[CiteAndPrune]

*heeds the warnings issued* Thread-hogging I never suspected. I'll keep this in mind from now on for sure... CAP is so different taking part in compared to what it looks like just casually following it.

1. What should the build of our Pokemon be? (eg. Bulky offensive, hyper defensive, etc)
I think a good rule of thumb is to give it inherent weaknesses that on the one hand, can be expoited by the enemy, and on the other hand, can be compensated for if the player puts effort and skill into CAP4. In general frailer Pokemon can't afford to make mistakes because they risk getting KOed, so a fast and frail risk-mon would work. On the other hand, low Speed introduces inherent risk also, and relying on moves like Sucker Punch or even Trick Room to be effective is risky but the payoff is good. There's a risk to using a Pokemon with no real offenses too (Chansey is a good example) because a Taunter can practically shut you down.

In short we have a lot of options, but whatever route we take, the build should have exploitable holes in it that EVing can't fully patch up. But EVs and Nature should be part of the risk management part of the project, so players must strategically decide whether to enhance CAP4's already strong sides or to offset the weaknesses a little bit.

2. What facet of strategy should this Pokemon's focus be on? (eg. Support, hyper offence, etc)
I still think that a setup sweeper is a good fit for the concept we're working with... however some arguments presented above make me think other routes are possible. Risky support is definitely viable (Prankster and Magic Bounce users are risky but rewarding when played right) or a blend of these two.

3. In which parts of the process should we focus on Reward, and in which should we focus on Risk? (eg. Low Speed so reliance on set-up moves, etc)

I thought about this before the question was raised, so I'll try to be brief here while giving reasons also:

Base Stats - These should be mostly focused on risk, leaving an exploitable weakness, but some reward should be allowed here too (such as frail but strong risk-mon). Base Stats will always be the same for CAP4 so players can't really do much in this area (aside from choosing EVs and Nature meaningfully, especially if Speed is placed at an interesting break point). By making the Base Stats risky we will make CAP4 itself naturally risky to use.

Typing - It should lean towards risk, leaving many weaknesses to common attacking types, but with a handful of useful resistances that can be used right too. An example of the level of risk I'm looking for is Mamoswine's Ice/Ground typing: weaknesses to Water, Fire, Grass, Fighting and Metal are major setbacks, yet the handy Electric immunity and (Stealth) Rock neutrality come in handy and give it some switch in opportunities allowing it to perform throughout the match, with the fringe benefit of being immune to weather damage. This is because the typing, like the stats, is set for CAP4 from the start and players can't really do anything to change it (maybe compensate for it through Air Balloon or something?) so focusing on risk here will keep CAP4 naturally risky.

Ability - Here I believe we could shift the focus more to reward, and/or risk management. If CAP4 gets more than one ability, the choice of its ability will have a strategic purpose behind it, reflect the player's decision. In fact, I would support giving CAP4 2 abilities that represent risky rewards (say Hustle and Unburden) and a Dream World ability that's entirely concerned with risk management (say, Anticipation to alert it to super-effective coverage that could spell its doom when it would otherwise be safe to use it). More freedom here will give more opportunities for players to tailor CAP4 to their own style and make meaningful decisions about what kind of risk they want to take and how exactly to manage it.

Movepool - This is the area where we could most compensate for all the previous weaknesses. Giving CAP4 a lot of moves to choose from, including pairs of safe + risky moves to let players decide what they adopt, will make it unpredictable. Some risky moves in particular rely on being options in a mon's arsenal that are good enough to use but not 'standard' because they have competition: coverage options could be given up in place of Counter/MirrorCoat+FocusSash, or Trick, or Perish Song + trapping move (which is risky to use, as opposed to the safe option of a trapping ability) or SubTorment (relying on the fact the enemy only packs one key move to hit it with)... there's so much room for variety and for player skill to show in selecting just the right moves to carry out a winning strategy. With this in mind, I will be supporting a broader movepool if possible.