[OSM-talk] Metrics

[Mateusz Konieczny]

Oct 17, 2021, 16:09 by andy at pigsonthewing.org.uk:

> On Sun, 17 Oct 2021 at 12:08, Mateusz Konieczny via talk
> <talk at openstreetmap.org> wrote:
>
>> Oct 17, 2021, 12:01 by andy at pigsonthewing.org.uk:
>>
>> > On Sat, 16 Oct 2021 at 06:20, Mateusz Konieczny via talk
>> <talk at openstreetmap.org> wrote:
>>
>> >> that is why quality of paid editing is usually problematic
>>
>> > Is it? Or do we only see low-quality paid editing brought here for discussion
>> (or to the Data Working Group for reversion), while good quality paid editing
>> goes unremarked?
>>
>> Maybe it depends on area but in Poland overall value of paid editing
>> is negative in my impression.
>>
>
> But I'm not asking about your impression - which you'd already stated
> as fact - but about the actual facts; or indeed, whether it is even
> possible to measure the actual facts, with the data available to us.
>
As you are well aware noone has data of such quality.

Unless you actually have data for Poland that measured

- quality of added data
- time needed to clean up wrong data
- mappers discouraged by low quality mapping by paid editors
- mappers encouraged by low quality mapping by paid editors
- mappers discouraged by presence of paid editors at all
- mappers motivated by presence of paid editors at all
- etc

As usual with sociology getting some useful, objective and not trivially gamed
metric is incredibly hard, and far more serious fields than OSM mapping
have failed to achieve them.

And it is typical that some effects that most missed is actually 
stronger that measured ones.

So it will be hard to get better that set of biased impressions.

Unless you can provide "actual facts" about quality of paid mapping?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20211017/83d2df00/attachment.htm>