Show HN: I scrape Steam data every month and it's yours to download for free | Hacker News Reader

Show HN: I scrape Steam data every month and it's yours to download for free

gginsights.io

161 pointscsmets60 comments

Yeah, there's AI, but I added it because I found it easier to find answers I'm looking for. For the data scientists, you can download the CSV and go crazy. Would love to know what discoveries or learnings can be found from it.

To download the raw scraped data you need to become a paid member but you don't really need it unless you're wanting to finesse a table of data for a particular need. The cost is mostly just an incentive to help me pay the bills for running the website.

The bunch of available CSV files contain large amounts of data which has everything from tags, genres, pricing, wishlists, estimated revenue, etc. It's what the AI is reading from.

Hope you find it useful :-)

Open articleView on HN

Do you have data that https://steamdb.info/ doesn’t have?

Steamdb lacks an API for one, and the devs officially have a policy that they'll never make one, saying you should just scrape Steam directly instead of bugging them about it[0].

It means that steamdb, while extraordinarily useful for casual prodding at what's stored on Valve's servers, isn't very good if you want to run data analysis or something like that on the metadata of Steam games at scale.

Not sure if it's legal to charge for the raw scrape when OP doesn't seem to be affiliated with Valve, but that's not up to me to figure out.

[0]: https://steamdb.info/faq/

That seems pretty reasonable, it's their data, they just make useful visualizations

This whole time I was under the impression that SteamDB was owned by Valve. Huh.

All run by one person who doesn't even take donations for the service they provide.

They joined here on an Ask HN about "Those making $0/month or less on side projects – Show and tell"

https://news.ycombinator.com/item?id=34548908

If you need to be a paid member to download csv file, then it is not free :) lol

If you need to make an account and give this guy personal information (a digital commodity like oil) to see the data it's not free lmao

> If you need to make an account and give this guy personal information

In this case, you don't. That's just to weed out people who can't figure out temporary emails. I just used one to create an account without turning over any PI.

free tier allows you to download the csv.

I got some answers that weren't specifically about my questions in some instances. As someone who's just trying out the free demo, it's not a big deal, but maybe you can provide a way to flag answers for to redeem their credits? It would probably increase retention and help people chase down bugs.

I guess the main differentiator over steamdb is getting the data in CSV?

Might be good to clarify in the FAQ because the people I know who would pay for this are not the most techy types.

Hi, I'm interested in scraping steam too. Do you have the scraper code available open source or one you recommend?

Have you looked over the data that OP is providing here and determined that it doesn't meet your needs?

Generally it's polite to avoid scraping if you can help it, so I'd start by considering whether OP is already providing what you are looking for.

On the other hand you need to be a paid member to download the raw scraped data, so it isn't unreasonable to want to learn how to scrape it instead.

Good idea, let's save $5 and sink dozens of hours into building our own instead

Isn't that the hacker spirit, wanting to put things together yourself? Let's revisit the comment that started this thread:

> Hi, I'm interested in scraping steam too.

I was responding directly to your objection with having to pay for it as the reason to do this, not hacking for the sake of hacking.

Hacking for the sake of hacking VS hacking for the sake of saving money, why does it matter?

Jesus Christ, it's 5 bucks. 5 bucks is not a reason to roll your own version.

Roll your own version because you want to roll your own version, not to save 5 bucks.

you seem to be really hung up on the 5 bucks while at the same time being angry that people are hung up on the 5 bucks, it's just 5 bucks man, if people want to pay it or not doesn't matter, let people hack it away if they want

Its not $5. It is free or $5/month. The reason OP wants to scrape doesnt matter. His question was reasonable to ask here and could lead to finding out about some open source projects. You have not added anything to the conversation besides being wrong about the price.

Let's expand our own skillset by investing time instead of money (paying someone else). Sounds like a reasonable proposition to me.

I wrote a simple scraper for a 'steam game semantic search' app I built a while ago.

