Strategy Riddle me this (gaps vs parries)

[Sami]

The Riddle

1. The gap in Raven's 223 string (between the 2 and the 3) is blown up by Batman's and Killer Frost's parry when used as a reversal. The gap in Raven's b123 string (also between 2 and 3) however beats both reversal parries.

2. The fact that the reversal triggers indicates there is a gap between the 2 and 3 in both strings. Both parries have a 1 frame start-up.

3. So, in the gap in the string the parry starts up, and then on the next frame (at the earliest) 3 hits as the parry becomes active, but in one of the strings the parry wins, but on the other the combo wins.

4. For reference, you can parry the 3 in b123 if b1 and 2 both miss, so the 3 is not unparryable.

5. Both 3 attacks hit mid.

Possible Answers

1. The frame data for both parry moves is incorrect. Assuming a 2 frame start-up, 223 could have a 2 frame gap (meaning the parry would work), but b123 has a 1 frame gap which blows up the parry reversal.

2. There is a bug with with reversal attacks in training mode (I only tested it in this mode).

3. The first frame of 3 in b123 is unparryable for some reason (hitbox, magic, etc).

4. Something else, likely involving it being a long week at work for me and missing something obvious.



Any ideas? Because it seems that something doesn't add up here. Also, this also calls into question the oft-repeated claims that start-up frames include the activation frame.

 

[Kiko]

Is superman's super 1 frame or 2? (I hear both all the time)

So if its 1 frame couldn't you technically test with super on those parries (f23 > super) get your validation and prove/disprove #1, right?

By the way lolmagic

 

[Sprint]

You're reading the framedata wrong.

1 frame startup does not mean a move is active on its first frame.
1 frame startup means there is 1 frame before it becomes active, therefore it is active on frame 2.

I tested this a while ago by playing back 60fps video one frame at a time. All the moves I tested hit one frame after the listed startup value.


Late edit: The whole "startup includes the activation frame" is nonsense. There's no consensus as to whether or not startup should include it, so you have data floating around in both formats. Personally I'm of the opinion that it should not be called startup if it includes the first active frame.

For the most part, the framedata I've seen that includes the first active frame has been published with the column named something else (Execution, Impact, Hit, etc.), but there's always some dickhead at eventhubs or whatever who copies the data with the column renamed to startup because fuck you for trying to avoid confusion.

 

[Sami]

Sprint I had always assumed this was the case (start-up doesn't include the first "active" frame), but was just throwing out ideas.

This still brings us back to the original question: why does the gap in 223 get caught by parries but b123 does not? Assuming the smallest possible gap (1 frame) between 2 and 3 in both moves (skipping all of the frames prior to the one before the gap):

Frame 1: last active frame of 2's block-stun
Frame 2: GAP. Reversal parry kicks in, has it's 1 frame start-up
Frame 3: First active frame of 3. First active frame of parry.

Why doesn't the parry always beat the first active frame of 3? The only time the parry should fail is if 3 can't be parried or if 3's start-up is sooo long that the parry window ends (definitely not the case).


P.S. Kiko superman's super will activate as the reversal, and he will still take damage from the third hit. Bizarrely, the damage from the third hit is applied after the end of the super move. The damage applied is block damage, and doesn't flash up with a number in training mode. This applies to 223 and b123.

 

[Sprint]

I mentioned that the framedata startup does not include the first active frame. I would like to revise that statement to say that the game engine is a mess and I have no idea what is going on. But I have an unconfirmed theory:

- Normal moves include the first active frame in their startup.
- Projectiles do not include the first active frame in their startup.

I had previously just tested Raven's specials, which are all classed as projectiles.
Raven's df2 cannot punish Cyborg's db1 on block, but her f2 can.
They are advertised as having the same startup, but df2 is definitely slower.


With regard to reversals from blockstun, this appears to be happening:

Frame 1: Last frame of b1,2's blockstun.
Frame 2: Forced neutral frame.
Frame 3: Parry's startup frame. First active frame of 3.

- A reversal backdash will start to animate on frame 2, but if you get hit on said frame then it doesn't come out and you block instead.
- If a reversal special is performed you don't see it happen. The move is stuffed on the same frame that it starts to animate, so the character gets dumped straight into their hitstun animation instead.
- Inputting a normal move will not get you hit in this situation. I'm not sure why, probably a processing order quirk.

- Raven's f2 does seem to hit on frame 13, but it cannot punish anything that is -13 on block.
- Cyborg's db1 is -14 on block and can be punished by Raven's f2. This seems to support the idea that the game forces you into a neutral state for one frame after blockstun (Not entirely surprising; some other games demonstrate similar behaviour, like Street Fighter 2).

More silliness.
- DS f2 > df1~b
When f2 is blocked high there is a gap as described above. When f2 is blocked low there is no gap. Basically f2 gains a frame of advantage when it is blocked low. This behaviour seems to apply to other moves, but not everything.