but rather for those bots and crawlers that impersonate Mac users and are counted as "guests"
And you have any proof for that? What's the motivation? Who is gaining from this? Who does this? Why impersonate mac users? Would it not be more clever for bad bots to impersonate the expected user agents representing a general audience?
You just said the site was previously more popular. Maybe it is still popular for mac users for whatever reasons. At least more popular than for windows users, that is. A few hundred users is not a big number to do statistics. A single school or university fitted with computers for the students could dominate such a statistic.
What I was trying to demonstrate in the patterns between the statistics of my games, Padman and Garry's Mod, are due to the bots from guest counter
You build a premise that itself is based on assumptions. What you showed as evidence is not conclusive and the big flaw is: even if there was a correlation between games, you could not prove that it would be because of a bot traffic wave hitting those games.
So am I to believe that this alleged bot wave traffic selectively hit some games with 2000 hits and others with 200 hits and still others not significantly at all? How does your hypothesis explain this mismatch of supposed bot activity?
This is almost as bad as: I see lights in the sky, therefore aliens. You just jump to your conclusion, it must be the aliens bots that are visible there.
I am critizising your chain of logic here. Your observations are not explanations. They are observations. And the only thing you show here is, that the view to download ratio is rather bad on that site. And even that is not true, it seems.
AI are good for that sort of summary. I asked what ratio to expect on indie gaming sites. It gave me these numbers.
1–3% downloads per view — common for an average listing
3–10% — solid performance
10%+ — very strong, usually a highly targeted audience, compelling page, or game-jam/viral traffic
Below 1% — not necessarily bad; often caused by broad exposure, weak store-page messaging, paid games, or visitors who already own the game
And your so called unusal spikes of bot activity fit this ratio. You do have like 1 download per 100 views. Sometimes 4 downloads for 53 views. Then one day 216 views and 3 downloads, next day 162 views and 1 download. And the traffic below 50 often has 0 downloads.
Nothing in the data of that Randy game seems suspicous to me at the moment. You do not have 1000 views and 0 downloads. And if your views pass into the hundreds you do have a proportional download ratio near 1%, which is to be expected.