Complete nothingburger of a study, which itself is locked behind a $25 paywall to access it. And the author of the article obviously didn’t cause there’s 0 mention in the article itself about the methodology used to determine the 20% revenue lost (nice round number might I add). The only thing that even alludes to the methodology used in the abstract is

When Denuvo is cracked very early on, piracy leads to an estimated 20 percent fall in total revenue on average relative to an uncracked counterfactual

Which really doesn’t tell us much, how are these counterfactuals selected in the first place? What is the cirteria? How are you determining that the differences between revenue of a game that was cracked and that went uncracked are due to one game being cracked? How can anyone even confidently claim that they’ve normalazied the data set enoguh that these differences in revenue are mainly caused by a game being cracked, especially with how rare early denuvo cracks have been in the past few years. Statistically this sounds dubious at best, especially when we have fully open studies (like the one funded by the EU a few years back) that have found no statistical proof that piracy has any impact on revenue ( with the exception of box office revenue of big new movies being leaked and pirated while still in theaters). Surely they wouldn’t have missed a 20% meadian difference in revenue.

Lastly you have major tech news outlets all reporting on a study less than a month after it was made available online. For context the journal containing this study will only be published in jan of 2025.

Didn’t even open the link probably

That’s a fair point, I guess I used binary numbers so much i uni that I just know the small ones by heart and that’s why I find it easier. Following the example, I never convert 101 as 4+0+1, I just see it and know it’s 5.

Read, write, and execute are represented by the numbers 4, 2, and 1, respectively, and you add them together to get the permission

Maybe I’m the weird one here but this seems like a counter intuitive way to remeber/explain it. Each octal digit in the three digit number is actually just 3 binary digits ( 3 bit flags) in order of rwx. For example read and execute would be 101 -> 5.

I’m on ddg and get no such issues with the same query: