[Sigia-l] 800 or 1024 - Which min. display resolution to build for?
Paola Kathuria
paola at limov.com
Wed May 10 13:55:17 EDT 2006
John McCrory wrote:
> http://www.johnmccrory.com/miscfiles/visitor_viewport_data_sample.txt
I found this today when catching up on list mail and I
ended up procrastinating by playing with the data in a
spreadsheet.
I've prepared a summary of the top 20 combinations of screen
resolution and viewport size. They account for 91% of the
logged visits.
The summary shows viewport size as a percentage of the
screen size (e.g., a viewport of 400 x 300 on a 800 x 600
screen is 25%). The last column contains actual viewport
dimensions (the most popular) for the screen/viewport % combo.
========================================================================
Screen Viewport # visits % visits Example viewport (# visits)
-------------------------------------------------------------------------
1024 x 768 80% 1509 22% 1000 x 600 (517)
1024 x 768 70% 1250 18% 1000 x 570 (303)
800 x 600 70% 1177 17% 780 x 430 (540)
1024 x 768 60% 480 7% 800 x 600 (31)
800 x 600 60% 416 6% 780 x 380 (100)
1280 x 1024 80% 229 3% 1260 x 850 (34)
1024 x 768 40% 174 3% 770 x 430 (14)
1024 x 768 50% 151 2% 840 x 470 (7)
1280 x 800 80% 146 2% 1260 x 630 (41)
1280 x 800 70% 125 2% 1260 x 580 (29)
800 x 600 80% 122 2% 780 x 460 (83)
1152 x 864 80% 93 1%
1280 x 1024 70% 68 1%
800 x 600 50% 65 1%
1400 x 1050 80% 59 1%
1024 x 768 30% 44 1%
1280 x 1024 60% 44 1%
800 x 600 30% 40 1%
1280 x 1024 90% 39 1%
1280 x 800 60% 38 1%
-------------------------------------------------------------------------
(This summary is based on viewport percentages rounded to the
nearest 10% based on viewport dimensions rounded to the nearest
10 pixels.)
You can get the collated data that this summary is based on from
http://www.limov.com/share/viewport-percentages.csv (2200 rows)
It contains three columns:
- screen resolution
- viewport as % of screen
- # visits
Example row: "800 x 600",56,30
Explanation: 30 visits had a viewport 56% of 800x600 screen
Note that the logged raw data contains repeat visits from
visitors and I suspect around 30% is duplicate, which will
skew any analysis. This means you should take any stats
resulting from this data with a pinch of salt.
John, if you collect data again, I suggest you adjust the script
to only collect the screen/viewport stats once per visitor
(by visitor id in a cookie and/or session id). Also collect
IP address, date and time (H:M:S) to help eliminate duplicate
visitors from the stats.
Thanks very much for making your data and your script available.
Paola
More information about the Sigia-l
mailing list