I’m sure there is a misunderstanding.
You cited a piece of a paragraph from the “Main conclusions” section of that document:
“In general, the results do not show robust statistical evidence of displacement of sales by online copyright infringements.”
I just cited how that same text (which appears only in that section of that document) continues and what’s around it:
In general, the results do not show robust statistical evidence of displacement of sales by online copyright infringements. That does not necessarily mean that piracy has no effect but only that the statistical analysis does not prove with sufficient reliability that there is an effect. An exception is the displacement of recent top films. The results show a displacement rate of 40 per cent which means that for every ten recent top films watched illegally, four fewer films are consumed legally.
I would like to sum up the actual results of that document. It contains an executive summary about the results and several detailed sections about specific aspects of the study.
Executive summary (results about all media):
The executive summary provides final statements that apply to all media, including games. The paragraph that both you and me cited comes from this executive summary:
In general, the results do not show robust statistical evidence of displacement of sales by online copyright infringements. That does not necessarily mean that piracy has no effect but only that the statistical analysis does not prove with sufficient reliability that there is an effect.
This paragraph means that the method used to measure the effects of piracy has provided results that are not sufficiently reliable. Again, this applies to games as well.
If you read the study, you’ll also find why the results cannot be considered reliable, because each section of the document explains which kind of data was collected, how much the acquired information can be considered reliable, which features suggest that some data might be not reliable, in which way this fact affects the results of the methodology chosen by the authors, and so on…
Displacement rates
This section provides, among other things, two main results:
-
estimations on how much legal transactions are affected by illegal transactions
-
how much the estimations probably match with reality.
About the estimations (point 1):
These effects are significant at even the 1 per cent level and are similar for illegal downloads/streams and games played on chipped consoles: 34 and 38 extra legal downloads per 100 illegally accessed games, and 60 and 63 extra cloud games played per 100 illegally accessed games.
[…]
The overall conclusion is that for games, illegal online transactions induce more legal transactions.
[…]
[About all kind of games] Games: out of every 100 online copyright infringements, 24 induce an extra legal transaction.
About how much the estimations match reality (point 2):
It is striking that all coefficients are insignificant. This raises the question to what extent estimates suffer from problems of respondents to recall exact numbers of transactions, an issue that also applies for individual channels discussed earlier. This means that estimates are only most likely effects given the data, but with no certainty that the true displacement rate is close to the estimate. Therefore the conclusion is that not too much significance should be attached to the estimates.
In other words, the authors of the study are telling that we might be in the presence of of a very famous phenomenon. This is the reason why in the executive summary, the overall conclusion (which applies to games as well) is “That does not necessarily mean that piracy has no effect but only that the statistical analysis does not prove with sufficient reliability that there is an effect.”.
Why some articles report the result of the study as if it was a reliable one?
I’m not referring to the article cited by you, which I have not read, but about other articles that I’ve spotted online. They don’t report the part of the study that states that the results are not reliable. Why?
Now, trust me on the following statements, because fighting misinformation is part of my daily job: if my objective is to make the reader think something that I would like him to think, I can selectively extract specific sentences from a scientific study that have an higher chance to reinforce a belief or to satisfy a hope.
As a consequence, I (writer of a news website) have decided to report the sentences that states that the results “do not show robust statistical evidence” and that “out of every 100 online copyright infringements, 24 induce an extra legal transaction” but I also decided 1) to avoid to explain to the reader what that means/implies and 2) to avoid to cite the part of the study that states that the methodology didn’t provide reliable results, which is a very very important part of the story.
This behaviour is an extremely common example of how scientific results are usually manipulated so that they can become a tool to support a political agenda. Sometimes the subject that reports an information is in good faith and he/she just jumps to a conclusion because he/she is biased or because he/she didn’t actually read carefully the document or because he/she is not competent enough to interpret correctly the scientific results of a study. Other times the manipulation (of both the information and of the reader’s beliefs) is intentional. In any case, the final effect is the same: the reader blindly reports a false or misleading or wrong information somewhere else.