Problem with 'choking' DSL as proposed at public meeting

Thought I'd move my comment from the Public Meeting thread to a separate topic, to allow the Public Meeting thread to focus on the Public Meeting

<font face="Verdana, Arial" size="2">
David went through the main points in this and explained in more detail some of the issues relating to ADSL particularly the concept of ?choking? where, rather than imposing capping limits, ISP?s could offer a reduced bandwidth option e.g. 64 or 128 Kbs at a lower cost.
</font>

This is an interesting point. The problem with the proposal, IMHO, is that, from the ISP's perspective, there's no economic reason to do this:

Short explanation:
When you buy 512k DSL you're actually only getting a 64k or so service anyway. All the 512k means is that when the ISP has spare bandwidth, you can use it, up to a maximum of 512k

Long explanation:
Large ISPs typically pay for bandwidth based on the amount needed at peak times. For example, at 8.00pm at night they may need 100mb of connectivity to satisfy all their customers. At 8.00am the next day though, they will still have (and pay for) that 100mb of connectivity, although nobody will be using it.

DSL typically takes advantage of this by overbooking - at an overbooking rate of 50:1, for example, at 8.00pm when every customer connects, they may only be able to get about 50-60k of that 100mb, but later on, when there's tonnes of bandwidth spare, they can use as much as they can take, up to 512k.

Thus, if if you buy a 512k package, you may only add about 50-60k to the ISPs peak time demand, which is about the same as the person who bought a 64k connection, even though they paid less for it.

So, essentially, with a 64k DSL service, the ISP will be asked to buy as much bandwidth as for a 512k user, but will only recieve about half as much in revenues from him or her.

So pushing this idea is probably a little pointless, IMHO

Congratulations to all, by the way, on an excellent Public Meeting. Looking forward to the seminar next month.