It definitely won't fetch all the data that this person does though. It only fetches the current list of games on Steam, their store page information and some reviews for the game.

The code quality probably isn't amazing, but it might give you an idea of how to get started with your own scraper.

https://github.com/Netruk44/steam-embedding-search/blob/main...

Thanks! That's perfect, just want somewhere to get started.

https://steamdb.info/faq/#how-are-we-getting-this-informatio...

I found this explanation from steamdb that points to the various projects and libraries they use to gather all the data they have. It's not a how-to, but it has very useful info.

I had to refresh before posting, because I wanted to see if someone else beat me to being that HN commenter but...

From the Terms of Service (emphasis mine):

6. Restrictions on Use

You agree not to:

    Use the Service for any unlawful purpose.

    Attempt to reverse-engineer, modify, or *create derivative works of the Service.*

    Share, resell, or distribute downloadable data provided by the Service without explicit written permission.

Do you intend to delineate the data provided by the service from "the Service" itself? It seems most fair that data received via Fair Use remains in that arena, pun fully intended.

That aside, it's an intriguing dataset nonetheless, but I'd prefer to see a sample of the data before signing up.

At a glance, it appears the product is the “chat with the data" feature; The CSV is free.

What I don't understand is the difference between 'Download all CSV data' in the free tier and 'Download CSV data' / 'Download raw data' in the paid member tier. It seems that the free CSV data is likely an extract or digest of the raw data offered as a sample.

I might be inclined to seek the raw data, should it be more cost effective than scraping Steam myself.

Being a user, free, paid, or anonymous, can still be under the thumb of their ToS, especially so if they force a dialog in front of you to agree to the ToS while signing up. I'm merely pointing out hurdles to the OP that may obstruct some of the people they are trying to reach.

Steamdb.info displays graphs etc. Is that considered a “derivative work”?

I am not sure what is considered derivative work and what isn’t

IANAL but I am someone who deals heavily in 1) scraping and 2) data and the analysis, enrichment & brokerage thereof. As such, I like to consult this for anything regarding US Copyright law: https://www.copyright.gov/circs

Circular 14 addresses derivative works, including those based on data: https://www.copyright.gov/circs/circ14.pdf

Steamdb.info is a derivative work, yes. And scraping is usually accepted as Fair Use, so both services are presumably within their rights, but they have no claim to the underlying data, only their process of enrichment. If someone were to build a new service based on the data presented on either site, there's not much they could do to stop them... short of getting them to agree not to do so via their ToS.

OpenAI is a great example of a company who built a derivative work on scraped data available under Fair Use, and then subsequently gated their data via their ToS. With such a popular precedent at play, I'd rather not use any services doing anything similar, especially when steamdb.info doesn't even have a ToS.

Thank you. Does this still hold good if steamdb was making money (ads, for example)?

Also, I am wary of using big companies like OpenAI as precedent. Big companies can do whatever they want and get away with a lot of stuff that individuals and smaller companies can only dream of

Yes, within some limits, but if one were to set up a business like that, it's a very good idea to seek out a consultation from a local copyright lawyer to know exactly what one can and can't get away with. Datasets are addressed as a "collective work", which lumps them in with everything ranging from art books, to hackernews, to scientific journals.

Personally, I wouldn't sell anything I gathered from a publicly available source anyways, mostly out of principle, but doubly so if that source is as well-paid as Valve.

> Personally, I wouldn't sell anything I gathered from a publicly available source anyways, mostly out of principle, but doubly so if that source is as well-paid as Valve.

Market reports are an entire industry, and people pay for them solely to avoid ingesting a tangential domain. It's ok to sell your transformations.

My advice is free, my custom tooling is dirt cheap with public examples, and my finished product costs money every month. It's basically price tiers based on your interest level.

Thank you for highlighting this. I've updated the terms to align with the values of this service.

Regarding Steam data, I am curious about how games are being played (hours spent) and, even more, about their co-occurrence (i.e., player X spent both time on game A and game B). I would love to make a visualization like https://p.migdal.pl/tagoverflow/?site=gaming&size=32, but for Steam data.

