Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Building an AI game studio: what we've learned so far

braindump.me

Post reply on HN
Print view View original

281–290 of 312 posts

Re: Building an AI game studio: what we've learned so far

#281

Post by somenameforme »

Hammershaft wrote: post #271
hiatus wrote: post #189

Earlier quoted context omitted.

> For every Kenshi there are many dozens of excellent indie games that languish. Can you list some here?

Sure! Pinning down some shared criteria for quality and obscurity might be a challenge, but I can give it a try. I'll only include games that I've played recently with an overall steam review score of >=%94 and less than 1000 cumulative steam reviews. Deadeye Deepfake Simulacrum: https://store.steampowered.com/app/1545990/Deadeye_Deepfake_... Tandis: https://store.steampowered.com/app/1297210/Tandis/ Nidus: https://s…

I'd drop the max reviews down by quite a lot. Sales:reviews tend to be around 60:1. So a game with 1000 reviews has sold tens of thousands of copies. We can look at something like the PS2 for some reality checks. It sold 1.5 billion games through 4000 titles. So an average of 350k sales. And of course that's going to follow something like a typical 80/20 ratio, which would leave us with a median sales in the ballpark of 70k! Of course Steam has far more players, but also way more games. The PS2 was a ring-fenced minimal competition domain with an attentive and [relatively] content starved audience!

I decided to look up Tandis first. First off the game has been featured on Ars Technica [1], Rocker Paper Shotgun [2], and lots of other places. So it's definitely not like nobody heard of it. But it just doesn't really seem to drag people in. The reviews, positive as they are, all seem to stop playing after an average of around 3 hours. In fact a huge number of the positive written reviews have less than 1 hour of lifetime gameplay, which is pretty weird for a game that comes at a relatively premium price. And the peak players, a month after release was 7. [3] I'm guessing the players that moved on after such short timeframes probably weren't actively recommending this game outside of those reviews. Would you have actively recommended it, outside just filtering your library and seeing it there?

[1] - https://arstechnica.com/gaming/2022/04/tandis-review-the-ult...

[2] - https://www.rockpapershotgun.com/geometry-puzzle-game-tandis...

[3] - https://steamcharts.com/app/1297210

Re: Building an AI game studio: what we've learned so far

#282

Post by reitzensteinm »

Jensson wrote: post #233
programd wrote: post #226

Earlier quoted context omitted.

Hex of Steel. Turn based hex wargame, very niche but with a fairly large worldwide community into such games. Only the solo developer is doing plays on youtube and twitch. The game has been seeing a slow upward sales trend as the niche wargame community realizes how good it is. If you've played SSI and Avalon Hill hex games as a kid you should get this. https://store.steampowered.com/app/1240630/Hex_of_Steel/ https:/…

That is an indie success, almost 400 reviews for that price is good for a solo developer, he can live well on that as long as he continues to create and sell content for. Failures are games with 50 or less reviews or a bit more if they have much lower price points, there aren't many gems at that level. Note that a large majority of indie game developers don't live in USA, indie gaming isn't dominated by American deve…

Hardcore games are much more likely to convert plays to reviews. Many of the units would have shifted during the 50% sales.

The game is multiplayer which is a whole other layer of complexity.

Based on my experience in this market, I really doubt the author is happy with the outcome as anything other than a stepping stone to something more sustainable.

I think you're an order of magnitude too low setting the bar at 50 for professional developers where this is their full time job - even outside of the US.

Re: Building an AI game studio: what we've learned so far

#283

Post by kibbi »

Hammershaft wrote: post #271
hiatus wrote: post #189

Earlier quoted context omitted.

> For every Kenshi there are many dozens of excellent indie games that languish. Can you list some here?

Sure! Pinning down some shared criteria for quality and obscurity might be a challenge, but I can give it a try. I'll only include games that I've played recently with an overall steam review score of >=%94 and less than 1000 cumulative steam reviews. Deadeye Deepfake Simulacrum: https://store.steampowered.com/app/1545990/Deadeye_Deepfake_... Tandis: https://store.steampowered.com/app/1297210/Tandis/ Nidus: https://s…

Surely you're not claiming that a game with fewer than 1000 reviews is "languishing"? Any indie game with 1000 reviews is very successful.

Picking your first example, Deadeye Deepfake Simulacrum has about 460 reviews, roughly 20,000 copies sold, an estimated revenue of about $120,000 while it's still in early access, and apparently a developer team size of 1. I'd guess that this game will net the developer several hundred thousand dollars over the next 10 years.

(Regarding multiplayer games: yes, these are a different case because they need to reach some critical mass to have a chance.)

Re: Building an AI game studio: what we've learned so far

#284

Post by somenameforme »

johnnyanmac wrote: post #279
somenameforme wrote: post #274

Earlier quoted context omitted.

