Using Cost/Mbps/GB as the comparison criteria is possibly fair when comparing different broadband packages but is not that meaningful when evaluating the best option to select, and can be confusing to the general public. IMHO this comparison places too much emphasis on the speed component. For the average user this comparison implies that telkom's do3 option is almost 9 times better value than the do2 option which is not true.
I agree. Cost/Mbps is valid in an uncapped environment. Just dividing this again by the cap is not that meanigful.
Scoring Telkom 4Mbps/3Gb at 55.49 and Telkom 384/1Gb at 829.07 is completely misleading.
There is, of course, a benefit to having greater speed in a capped environment, but not nearly as much as in an uncapped environment. Double the speed means double the thruput when uncapped. But the thruput is the same when capped. You do however get the data marginally quicker.
The idea is that (Speed x Data) is the "benefit" you receive from your broadband connection. But while data is capped most people would value data much higher than speed.
If you can imagine a service that has infinte speed, but is still capped, it would score 0 in the article, regardless of the cap. That clearly makes no sense at all; it is the cap that would be all important.
Consider a graph that gives time on the x axis, and thruput (ie gb's downloaded) on the y-axis. The line would slant up to the right as time progressed. For an uncapped service, the line would contunue at this angle until the month is over. A capped service, on the other hand, would level off and run parallel with the x-axis once the cap is reached. One could consider the "benefit" of a broadband service to the be the area under the curve. A 4Mbps service with the same cap as a 384kbps service would give greater benefit, but not to the extent in the article.
How could this be measured? You need to determine an uncapped (max) thruput. Firstly it is unlikely that one would download 24/7 (some would, of course!). Or that one would normally be utilising 100% of the line speed. So if you assumed an optimum (uncapped) "full-speed" internet usage as say 1 hour per day. That is 30 hours a month. 1 800 minutes. 108 000 seconds. Divide this is by 8 (bits to bytes) x 1024 (bytes to kbytes) gives a constant of 13.1836. Multiply this by the line speed (in Mbps) to arrive at uncapped Gb usage. For 4Mbps this is 52.73Gb. For a 384kbps line this is 4.94Gb
The benefit (area under the curve) is ( 1 - cap/(2*max)) * cap.
For an uncapped service cap = max, and the benefit is cap/2. this is directly proportional to the line speed, and can be expressed as the average amount of data you have "received" over the month.
For capped services, things look very different. Some examples:
For a 4Mbps 3Gb cap: Benefit is ( 1 - 3/(2*52.73))*3 = 2.915
Cost is R665.90. Cost/Benefit ratio = 228.44
For a 384k 2Gb cap: Benefiit is (1-2/(2*4.94))*2 = 1.595
Cost is R369.90. Cost/Benefit ratio = 231.91
For a 384k 1Gb cap: Benefiit is (1-1/(2*4.94))*1 = 0.898
Cost is R310.90. Cost/Benefit ratio = 346.21
For 1.8Mpbs 1Gb cap: Benefit is (1 - 1/(2*23.73))*1 = 0.979
Cost is R289.00 Cost/Benefit ratio = 295.2
What is considered to be max/ideal/uncapped usage is debatable (and adjustable). I guess it depends on the Average Joe that the comparison is aimed at. But I think this kind of approach would give much more meanginful results, and would more closely reflect how broadband users evaluate the respective services.
I'm pretty sure that Telkom Do1 users do not consider themselves 15 times worse off than Telkom Do3 users.
Sign up with Google
)