Re: js/init-defaultbranch-advice, was Re: What's cooking in git.git (Dec 2020, #01; Tue, 8)

From: Felipe Contreras <felipe.contreras@gmail.com>
Date: Dec 11, 2020, 00:03 UTC
Message-ID: <CAMP44s3Wp5HfmHB2UZFJyZSVv7xE9YAKVOWfDUCGOwHtQV6Gzw@mail.gmail.com>
In-Reply-To: <xmqqeejxwlx9.fsf@gitster.c.googlers.com>

On Thu, Dec 10, 2020 at 12:38 PM Junio C Hamano <gitster@pobox.com> wrote:

> Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:

> > Obviously, I have no idea how many users chose the first option because we
> > do not collect any usage data in the Git for Windows project. However,
> > anecdotal data suggests that most users go with the first one, not caring
> > one iota and simply going with what Git decides for them.

1. The first option says let git use "master". The user might have
stopped reading there.
2. Even if you had collected numbers, anyone with a minimal
understanding of statistics would tell you that you can't extrapolate
numbers from a subgroup of the population to the total population. You
need a random (non-biased) sample.
3. Anecdotal evidence is the weakest form of evidence.

So yeah, even if we suppose that most users (what is "most"? 51%?, 75%?, 90%?) of Git for Windows did choose the first option, we can't assume the rest of the users are going to choose that option, and for the same reason.

Something that very often happens in these types of questions is that the user simply did not understand the question, and just chose the first one.

Cheers.

-- 
Felipe Contreras