Hacker News Reader

Comment by douche | Hacker News Reader

The bottleneck is art, from what I understand. It's not trivial to generate enough skinned, rigged, and animated, good-quality 3D models. Even if you can reuse the bulk of your engine code, or even build off an existing game, as in the case of total conversion mods, the art demands have been getting steeper and steeper over the years.

Looking at really prolific game studios, they tend to be in less art-intensive genres. 8/16-bit RPGs, pixel-art platformers, etc.

And the people that really crank out games on a limited staff and art budget tend to reuse the same engines and assets with minor iterations. Think Spiderweb Software[1], or SSI with the Five Star games[2].

[1] http://www.spiderwebsoftware.com/

[2] https://en.wikipedia.org/wiki/Panzer_General

View on HN

This might be true on a small scale, but for a large AAA-sized studios economies of scale kick in. I'm not saying it's not expensive, but beyond a certain scale it becomes less prohibitive.

Art costs (measured in man-hours) are one of the more predictable aspects of game development, they scale well and can be outsourced easily.

Maybe that's why Hearthstone was such a success, models and art for all the characters already existed!

You have to give it up that team: A lot of people complained about how simple Hearthstone is compared to Magic, but the reality is they did a pretty good job in creating accessible card game that works on multiple platforms.

I certainly have my criticism for the game such as the slow card releases, RNGfiesta, and lack of QOL updates, but ultimately the game is really well put together.

If you have evervplayed MTGO you might see that Magic's complexity doesn't transition super well to digital.

Though Hearthstone does have its own set of issues like you mentioned, I'm very glad they drew the lines on complexity where they did (notably no user interaction on opponent's turn except for secrets)

What's a glaring problem from a cardgame standpoint and not due to simplification is there is no defined order of action in a turn. You can play cards, attack,trigger effects and then play other cards immediately and the outcome of such actions is not deterministic.

An exact sequence of events can be replicated with the timing of them altered and you'd get different outcomes.

Whole board affecting secondary effects can also have unpredictable interactions with cards in play's triggered abilities.

In MTG parlance, the "stack" is broken in Hearthstone.

I play hearthstone quite a lot and what you are saying is absolutely not true. I'm guessing what you are talking about is that the game let's you play ahead while animations are still running, but the actual effects and results are the same regardless of the animations. It can make it appear that if you play a destroy all effect and then a minion the animation looks like it should destroy it but of course it doesn't since the animation isn't representing the game state.

Unless you are talking about something else and I misunderstand you? Maybe something like the order of deathrattles (they are triggered in order of the cards played, just like the few instances of time-stamp-matters interactions in MTG). Or do you mean how the max minion limit works where "dead" minions are still there during "on damage" effects but not during deathrattles? There is a few nonobvious rules like that, but I'm pretty sure the rules are actually pretty solid and not bugged.

At a concept level they existed, but the artwork was new.

I mean, everyone at Blizzard knows what an "Orc Warrior" should look like. Green skin, muscles, some type of weapon. But someone had to actually create new drawings of it for the game.

This means that the main challenge behind coming up with a new franchise is to come up with a consistent, compelling concept for its various elements.

A large amount of art from the original set was incidentially cropped and reused from the ill-fated WoW TCG.

I have heard this as well, it suggests the single most impactful thing someone could build today would be a tool to generate game art cost effectively and quickly.

I would certainly throw money at such a tool, but the question then becomes scope. Autodesk has their Character Generator and Adobe bought out Fuse that has a competing product. But there is still nothing for backgrounds, interiors, etc short of modeling software and (if you have money rather than time) purchasing quantities of assets that you then still spend time customizing.

This is really interesting considering the "starving artist" stereotype.

Is the problem that there are lots of mediocre artists but not many who can actually execute projects? Or that it is hard to parallelize things like this? Or that once you parallelize too much, the amount of value added by a marginal artist is not enough to pay them?

>Is the problem that there are lots of mediocre artists but not many who can actually execute projects?

I think part of it is that standards keep increasing. Compare this demon concept art from Warcraft 2 (1995): http://vignette2.wikia.nocookie.net/wowwiki/images/c/cb/Daem... to this one from WoW: Legion (2016): https://s-media-cache-ak0.pinimg.com/736x/56/f2/7f/56f27f98a...

There are probably a lot of artists that can execute at the 1995 level who aren't employable in 2016.

It's just a hard problem, especially when you have to try and get a number of different artists to maintain a single style for a project. Disney was struggling with this problem for decades and financially they were hardly breaking even with their studio part of the business because of the cost of generating all the art they needed. Computers lowered the costs and the manpower quite a bit, but then increased the quality expectations.

There's also the problem of working within technical limitations. Even with standard file formats there can be much more involved than just dumping files into a project.

I would imagine artists need to be involved in technical discussion and would need to learn & bug-test custom software for importing and processing assets. It seems like a different skillset to traditional art & design backgrounds.

For a finished game product, absolutely. But for a prototype having access to the in house library of thousands of models, animations etc is a huge step up.

Reply on news.ycombinator.com