Broadband penetration rates - percentages?

LoBo

Interesting reads - thanks for the transcripts.

"..that’s maths that you’re putting forward rather than spin.."
heh, obviously maths is a difficult proposition for them - where's the 5% coming from? 75000/4000000 = 1.875% ..

Ripwave

LoBo wrote:

Interesting reads - thanks for the transcripts.

"..that’s maths that you’re putting forward rather than spin.."
heh, obviously maths is a difficult proposition for them - where's the 5% coming from? 75000/4000000 = 1.875% ..

Look, if you're going to slag off McRedmonds maths, you should get your own right.

There are 4,000,000 people in Ireland, but there are only about 1,500,000 households in the country. Unless you want to argue that only one person in any household should be allowed to use DSL, and a household with 5 people should have to pay for 5 DSL connections, then you have to allow the 5% figure.

(Of course, McRedmonds 75,000 includes Business DSL connections, but as far as I can tell, that's trough for the figures from other countries too).

DonegalMan

Ripwave wrote:

Look, if you're going to slag off McRedmonds maths, you should get your own right.

Read EircomTribunal's post above where he corrected me on this.

Your maths are quite correct (as mine were) but if we are making comparisons with figures published for other countries, then we have to use the standard measure which is the amount of broadband connections per 100 population.

Ripwave

DonegalMan wrote:

Read EircomTribunal's post above where he corrected me on this.

Your maths are quite correct (as mine were) but if we are making comparisons with figures published for other countries, then we have to use the standard measure which is the amount of broadband connections per 100 population.

Yes, I've since verified that.

But I've also re-read the interview, and McRedmond never said 5% for "broadband connections per 100 inhabitants" (in fact, that's not generally expressed as a percentage figure). And the interviewer said "for example in Denmark, which has twelve point seven percent of the population having access to broadband". I don't know if Downes meant that there are 12.7 broadband subscriptions per 100 inhabitants in Denmark (700,000 DSL connections more or less) or not, but if he did, then his statistics are skewed too, because a "score" of 12.7 of the "BB subscriptions per 100 inhabitants" would actually imply that about 35% of the population have access to Broadband. (That's what the OECD says - referring to Koreas "score" of 21, it says that at 30 broadband connections per 100 inhabitants, 90% of the population would have access to Broadband).

So what I said still stands - if you want to refute McRedmond, there's no point in using a different standard than the one he used.

His figure of 75,000 DSL connections representing 5% of households stands (ignoring the substantial issue of residential versus business connections for the moment). Maybe that particular frame of reference happens to flatter eircom, but that, in and of itself, doesn't make it invalid.

Does anyone have up to date numbers for the actual number of broadband connections in other EU countries?

eircomtribunal

Ripwave wrote:

Yes, I've since verified that.

But I've also re-read the interview, and McRedmond never said 5% for "broadband connections per 100 inhabitants" (in fact, that's not generally expressed as a percentage figure). And the interviewer said "for example in Denmark, which has twelve point seven percent of the population having access to broadband". I don't know if Downes meant that there are 12.7 broadband subscriptions per 100 inhabitants in Denmark (700,000 DSL connections more or less) or not, but if he did, then his statistics are skewed too, because a "score" of 12.7 of the "BB subscriptions per 100 inhabitants" would actually imply that about 35% of the population have access to Broadband. (That's what the OECD says - referring to Koreas "score" of 21, it says that at 30 broadband connections per 100 inhabitants, 90% of the population would have access to Broadband).

So what I said still stands - if you want to refute McRedmond, there's no point in using a different standard than the one he used.

His figure of 75,000 DSL connections representing 5% of households stands (ignoring the substantial issue of residential versus business connections for the moment). Maybe that particular frame of reference happens to flatter eircom, but that, in and of itself, doesn't make it invalid.

Does anyone have up to date numbers for the actual number of broadband connections in other EU countries?