No I don't, and this is one of the reasons why. When you can work online and get paid in $, that dollar goes so much further in the other 95% of the world. What people who don't travel much may not understand is that it's not just the exchange rate. What really matters is PPP - purchasing power parity. $100 might translate to e.g. 8000 rupees, but those 8000 rupees go far further in India than $100 would in America.…

>So by living outside of America and getting paid in USD you basically just multiply every dollar you make, by a very large amount. That $168k easily becomes a $million+ in the overwhelming majority of the world That's a nice sentiment for those who live in the not-US. But sadly most game development is in fact located in the US and Japan. For the US, many of the developer scene is in fact in higher cost of living ar…

Going back to the games I mentioned in the other conversation we were having, Mount and Blade is Turkish. Siralim is made by a guy living in Cortland, Ohio - population 8000. Japanese stuff is obviously in Japan, which has economically become much closer to a developing country following 3 decades of stagflation. Nominal GDP/capita is $33k. Average household income after taxes is $29k. [1] It's why I kind of snicker when people reference Japan as having 'achieved affordable housing.' They sure have, so long as you aren't working for Japanese wages!

I don't know how to get meaningful stats on game development locations (because you need to exclude shovelware), but I know that in my library of games - American companies are few and far between except for big AAA titles, though even there things are getting a lot more diverse. For instance Kerbal Space Program is Mexican, Starpoint Gemini is Croatian, Space Rangers is Russian, Conquest of Elysium/Dominions is Swedish, Endless Legend is French, Lost Castle is Chinese, Crusader Kings is Swedish, Battle Brothers is German, and so on endlessly.

And in my experience lifestyles only improve abroad. All the niceties you're used to still exist in pretty much every semi-major hub around the world, and then much more on top. As do large communities of other expats. If somebody just wanted to be in a mini America, or even Silicon Valley, each and every day - they absolutely could. Think about how in the US basically every major city has a little 'China town.' It's the exact same thing abroad with Westerners, often with a tech bias.

[1] - https://www.oecdbetterlifeindex.org/countries/japan/

Re: Building an AI game studio: what we've learned so far

#285

Post by pierd »

hdlothia wrote: post #126

This is very impressive. If i'm understanding the post they basically created an inrernal api and just have the llm use that api. Seems like a great way to workaround llms struggling to keep track of large codebases.

Yip. It's like a tiny game engine and then the LLM makes games using it.

Re: Building an AI game studio: what we've learned so far

#286

Post by pierd »

arvindrajnaidu wrote: post #81

Could you leave the tapping and chatting to the AI? This way, I can let it work on some basic guidelines and let it surprise me with a game every day.

Since it's all chat based you can actually start by asking it for game ideas and tell it to proceed with making it. Unfortunately and definitely not surprisingly the resulting game is usually quite bland.

Re: Building an AI game studio: what we've learned so far

#287

Post by antifa »

johnnyanmac wrote: post #244
antifa wrote: post #225

Earlier quoted context omitted.

That probably involves more of a corporate marketing vanguard that picks winners more so than it organicly finds them.

And you think Games are different?

The corporate marketing vanguard that picks winners in the gaming industry is almost entirely dedicated to microtransactions, pay to win, pay walls, subscriptions, and battle passes.

Inorganicly chosen winners is a problem, but in the music industry, the end product itself doesn't exploit the listener beyond just that.

Re: Building an AI game studio: what we've learned so far

#288

Post by johnnyanmac »

antifa wrote: post #287
johnnyanmac wrote: post #244

Earlier quoted context omitted.

And you think Games are different?

The corporate marketing vanguard that picks winners in the gaming industry is almost entirely dedicated to microtransactions, pay to win, pay walls, subscriptions, and battle passes. Inorganicly chosen winners is a problem, but in the music industry, the end product itself doesn't exploit the listener beyond just that.

Sure they are now, that's simply where the money is nowadays. 5 years ago it was battle passes 5 years before that it was an attempt to recoup costs with used game. 5 years prior to that it was experimenting with DLC types and seeing what would stick. And 5 years before that it was trying to have long term subscription attachments to a game.

Games are ultimately tech. And games grew along with the internet. It was inherently going to be run more like a tech company than a media conglomerate for that reason. I see it less as games being exploitative than a double edged sword, like the internet itself. It can be a lot more exploitative. But it also opened up entire mediums of ideas and arguably tore down world borders.

Re: Building an AI game studio: what we've learned so far

#289

Post by jncfhnb »

stubish wrote: post #270
jncfhnb wrote: post #14

Earlier quoted context omitted.

The majority of those games are now visual novels / pornography.

Does that actually matter? You see the same sort of embarrassingly low numbers in all artistic mediums. Writer, painter, musician, sculptor, interpretive dancer, comedian, indy game dev - on average none of these pay enough to count as a career and you only hear about stuff on the successful end of the bell curve. Even if you improve your odds by making commercial art. In many cases even the artists considered succes…

It matters in the context of the original claim. Median revenue per game has not changed for the genres that are not VN and porn games.
Post reply on HN