[Talk-us] MapRoulette new challenge: connectivity

Martijn van Exel m at rtijn.org
Mon Oct 29 21:03:56 GMT 2012


Hi Frederik,

On Mon, Oct 29, 2012 at 2:33 PM, Frederik Ramm <frederik at remote.org> wrote:
[..]
> If I may add a word auf caution.
>
> I think that tools like this are a nice way to bring a bit if fun to
> otherwise rather tedious mapping tasks, and I have often talked of the
> "gamification" of such things myself.
>
> Recently, there was a rather unfortunate incident involving such a tool in
> Thailand, where a contributor unfamiliar with the local language, community,
> or customs "fixed" a large number of "bugs" that a roulette-like tool had
> sent him to. Mistakes were made (because part of the "play" aspect was also
> to get a good "score" and be quick), an email standoff between a local
> mapper and the remote "fixer" ensued, and the local guy reverted the edits,
> DWG was drawn into the issue and didn't make any friends on either side
> either (of course everyone believed that they were doing the right thing)
> and now it looks like we might be losing a long-time community member with
> thousands of good, sweat-of-the-brow, old-style survey local edits.

You make a good point, and we have discussed this before as well.
There is a thin line between providing incentives and ways to get
tedious work done more quickly and efficiently, and real
'gamification' where there is a real risk of quantity over quality.
This is why MapRoulette does not do any of these things - there is no
ranking, there is no badges or prizes, exactly because I don't want
folks to go fix lots of things quickly just for the sake of it.

That said, we did have success with the Baseball challenge - at least
I did not hear anyone complain about bad baseball fields added after
that was done. So I think there is room for a little more game
incentives in this kind of tool, but I definitely don't want to rush
to achieve that, especially because it seems to work just fine as is.

> What do I want to say by that: Playful armchair bugfixing is nice and may
> often be helpful, but we must not forget that local knowledge trumps
> anything you can fix in a roulette. If one road ends near another road, that
> might actually be for a reason, and what looks like a shadow on the aerial
> image is in fact a fence - or the aerial image is outdated...

By exposing these bugs this way, you put the responsibility in the
hands of the armchair mapper, and that is a leap of faith I am willing
to make considering the situation here in the US. We don't have the
luxury of 5 mappers per square km here and we will never have that -
not only because we don't have the population density, but also
because of the vast stretches of land that do have road coverage but
almost no people, and it is very unlikely for someone to take
ownership / stewardship of those stretches. But we still want the bugs
fixed, so I am happy to resort to the second best solution, which is
to carefully direct armchair mappers to the bugs and leverage the
crowd to fix them.

For the same reason (and other ones that are irrelevant here), I
restricted the scope of MapRoulette to the US initially. I think that
a tool like this makes much less sense for a mapper-dense and
population-dense country. I am looking at integrating with KeepRight,
but we will need to have this discussion before that can happen.
>
> So I would appeal to every writer of "roulette" type tools: Educate your
> users; don't put too much "fixing pressure" on them (give them a chance to
> back out and tell them that when in doubt, don't fix something). Always make
> sure that you don't air-drop your users into countries they're not familiar
> with or even worse, where they don't speak the language. When in doubt, let
> someone draw a rectangle on the map first and say "this is my area of
> interest" or so. (I know this doesn't apply to your particular tool,
> Martijn, since the US is probably coherent enough for someone from Florida
> to make a good edit in Colorado... or is it?)
>

There is already a SKIP button in MapRoulette for that very reason.
What I will also do is add another word of warning to MapRoulette to
please skip stuff if you're not sure how to fix it or you can't
appreciate the situation on the ground from aerials. And yes, I do
think the US is predictable enough in its road layout that I would
trust someone from say Florida to clean up the beautiful state of
Utah.

-- 
martijn van exel
http://oegeo.wordpress.com



More information about the Talk-us mailing list