Ripwave,
if you listen again through the interview or read it, then can see that they were discussing on the basis of the Broadband penetration figures of connections per 100 persons.
They referred to the 0.1 % of Greece and to the 12.7 of Denmark, which are from the most up-to date official EU–15 stats for January 2004. They were not talking about household figures, and while they were not saying that after quoting each of the figures it is clearly implied.
McRedmond knows exactly that the Irish figure in this comparative context is 1.8% and not 5%!
Please have a look at "There is something Rotten in the State of Broadband Ireland", downloadable from the main page of www.comwreck.com in a small and a print version.
And yes, the Danes have 12.69% of bb connections per 100 persons and SKorea has 24%.

P.

Ripwave

LoBo wrote:

Some very good graphics in your .pdf on comwreck.com (direct link) - exactly what I mean by a fact sheet for journos who are interviewing Eircom- eg:
<image deleted for brevity - see above>
I've helpfully labelled luxembourg there too

If you're emailing Morning Ireland, why not add a link to the pdf (see above) for further info.

That's all very well, but McRedmond will simpley point out that 6 month old statistics are totally meaningless in this game, given the rate of increase in the number of connections (if the "635 we signed up yesterday" is actually true, you're looking at 13,000 a month, even on a 5 day week basis).

Now I'm not supporting McRedmond here, I'm simply playing devils advocate. The metric "BB subscriptions per 100 inhabitants" shouldn't be expressed as a percentage (because it isn't a percentage), and it's obviously confusing when it is used that way (as evidenced by Richard Downes reference to Denmark and 12.7% "having access to Broadband". McRedmond clearly used misleading statistics, but even if you are extremely generous and say that he doesn't understand them himself, you still end up with the problem that the metric being used is not intuitive to someone like Richard Downes (hell, it's not intuitive to someone like me - I thought that Koreas score of 24 means that a quarter of the population had access to Broadband, but it actually means that over 2/3rds of Koreans have access to Broadband).

A large part of the reason we're having these discussions is because of eircoms refusal to actually release usable statistics. But if there really are 13,000 new connections a month, then we could easily be at 200,000 connections by mid-2005. That's not quite the 240,000 that would represent a 6 BB subscriptions per 100 inhabitants (note NOT 6%) but it would be a pretty impressive position to be in if the target set by DCMNR is 100,000.

So the statistic that I'd like to see explained at this point is - was yesterday just an exceptional day? Does this 635 only include "sign ups" to eircom, or does it include all connections to the eircom infrastructure, including connections sold through resellers? And what's the actual monthly target?

jwt

Uggggg I'm slightly confused now

See if I have this right

(I'm simplifying this leaving out businesses etc.)
If there approx 2.5 inhabitants per house then if 1 house in a hundred had BB then the reality is 1 in 100 houses but actually the score is 1 in 250 inhabitants

Multiplying this up a score of 24 actualy means 24 by 2.5 is 60 percent of lines assuming one line per house
Put another way a score of 24 in 100 inhabitants is 24 bb lines per 40 (100/2.5) houses again assuming 1 line per house.

Yes? no? maybe?

So for very rough calculations we can assume in future that when someone says say 6 bb per 100 inhabitants what is really meant is 6 bb per 40 possible bb enabled lines or actually 15 percent!



John

eircomtribunal

jwt wrote:

Uggggg I'm slightly confused now

See if I have this right

(I'm simplifying this leaving out businesses etc.)
If there approx 2.5 inhabitants per house then if 1 house in a hundred had BB then the reality is 1 in 100 houses but actually the score is 1 in 250 inhabitants

Multiplying this up a score of 24 actualy means 24 by 2.5 is 60 percent of lines assuming one line per house
Put another way a score of 24 in 100 inhabitants is 24 bb lines per 40 (100/2.5) houses again assuming 1 line per house.

Yes? no? maybe?

So for very rough calculations we can assume in future that when someone says say 6 bb per 100 inhabitants what is really meant is 6 bb per 40 possible bb enabled lines or actually 15 percent!

John

The official stats used as the primary figures by the EU and the OECD are simple and clear.
12.7% bb penetration for Denmark means: per 100 Danes there are 12.7 broadband connections.
It is a simple and clear percentage figure and I can't understand Ripwave's post questioning the validity of using %.

