Back on topic:
B23~4 is safe against Aquaman's From The Deep (ground trident) as you have just enough time to block it. It seems he can block a follow-up Event Horizon (frame data is a lie). However, if he ends up in Soul Crush range after b23 and tries to trident you, you can block it and follow up with a crush.
Not sure whether Event Horizon data or From The Deep's frame data is incorrect. EH has a start-up of 22 frames, FTD is -24 on block. I'm going to assume this is the game extends EH start-up depending on distance, although it doesn't work at point blank range either. Tested it in the corner too to no avail.
Negative Mass (23f start-up) gets blocked as well :/
From what I can gather based on watching back a few 60fps recordings:
- Listed startup for projectiles does not include its first active frame. Add 1 frame.
- Listed frame advantage for projectiles is miscalculated due to the above. Add 1 frame to that too.
- When returning from blockstun, the game forces a neutral state for one frame and this is not taken into account by the framedata. This effectively adds 1 frame to every block advantage with regard to reversal attacks. I have not yet tested to see if this also applies on hit.
Applying this theory to Raven vs Aquaman (not specifically tested, but seems to add up):
- FTD is listed as -24 on block.
- Projectile. Add 1 frame to listed advantage. -23.
- Add 1 frame to advantage due to forced neutral frame. -22.
- Event Horizon startup is listed as 22 frames. Actually hits on frame 23. Blocked.
- Soul Crush listed as -20 on block.
- Projectile. Add 1 frame to listed advantage. -19.
- Add 1 frame to advantage due to forced neutral frame. -18.
- FTD startup is listed as 18 frames. Actually hits on frame 19. Blocked.
Edit @ 2013/06/03: There is also a possibility that the attacker may also have a forced neutral frame after recovering from their move. If this is the case then the forced neutral frame does not affect the frame advantage, but as it would be possible to block on this frame a reversal needs to be 1 frame faster in order to punish. The result follows the same math as above, but it would be more accurate to think of it as adding 1 frame to the startup of the reversal rather than to the recovery of the blockstun.
I'll try to test a bit more thoroughly later. My equipment isn't fully set up at the moment due to taking my console to an event yesterday.