Post by KozmoWorks in Bot/crawler activity detected on IndieDB/ModDB!

Viewing post in Bot/crawler activity detected on IndieDB/ModDB!

I’m a developer on this platform too, and I completely agree with Ofihombre here.

When you spend time making games or mods, you look at your stats to see if real people actually care about your work. Seeing a sudden jump of 1,000 views but getting exactly 0 downloads and 0 comments is just a ghost town effect. That's not normal human behavior, and it's definitely not just "people using private browsing."

Redoni The reason this is a big deal for us devs is that it ruins our data. If we can't trust our view counts, we can't tell if our projects are actually getting popular or if it's just bots scraping the site. It makes the whole dashboard useless.

Thanks for showing the data from Garry's Mod and World of Padman, Ofihombre. It proves this is a real platform issue and you’re definitely not the only one noticing it.

Like Reply

Ofihombre5 days ago

Thanks for helping me, although we also have to understand Redoni's point. He was trying to demonstrate points that perhaps it works in a different way than I do in the case I am reporting on.

Like Reply

KozmoWorks5 days ago(+1)

Well said, Ofihombre. And yeah, fair point, Redoni raised some good arguments about how small stats can be super random and tricky to read. It's cool to see how different people look at the same data.

In the end, I think we all just want the platform to be as accurate as possible for creators.

Like Reply

redonihunter5 days ago

Complaining that the statistical data on that platform is useless is not the same as calling out the platform as useless. When searching a game from OP that game's indiedb page only appeared at #9 in results. That "low" number would be the indicator that indiedb is far less important than Itch or Steam. But it does show up. People preferring that site can go there and look up information about the game there. If they download there, is another question. I would not. I would go to other places for the actual download.

My main issue was the chain of logic. It does not hold up. You can't infer bots from those patterns and the other oddities. The only thing you can direcly see, is, that people do not use indiedb to download a game very much and there are a lot of other plausible reasons for that. Also, there actually was a rise in downloads with the November spike correlating with the bigger spike in November.

Like Reply

KozmoWorks5 days ago

Redoni i think you're still missing the most definitive proof here. You're saying we can't infer bots just from patterns, but you are forgetting that Ofihombre literally stated the DBolical administration already confirmed the issue and actively tried to block this specific traffic.

Platform admins don't spend time writing filters and blocking IPs for "random fluctuations" or "private browsing humans." If the people running the servers classify it as bot traffic and take measures to stop it, then it IS bot traffic. That completely validates the logic.

As for Google ranking, ranking #9 doesn't justify a platform having corrupted analytics dashboards for its creators. We aren't arguing about whether the site is completely useless to find a game; we are arguing that the stats system is broken, which is exactly why Ofihombre's points are valid.

Like Reply

redonihunter4 days ago

that Ofihombre literally stated the DBolical administration already confirmed the issue

Then why not simply cite this. Instead a non conclusive mess of oddidites is presented as evidence. Did those admins say how much such traffic there is, and if they remove such traffic from statistics?

We aren't arguing about whether the site is completely useless to find a game; we are arguing that the stats system is broken

No, we do not. You might. Ofi is not. And I am pointing out the flaws in Ofi's logic.

The thread's title is:

IndieDB/ModDB is a bot/crawler farm! Don't public any game/mod!

I think that is supposed to mean publish and not public. Ofi calls for other people to not use this platform and claims it is used to simulate traffic, rather than there being genuine traffic.

Ofi is not arguing that the statistics of that site are broken and should not be used for marketing analysis. It is a call to boycott that site.

Like Reply

redonihunter4 days ago (1 edit)

Redoni i think you're still missing the most definitive proof here. You're saying we can't infer bots just from patterns, but you are forgetting that Ofihombre literally stated the DBolical administration already confirmed the issue

That there is a confirmed bot problem on this site does not mean, you can infer it just from patterns. One does not follow from the other.

And most importantly, I do not happen to see those alleged patterns. What I see is perfectly normal traffic to download ratio for an indie game on a less popular platform.

When the views go up, so go the downloads. That directly contradicts this assumption that the views are mostly patterns of bot traffic.

So from what I know currently, I assume that most of this bot traffic is either removed from the statistic, or the games of Ofi are just not hit by that fake traffic.

I did check out a few other games from the currently popular sorting of that site. They had even less traffic than Ofi's game. And no such patterns.

Like Reply

KozmoWorks4 days ago

You are replying to yourself now, which shows you're just trying to force your theory.

You say downloads went up with the views, so it can't be bots. But as developers, we know exactly how web crawlers and modern AI scrapers work. They don't just look at a page; they script click every single link on it, including the download buttons, to scan the files and data. That’s why both metrics inflated artificially at the exact same time.

Ofi already proved with the Garry's Mod and World of Padman charts that these spikes are synchronized across completely different games. If you choose to ignore synced charts, admin confirmation, and basic web crawler behavior just to protect your own opinion, that's fine. But the developer community sees the ghost traffic for what it is.

I'm done debating this.

Like Reply

redonihunter3 days ago

I replied to you twice, instead of editing the post. You need to look at the indentation which post was replied to.

And please look at the data. I did align the 3 games above https://itch.io/post/17400576 There is no pattern. There is no proof.

Please understand the difference between opinions, facts, conclusions and proofs. The pattern was used as a premise for argument. The premise is wrong. Any conclusions based on a faulty premise are formally wrong. The conclusion could still be true, but not by that chain of logic.

Like Reply

KozmoWorks3 days ago

i think i said am done debating

Like Reply

Ofihombre4 days ago

I admit I was sensationalist in choosing the title to grab attention. And what you call a boycott wasn't done to make the site sink even further; what I wanted was for it to improve and regain its credibility.

Like Reply

redonihunter3 days ago

You can always delete the thread, if you want. Accusing a site to be a bot farm is not the way to make it improve. It implies that you might think the site owners are responsible for the bot activity. Maybe to cash grab on ad money, so they are acutally scamming the advertisers with inflated views.

Also, I think the aligned graph shows that your premise of patterns is not a foundation to argue.

At most I would agree that the statistics of that site are questionable and not good for marketing analysis. Maybe that angle is more promising for improvements.

Like Reply

Ofihombre3 days ago

Well, I don't think deleting the thread is the best option, but I will change the title since the one that says "IndieDB/ModDB is a bot/crawler farm!" was more due to the excitement of the moment after seeing so much traffic inflation for a year (besides updating some things too).

I may not have found enough evidence, but I'm satisfied that you at least acknowledge that the statistics are questionable.

Like Reply

redonihunter3 days ago

Of course they are. Most "problematic" imho is the links to Steam (and Itch). Views on that site can mean anything. Or nothing. It is not a primary release channel. It's even in the name (at least for me). A site with db suffix is an information site in my eyes. Not a download site.

If the site tries to be a publishing site, this was not my impression. But I did not spend much time there.

As a recommendation for your r&m site. Ditch the older downloads. That many downloads confuse people. Have a stable and an experimental version if you must. Similar to your Itch page of that game.

And as I said before, you do not need to find extra evidence. You can just use the verbatim statements of the site admins. If they would say something like "we had a spike of suspicious bots mascerading as guest users in early November", you could corraborate that, with "yeah, I even see a spike in some games at that time, which might be that activity, so that problem those admins talk about seems real and it is affecting this and that statistic".

Like Reply

View more in thread