PRs made against own repos and repos in orgs you belong to should not count

[GBH]

The whole point is to "gift" PRs to other open-source projects, right?

Just look at this: https://24pullrequests.com/mergers and https://24pullrequests.com/users Try to find a person who actually committed to a project that they don't own. Gonna take you a few tries.

Also there are repos that are clearly not software projects :/

There needs to be a better gamification of this event. With clear rules and a fair leaderboard of sorts.

[adamyeats]

I seem to remember that we changed this last year as maintainers weren't properly being included (see #1734 and #1611), but if there is a way to tell project maintenance (e.g. pull request merges, documentation etc.) apart from just arbitrary commits to a repository, then I'd be interested in reviewing it.

Either way, in my view, further gamification of 24pullrequests is out of scope for the project; the point is the spirit of contribution rather than "winning" anything specifically. Although it's not my project, I'm keen to hear what @andrew and others think!

[andrew]

Agree with @xadamy, you're only gaming yourself as we don't have any prizes, we also randomize the top people/orgs on the homepage for that same reason

[GBH]

Of course it's not about "winning". But what drives participation and contributions? Is it really just the calendar thingy that's associated with the user?

Can't really tell if anybody actually participates in the event as the list of maintainers/developers (random or not) does not indicate anything. What's the incentive to participate? Right now it seems to be this list: https://24pullrequests.com/users (200PRs against own repo)

Funneling devs into projects for this event isn't great either: #2109 There are only 6 "featured" projects and 3 of them belong to 24PR org's members.

Having fun stats like:

• number of PRs per project users contributed to
• users per number of PRs (excluding own of course)
• list of projects a user committed to

"Gaming" participation drives more participation and improves project discovery. I just think current site can be improved a bit.

[alcaeus]

I agree with @andrew to the point where you're only gaming yourself, but I would also like to mention that the leaderboard suggests otherwise: it suggests that the goal isn't to give back or contribute to open-source projects, but rather to make as many contributions as possible.

As for "projects you own", this is difficult to handle. For example, I'm a maintainer in the Doctrine project. Does this mean that my contributions to the Doctrine projects during this month shouldn't count against my total? Are those worth less just because I contribute regularly?

On the other hand, we have cases where people get to the top of the leaderboard by obviously splitting commits (see https://github.com/changeworld/hackerrank/pulls?utf8=%E2%9C%93&q=is%3Apr+is%3Aclosed+created%3A%3E2017-12-01, where the author of these pull requests currently holds the number 2 spot in the leaderboard and each pull request only modifies a single file). I'm not saying they are doing this for the purpose of getting to the top of the leaderboard, but it does look like that. It also somewhat devaluates other people's contributions.

Maybe the system could flag pull requests to the same repository with similar titles that are created within a certain time: this would help not count automated commits as well as commits that were obviously split. It's not perfect but maybe food for thought?