The beauty with this "Broadband penetration figure" is that it is clear and measurable and measured in this way by the other countries.

Other definitions, like percentage of households with broadband connections are sometimes used, but are not at all as clear-cut. Reason: You could simply multiply your bb penetration figure by the person per household factor (3 in Ireland) to get roughly the "households with bb percentage", but in reality you'd have first to subtract the business bb figures. The Swedish regulator, if I remember rightly, gives such detailed data. But it is much better to compare with data that is available in the same definition across the Eu or OECD.

P

Ripwave

jwt wrote:

Uggggg I'm slightly confused now

See if I have this right

(I'm simplifying this leaving out businesses etc.)
If there approx 2.5 inhabitants per house then if 1 house in a hundred had BB then the reality is 1 in 100 houses but actually the score is 1 in 250 inhabitants

Sort of - if 1 house in 100 has ADSL, and there are 2.5 people per house the "score" is actually "1 BB subscription per 250 inhabitants", or "0.4 BB subsriptions per 100 inhabitants".

The notion of using a standard metric is admirable in itself, but the actual metric used is a bit confusion because it sort of compares apples to oranges.

jwt wrote:

Multiplying this up a score of 24 actualy means 24 by 2.5 is 60 percent of lines assuming one line per house
Put another way a score of 24 in 100 inhabitants is 24 bb lines per 40 (100/2.5) houses again assuming 1 line per house.

Yes? no? maybe?

So for very rough calculations we can assume in future that when someone says say 6 bb per 100 inhabitants what is really meant is 6 bb per 40 possible bb enabled lines or actually 15 percent!

Yes. Sort of. If you ignore the fact that some of the DSL enabled lines are actually used by businesses, and aren't going to residences, so the numbers don't actually mean what they seem to mean, because "business lines" will mean different things in different countries. The US has a very strong entrepreneurial ethos, and there are a lot of home based businesses. And there are significant numbers of "home offices" for people who work from home a couple of days a week. This is probably not as common in many European countries, and I have no idea how common it is in Korea). But these particular "business lines" are often used by the rest of the family out of business hours for "residential" use, which isn't the case for the small business in Ireland that uses DSL in the office, to replace the ISDN and dialup access that they were stuck with until very recently.

DonegalMan

Ripwave wrote:

The metric "BB subscriptions per 100 inhabitants" shouldn't be expressed as a percentage

That's the OECD way of doing it and they're the people that Comreg quote from

If Eircom want to play with figures, let them, we'll hit them on it, but if we go away from OECD then we'll have to explain it every time we make comparisons plus give Eircom a tool to hit us with, I can hear it now - ".. oh that IOFFL crowd, they measure things different from everyone else so their figures are meaningless ... "

(Plus EircomTribunal would have to redo his pretty charts every time the figures are released )

Ripwave

eircomtribunal wrote:

The official stats used as the primary figures by the EU and the OECD are simple and clear.
12.7% bb penetration for Denmark means: per 100 Danes there are 12.7 broadband connections.
It is a simple and clear percentage figure and I can't understand Ripwave's post questioning the validity of using %.

I'm pretty certain that the OECD does NOT express that as 12.7% bb penetration. There are no references to percentages on this table. for example.
http://www.oecd.org/document/33/0,2340,en_2649_201185_19503969_1_1_1_1,00.html

12.7 BB subscription per 100 inhabitants would imply a "broadband penetration" closer to 35%, assuming that most of those subscriptions were residential subscriptions. Otherwise you have the ridiculuous situation where you could have universal BB access with only 40% "broadband penetratraion", because there are far more people than there are phonelines, so it's statistically impossible to get 100 subscriptions per 100 inhabitants.

(And yes, I know that some media reports use this erroneous "broadband
penetration" label, but it is far from universal, and is clearlt misleading).

By the way, this ITU document is where I got the reference to "At a level of 30 per 100 inhabitants, more than 90 percent of households would have broadband":
http://www.itu.int/ITU-D/ict/papers/2004/ITU%20Statistical%20Activities%20by%20SP.pdf

Ripwave

DonegalMan wrote:

That's the OECD way of doing it and they're the people that Comreg quote from

Is it - where do the OECD express it as a percentage?

