Playing with the Personal Stats

[Phlik]

I was interested to see the extent to which the area/time I was collecting were "new" . To look at this I plotted the top 1,000's (based on total WiFi seen) WiFi seen versus "New to Wigle".

If over 61% of WiFi seen are "new" to Wigle that user is on the high side of the regression line and more likely to be collecting in a "new" area/time period. (lots of caveats!...)



Plotting total WiFi seen versus total Cellular seen might tell something about the WiFi density versus cellular in the area / time of collection. Some more variation on this and accepting some only collect one or the other.



Plotting the Cellular discovered might provide another way at looking the uniqueness of your survey in the area/time of cllection. Again more variation.



This gets messy when you take account that people are collecting over a long period of time and different areas - I just wanted a high level and easy to produce view.


There's got to be some great insights to come out of using this rich data source?

[arkasha]

Cool visualizations!

It would be really neat to distill these down to simple metrics people could use to compare "newness" and add that figure to the stats page - I've often been surprised by my low percentage "new" points vs. overall seen I have vs. other stumblers at my ranking... I feel like there's a cool "moneyball" style stat here that would allow people to compare "freshness of data" trends over time!

cheers,

-ark