Monday, May 29, 2006

Very Deceptive Broadband Statistics

I have to thank DPF for posting about these on his blog (where I have also commented briefly on their issues).
The full report is available here.

For anyone interested in the state of broadband across the OECD countries it makes a good read - and has plently of pretty graphs for those who want a brief read. However be sure to check what the scale being used on each one is as the authors have done a good job of attempting to make New Zealand appear that it is performing badly on almost every graph despite them leading that statistic.

As a case in point: (New Zealand is highlighted by being the gray bar).

Thats right - this statistic is the cost of a new connection. So smaller is better... The authors are using the innate perception though that this is putting NZ at the bottom of this statistic.

The worst however is when they calculate their final conclusion. This is done by ranking on the following formula:
Business [Download + Upload + Monthly Cost + Connection] + Data cap = Business Ranking
Residential [Download + Upload + Monthly Cost + Connection] + Data cap = Residential Ranking
Business Ranking + Residential Ranking = Final Ranking

This means the Data Cap ranking is given a 50% weight in the final score. The authors have decided that monthly cost, connection cost, upload speed and download speed are all only worth a 12.5% weighting versus the data cap.
And despite the datacap ranking for NZ being last, and this 50% weighting, we still get 18th in residential (and 24th in buisiness).

Furthermore even if the datacap was the most important thing to be ranked on they have only ranked on initial size - not charge per GB after breach, further distorting the results. (I know I prefer my orcon plan that only charged $10 per 10GB over the cap to Australia's "For example, a limit of 200MB per month was not unusual
and neither was an excess charge in the region of A$100 per additional gigabyte."

This is appalling statistics at its worst - and I deplore the Dominion Post and DPF both who should be able to do a halfway critical analysis on this before reporting the result as "fact".
Tags: , , , , ,

Ads by

No comments: