoning here on this thread, just in the reddit post I mentioned. So here's are two excerpts of my most fundamental problems with this system of SDI:
This is based around lack of consistency and feel of the game, as well as potentially causing certain moves to become imbalanced and remove options. And again, for nothing gained, no matter how small the SDI; the fact it exists at all still would potentially cause this problem.
This highlights what I already said in the last post to you. Potentially punishing a player for hitting their opponent, skewing the risk/reward system integral to all fighting games. I got hit with move A, therefore I take the full brunt of move A's effects as they are taken advantage of by the opponent.
This wouldn't be a problem, but how can you ensure that to happen? Multihit hitboxes traditionally have very low hitstun duration in order to allow a player to get out of them should the offending player find a way to cancel them, such as landing into the ground. Are you saying th hitstun of the move should linger on as long as the move is, even after the character visibly escapes? Because that still doesn't fix the readability and consistency problem.
An example with Bayo Up-B. If we look through the
Smash move viewer, we see that there is an initial hitbox and the multihit hitbox upwards that lasts for 14 frames and starts on frame 8. Let's ignore the initial hitbox for now and focus on the multihit. Now let's say the player hit by the move immediately escapes during frame 11, the second rehit of the multihit upwards.
If I'm reading your statement of "SDI out of a moves initial hit and still remain in stun", then the only way this sort of situation could work with your ideal while also making sure the Bayo is not punished for doing something right by hitting the player is if the escaping player is stuck in hitstun for at LEAST up to Bayo's first actionable frame for that move; in this case, 31 frames, since her FAF on Up-B is 42. This way, both players become active on the same frame, and at the
very least Bayo isn't punished for hitting the player, but she isn't rewarded either. This could definitely work with a bit of tweaking her Up-B.
The problem with this ideal you have to make SDI not punish a player for doing something right by making the escaper still be in hitstun is the fact that the escaper is now in such a long amount of hitstun for seemingly no reason. Again, not good for readability, and it makes the game feel clunky and unintuitive. On top of this, what if they escape much later into the move's multihit? Would they then take less hitstun than this example just to ensure Bayo is active at the same time they are? What if it's a multihit move that can be interrupted by landing? It seems like you would have to create an entire system of scaling hitstun based on when a move hit you AND what the player using the move ends up doing. At this point, it's FAR too complicated and convoluted than just removing SDI for that multihit move altogether, and just having the Bayo Up-B's last hit have DI exclusively. A far simpler, more elegant solution.
I knew what you were talking about and it still makes no sense to this situation because those moves aren't multihits. If you get hit by a singular Mario UTilt, you just DI normally. If you get hit by another one, you DI again. SDI has nothing to do with this scenario because SDI involves changing trajectory during the hitlag of a move, not trajectory after the move has hit you. If Mario's UTilt had a DI factor but absolutely no SDI factor, nothing would change, because the change in movement from the player being comboed here is in play only after Mario gets the hit off.
Except like I said, SDI on multihits rarely have enough hitstun to compensate for the rest of the move in its entirety, and I already went through why trying to make sure Bayo has no guaranteed followup even with them escaping the move is unintuitive and convoluted at best.
Again, going by what the difference between SDI and regular DI is, adding SDI wouldn't do anything for that scenario. The actual solution is heavily increasing DI angles for those specific moves to make sure they can't combo too easily into each other. And I already mentioned my mild disdain for stale move negation as is.
Honestly, your example has more to do with Smash 4's shortcomings than SDI's merit as a mechanic. And even with all of that, more DI is a much more elegant and sensical answer to this problem.
Except it has the capacity to move you to outside of a rehitting hitbox like I mentioned with Bayonetta's Up-B. Even though it retains stun, because these hitboxes rehit, if they move out of the hitbox after the hit, they obviously can't rehit anymore, no matter how stunned. And like I said, the hitstun for each and every rehit is not high at all.
Again, increase the DI without having to increase this mechanic that does exactly what DI does, but with the caveats I mentioned before. Just increase the DI. It doesn't have these problems I'm mentioning AND it still puts human error into the equation just fine. This is why I endorse DI but not SDI, because one is clearly superior to the other to the point of making the other objectively inferior and unnecessary, at least in a LOT of cases.
You keep bringing up single hit moves more than multihit moves here. I understand the use of SDI against single hit moves. There it might at least have some semblance of purpose, but what would it give that simply increasing DI deviation not bring exactly? And DI increase would be helping heavy weight characters, again without the baggage SDI can bring for some moves.
Most of this theorycrafting you are bringing up in this paragraph about the choices between SDI'ing one way or the other could easily just be substituted with DI anyway. Ignoring the specifics about movement during hitstun and multihits, let me just replace "SDI" with "DI" in your paragraph here as well as add some comments.
"Just because a player DI’s doesn’t mean they get out of the combo for free, they are still in stun, and if they DI wrong or if the opponent predicts it, it may put the opponent in a more disadvantageous position. Think about it, if you are stunned, take a hit and are giving the opponent the maximum amount of control due to DI’ing away, you are forfeiting more control of the stage at a possibly lower percent, at an attempt to be safe, or punish back. Of course this depends if the move would have killed or not, but mixing up an opponent attempting to repeatedly SDI just means resorting to use a move that isn’t multi hit to kill. (Again, just have the multihit move subject to DI on the launching hit, or if it's a rapid jab, then SDI is excusable at that point) Of course this is all character and positioning dependant, and it does hurt characters who have have set ups into multi hit finishers like diddy with d-tilt>u-smash, but at kill percents you can d-tilt > rar bair. (This dynamic wouldn't be lost without SDI allowing for escape prematurely by USmash) It’s a reoccuring issue that characters with several guaranteed kill set ups usually take up top placing in tournaments. Having some DI would help some heavy weight characters at least."
Quick edit towards
Quillion
. I think a better, and much more appropriate stale move negation system is to increase DI deviation for stale moves. This works especially well since most of what stale move negation works against is combo starters and kill moves. With more DI deviation, stale combo starters would have a harder time comboing, and kill moves that are stale would have a harder time killing as well.
Second quick edit: I see you talking about L-Cancelling's problem with relying on human error. I don't know if I myself bent you towards that view, but I'm glad to see that maybe I'm not as bad at explaining this stuff as I think I am.
Probably still am though ._.