[Manuel_D]

To be clear, what is measured is the proportional change in sales following the release of a crack, relative to other released games that went uncracked for longer. I don't know what percentage of games removed DRM after the crack was released, but let's assume they do.

If there's Y amount of people who buy if there's no crack available and pirate if there is, and X amount of people and who refuse to buy the game with DRM but do purchase it after DRM is removed. The net shift in sales is the difference in sales is X-Y.

We're not measuring Y, we're measuring Z. And since Z is negative, we can conclude that the group of people who just want free games is larger than the group of people who buy after DRM is removed. We know that Y is larger than X.

[AnthonyMouse]

> If there's Y amount of people who buy if there's no crack available and pirate if there is, and X amount of people and who refuse to buy the game with DRM but do purchase it after DRM is removed. The net shift in sales is the difference in sales is X-Y.

Now you're assuming that all games immediately remove DRM after they're cracked, which is definitely false. In a statistical study it would need to be all of them or any that don't would be skewing the average.

And again, removing the DRM doesn't fully undo the hit from having it to begin with. Existing negative reviews or forum posts panning the game don't disappear the instant you address the thing they were complaining about. The network effect and word of mouth in subcultures that don't buy games with DRM is already reduced.

[Manuel_D]

So if a game launches with DRM it's forever tainted and that group of people who supposedly would have bought the game absent DRM won't buy it - even if the DRM is subsequently removed. But on the flip side, if a game doesn't launch with DRM then it's pirated immediately and there's no way to measure the sales in the period pre and post-crack, because there effectively is no pre-crack period.

You've constructed an unfalsifiable claim here.