So.. good points. It's something we discuss a lot internally. What's the scope?
Thing is, VU's do take a while. Especially depending on the champion. In general, here's some basic things that can effect scope:
Concept iteration - How many variations or possibilities of the champion are there before we settle on a direction and start production?
Model complexity - Does this character have multiple states/stances/etc? How many actual models and components does the champion need? Guy with a sword is a lot simpler than girl with 3 weapons.
Animation complexity - Does this character require significant secondary or tertiary animations such as cloth movement?
VFX complexity - Does this character have a ton of VFX? A melee champion might have less complex of a VFX load than a caster.
Skins - How many skins does the champion have and what complexity level are they?
Those are a few points that we mull over. VU's, in terms of art production, are as much work as an entire champion minus the kit. Essentially a VU is a kit-locked champion. When we go through our lists of potential VU's we grade the existing champion. Our producer had a cool idea to create an excel spread sheet where different artists from the various disciplines (Concept, 3D modelling, Animation, VFX) to grade each of those aspects of the champion. We did that, then cross-referenced those grades across different aspects of the champion such as visibility, pick rate, etc. Depending on the variables you highlight, different champions come to the top. We don't absolutely abide by what the document tells us, it's just a tool to quickly parse the data.
Generally, if a champion gets a poor grade across all of these categories we take action and take them in consideration for pre-production.
Thing is, there's some champions which while they don't have failing grades across those parameters, we still want to give them love. However a full-blown VU might not happen to them because they technically aren't some of the worst offenders. In these cases is where we consider model updates. If a character's rig is healthy then we can make a new model that fits the proportions of the rig and create a cleaner, more detailed, better textured model. While a VU that requires creating an entire character from scratch can take months, depending on the criteria I've listed above, a model update might take weeks.
This is all still a discussion we as a team are having. What's the best course? We want to be constantly delivering value to players. Is that done best by inching through characters one at a time through VU's? Or is it a tiered approach that included texture updates, model updates, and complete visual updates? I haven't yet decided. Though my personal thought is that some champions don't need the complete overhaul. I want us to be able to deliver maximum satisfaction to players, and not let thier favorite champions stay untouched for years. Texture updates and model updates are a way to compromise and optimize.
Would I like us to be able to burn through VU's faster? Yes, of course. However, right now we're only in a position to deliver the cadence of big projects at our current pace.