(I've looked - I might well have missed it, because there's a limit to the number of documents that you can read, but none of the obvious ones expressed this "BB subscriptions per 100 inhabitants" as a percentage, because a percentage implies that there is such a thing as 100%, and that clearly isn't true of this metric).

eircomtribunal

Ripwave wrote:

I'm pretty certain that the OECD does NOT express that as 12.7% bb penetration. There are no references to percentages on this table. for example.
http://www.oecd.org/document/33/0,2340,en_2649_201185_19503969_1_1_1_1,00.html

Ripwave,

What are you at? This is leading us astray.

per 100 = percent. It is not apples an pears. Each of the population could order one broadband connection, in theory.
On top of the OECD chart it says broadband connections per hundred inhabitants. That is the same as percent to me.

P.

P.S. And in the EU chart (in "There is something Rotten..."), which I got from the Danish regulator they use the % sign. I didn't include it.

jwt

Can our one and only mod seperate this out to a new thread please?

John

DonegalMan

Ripwave wrote:

Is it - where do the OECD express it as a percentage?

There seems to be two arguments here, let's try and put them to bed for once and for all

1. What are OECD actually measuring?
   They are clearly measuring subscribers per 100 population. They don't actually define that but if you go back to their very first report in Oct 2001 (pdf downloadable here) they state it indirectly in a number of places:

   P5: "At the beginning of 2001 just one person per 100 inhabitants, on average in OECD countries, was a subscriber to high speed Internet access"
   
   P12 Footnote: "This indicator shows the total number of households which are connected to cable"
   
   P13 Footnote: "DSL connections for Australia are an estimate based on Telstra's reported broadband subscribers and industry estimates for overall broadband connection."

   When you think about it, it makes sense, they know the number of subscribers, they know the population but they don't have any conversion figure to get an accurate total people with access, they would have to estimate it.
   
   
2. Is it right to quote this figure as a Percentage?
   The mathemetician in me says no, the pragmatist in me accepts that that is what is in everyday usage. OECD certainly refer to it as "Broadband Penetration" and I would stick with their definitions for the reasons I gave earlier.

Note to JWT - I endorse what you are saying about moving it elsewhere, it's a becoming a pedantic/distracting argument and this is my last input on it

sceptre

Split from the Morning Ireland thread.

Ripwave

eircomtribunal wrote:

What are you at? This is leading us astray.

Where it's getting at is that if you try to play eircom at their own game by redefining what the statistics actually say, then eircom will win.

eircomtribunal wrote:

per 100 = percent. It is not apples an pears. Each of the population could order one broadband connection, in theory.

So we won't have "100% broadband penetration" in Ireland until there are 4 million seperate and individual DSL, wireless and cable internet connections in place? Any family of 2 adults and 4 kids, for example, would have to pay for 6 seperate DSL connections for it to count as "100% broadband penetration"?

Frankly, that's ridiculuous, and that precisely the sort of statistical tomfoolery that you're all accusing eircom of being guilty of.

There are about on the order of 2 million phone lines in the country. If every one of them was DSL enabled, you'd only have "50 broadband subscriptions per 100 inhabitants".

Now is that 50% "broadband penetration"? Even though it would include about 90% of all households in the country? What would 100% Broadband penetration be?

eircomtribunal wrote:

On top of the OECD chart it says broadband connections per hundred inhabitants. That is the same as percent to me.

It's not percent, because a single subscription can be shared between multiple inhabitants - so you can have 100% broadband access without having 1 bradband subscription petr person.

If it was per cent, then the OECD would say per cent - they sure as hell aren't using the clunky construction "BB subscriptions per 100 inhabitants" for fun.

eircomtribunal wrote:

P.S. And in the EU chart (in "There is something Rotten..."), which I got from the Danish regulator they use the % sign. I didn't include it.

Well then the Danish regulator doesn't understand percentages either.