Also, for deeper insight than sales volumes (e.g., game design, general trends, demographics, types of players), such things would be crucial.

and

It is not that difficult to scrape Steam data using SteamKit, right? I build a website around Steam data a couple of years ago, with a small scraper app which was hourly scraping new updates (using SteamKit) and putting it into my database.

The biggest advantage that SteamDB has, is that it has a ton of historical data. That is not retrievable from the Steam Network, so the only way to have gotten historical data is to start early.

My website is now defunct for a year, but I've kept the scraper running. I now have 7 years of historical data in my database.

Can you please provide examples for the raw data? As a user, I would like to know what I'm buying before paying.

It seems to be missing reviews? I have always thought about building my own recommendation engine from steam data, given how steam's own recommendation never works for me.

Nice! It would be nice however to see more detail about the data you collect and what exactly you provide on top of it using AI or through aggregation etc

Out of curiosity, what formula did you end up using for reviews:sales? I've looked into this a bunch and it's a very tough problem!

> Yeah, there's AI, but I added it because I found it easier to find answers I'm looking for. For the data scientists, you can download the CSV and go crazy.

This is kind of the only way I use AI really, to summarize things, and extract details, then review from the raw sources to make sure the LLM isn't misleading me. I find myself using this approach instead of Googling for things since Google crippled their search the last few years, it feels like every year its harder to find things with Google. I miss 2007 Google...

Give Kagi a try, it's basically Google before it went to shit.

You use the chat but the credit used isn't updated immediately in the lower left.

Question for OP, or anyone that considered it:

Do you think Steam reviews are coordinated?

I think for basically any possible online discussion, from Facebook to Hacker News to Steam Reviews, you should always keep in mind that some portion of it is probably astroturfed, to some scale

Anything from a small indie game to a huge AAA title, you can bet that the creators got their friends and family to post some nice reviews early, just to give it that positive bump

I was specifically alarmed by what looked like review bombing of a indie game. I just can't imagine it. I need to write a small llm plugin that collapses coordinated/astroturfed reviews.

The smaller the scale the easier to astroturf, honestly

If there are only 20 reviews it's pretty easy for one person to review bomb on their own if they want to

It gets much harder when there are 2 million reviews

> Do you think Steam reviews are coordinated?

Yes. It's not even a question. Steam flags outliers too.

https://store.steampowered.com/app/281990/Stellaris/

It got review bombed starting on Feb 14th because a different game that the company makes (HOI4) released DLC that upset the sensibilities of part of that player base. ( https://old.reddit.com/r/Stellaris/comments/1iqzih8/why_is_s... )

---

There are Steam review bots for discord ( https://www.codecks.io/steam-bot/ ) and that also encourages people who are members of a game's discord to leave a (positive) review.

---

It's a certainty that reviews are coordinated through a number of different means.

I thought this would a stavros post. Thanks for your efforts!

> I scrape Steam data every month and it's yours to download for free

Does not line up with

> To download the raw scraped data you need to become a paid member

Sooo, clickbait or just plain dishonest?

ye clickbait first i saw was pricing. this should be deleted (thread)

A CSV of this data is free to download per the pricing page, whereas the raw data (not sure what that looks like versus the CSV) requires a paid account.

So I guess it depends if you consider the CSV as fundamentally different from the raw data in a way that makes this clickbait.

> To download the raw scraped data you need to become a paid member

If I have to pay to download the data how is it mine to download for free?

looking into this thank you.

nice

Do the HN crowd read "for free" on the title, click it, scan the page in a milisecond, see "Pricing" on the top, and come back to complain in the comments? Geez

Yes, it's disingenuous. Free means $0.00.

There is a hope I would see just a single bold FREE when I click pricing still, so I check it first before complaining.

And yes, "for free" means free as in the air you breathe for free. Anything else falls into a "mousetrap free" category.

Reply on news.ycombinator.com