It would be great if a pro player could post a full analysis, but they'd likely want to keep the info to themselves to have the leg up in tournaments. Still it's never bad to have discussion, so I'm down for some theorycrafting until I get the demo. Let's just hope interest in Pika doesn't wane and activity on these boards die out.
Anyway, these are all based on observations, but I can attempt to start us off:
Pika's greatest strength right now seems to be how quickly he can attack, thanks to a combination of general ground/air mobility, good auto-cancel aerials, and especially the extremely little startup lag on his normals. Pikachu is immediately able to counterattack out of many situations, so he is very good at both whiff punishing and stuffing opponent's attacks. For example, I saw shofu's Pika beat out a lot of Zero's ZSS's moves before he could attack. Pivot F-tilts are really good for this reason, and I want to say Pika should be able to short-hop air dodge aerial (is there a name for this yet? SHADA?) with Fair and Nair, maybe Uair, but I'd like someone to confirm that.
Other strengths have transferred from previous games. Pika's speed and small size allows him to evade opponents' attacks well, so he can use hit and run tactics effectively. This would of course give him favor on large stages and stages like Battlefield with decently spaced platform layouts. His recovery is still godlike (Skull bash I think has even more distance on it, so it's probably even better than Brawl), so with proper DI he can usually always return to the stage. Standard QA should still be tricky enough despite the changes to ledge mechanics, thanks to the direction mix-ups and (speculating but) that new knockback on the second zip potentially threatening stage spikes. Skull bash's hitboxes also seem to cover his head now, so it has better protection. And of course, QA means he can go very deep for kills offstage, which complements edgeguarding tactics from his specials. Finally, most of Pika's moves are also quick in terms of endlag, so Pika's lower power (damage on moves) is compensated by DPS tactics.
I see 3 weaknesses for Pikachu, though the last one is probably the most unknown right now. First, Pika is a lightweight (I think 79 when the average roster appears to be in the 90s), so you can't take too much damage or else even the large blastzones won't save you. I'm also assuming that weight influencing hitstun is still a thing, so be wary if the opponent can combo very well. Second, Pika still has low range and disjoint, which forces him to go in for damage and limits his ability to space and zone. His shield size and grab range is also small, so the opponent will know that Pika would want to go in in order to get more options, meaning zoning particularly effective against Pikachu. Good fundamentals with DI, spacing/movement, and aggro mindgames can still help solve this as it has in the past given good APM.
The third weakness we should focus on first is the lack of kill power. Pikachu's U-smash, F-smash, and Thunder have all been nerfed, generally in range and/or distance, and combined with the larger blastzones this makes killing opponents raw with Pika not quite viable until around 120-150%. Thunder is particularly painful because when U-smash was nerfed in Brawl at least you could still Thunderspike. But thanks to the meteor in the cloud we have to settle for comboing into the thunder "splash", which seems to have reasonable knockback but the opponent could DI (maybe even meteor cancel?) before move finishes. The range nerf also means you can Thunder and miss because the opponent is too high (and as a long time Pika main that makes me sad), so you can't U-smash or U-throw Thunder at too high a percent. U-smash is actually slightly laggier on start-up, so I'm not sure how safe it is outside of shield anymore (probably okay, but still).
Because Pika can't kill at low percents anymore, this means the player will have to put in much more work to rack up damage, and it doesn't help that Pika got damage nerfs across the board either. So if you want to commit to Pika, don't be dissuaded that you'll need more effort and more button pressing to be effective with him compared to other characters (though honestly this applies to all games). That being said, we can work on Pika's meta by figuring out the following things:
- Figure out what percents Thunder stops and starts comboing into itself. Thunder is not Pika's primary move but it can be useful to master it now for potential future applications (B-reverse meteor edgeguards anyone?). It would also help get a feel of the roster's fall speed and weight since both are involved in the combo, so knowing which characters are particularly vulnerable will be interesting. Knowing if the opponent is always sent in front of Pika can help too. U-throw/U-smash/U-tilt to Thunder are probably combos for this.
- EDGEGUARDING! One's of Pika's greatest strengths, the edgeguarding game will be more dynamic than ever thanks to the new ledge mechanics and how everyone seems to have really good recoveries. We should learn to be particularly creative here since we have a healthy dominance offstage. For example, I'm curious about how good fastfalling Bairs are for dragging the opponents down.
- Solidify the low-percent game - basically make sure Pika racks up damage while the opponent is still suffering low knockback. Master Uair chains and tilts out of approaches and defense, to the point where you can start using higher knockback moves like Nair, Dair, Dash Attack, and F/B-throw to try to get the opponent offstage.
- Not new, but overall just make sure Pika's neutral game still works. Approachs will probably be melee-esque with SHFF aerials (Fair, Nair, Uair), especially since QAC is unconfirmed right now. Implementing Pivot F-tilts/F-smashes and "SHADA" might be the new meta, so see how effective that is.
Finally, keep note of which characters and playstyles you have trouble with. Pika is looking to be a particularly effective rushdown character with gimping prowess, so he's probably going to suffer against characters with large range or gimp-proof recoveries. I think the demo has a good mix of playstyles to test this out, but all the more power to you if you have the full version.
That's all I can think of for now. We'll need more research for a proper guide but hopefully this works as a good launching point.