Ripwave

DonegalMan wrote:

There seems to be two arguments here, let's try and put them to bed for once and for all

1. What are OECD actually measuring?
   
   They are clearly measuring subscribers per 100 population. They don't actually define that

What's to define, Martin? The description "Broadband Subscriptions per 100 Inhabitants" is a very clear definition, and I don't see anyone arguing over what it means. For the sake of simplicity, we can, for now at least, count it as DSL connections per 100 inhabitants (in Ireland), though that is slowly changing. As a metric for comparing countries, it's a perfectly reasonably measure. But it's not actually useful for understanding where we stand, because it's not necessarily obvious to peope that a "score" of 30 on that measure effectively means unversal broadband access.

DonegalMan wrote:

[*]Is it right to quote this figure as a Percentage?
The mathemetician in me says no, the pragmatist in me accepts that that is what is in everyday usage. OECD certainly refer to it as "Broadband Penetration" and I would stick with their definitions for the reasons I gave earlier.

, No, Martin, the OECD do not quote this as a percentage. Some people, who obviously don't understand it, are misquoting it as a percentage (including the Danish regulator, it appears). But there is nothing pragmatic about using bogus statistics, especially when it plays into eircoms hands.

When eircom comes out next summer and claims that there are 120,000 residential DSL connections, and that this means that 10% of the population have access to broadband at home, the ordinary man in the street can pretty much recognize that this figure is reasonably accurate - he doesn't need any formal training in statistics to do this. If the Sunday Business Post and Morning Ireland and Ireland Offline have been busy publicising misleading statistics that imply that Denmark has 12.7% "broadband penetration", then eircom looks pretty good. This is an "apples to oranges" comparison, because 40% of the Danish population has broadband access at home, not 12.7%, but the average man in the street isn't going to have a intrinsic grasp of that.

Result? eircom 1, the real worl 0.

And all because you think it's pragmatic to lable something as a percentage, when it clearly isn't (because you don't need 100 BB Subscriptions per 100 inhabitants to have universal BB access).

jwt

So if Eircom claim 120,000 bb is 10% (not sure where that comes from*) connections in a population of 4,000,000 it actually means
3 bb subscribers per 100 inhabitants. (120,000 / 4,000,000 = 0.03 0.03*100 inhabitants = 3)


So in that scenario Denmark is a factor of four ahead of us not a couple of points.

*
I assume were working on 1,200,000 lines in ireland?

John

P.S. Puting in the obvious maths in above is more for my benefit than casting aspersions on posters mathematical abilities

vinnyfitz

I thought Peter's correction of eircom, read out on Morning Ireland this morning, was quite helpful to the debate.

maxheadroom

vinnyfitz wrote:

I thought Peter's correction of eircom, read out on Morning Ireland this morning, was quite helpful to the debate.

Any chance of a transcript?

Ripwave

jwt wrote:

So if Eircom claim 120,000 bb is 10% (not sure where that comes from*) connections in a population of 4,000,000 it actually means
3 bb subscribers per 100 inhabitants.

There are about 1,200,000 houshoulds in Ireland - I think most people would accept that if there's a BB connection in a household, it would be fair to say that everyone in the house has access to Broadband? (obviously, there will be exceptions). So if 10% of housholds have a broadband connection, is that 10% Broadband Penetration, or is it 3% Broadband Penetration?

(120,000 DSL connections would give a "score" of 3 on the "BB Subscriptions per 100 inhabitants" table).

Just answer the simple question - how many BB subscriptions does a country with a population of 4 million people need top have 100% Broadband Penetration?

Is anyone seriously claiming that you need 4 million Broadband subscriptions to achieve 100% Broadband Penetration?

jwt

Ripwave wrote:

Just answer the simple question - how many BB subscriptions does a country with a population of 4 million people need top have 100% Broadband Penetration?

Ah...Ripwave I'm not disagreeing with you, far from it. Just making sure that if I'm asked I have it dead straight in my own mind.

Is what I said in my last post right or wrong? Its an honest question, I don't want to go off half cocked with this to some reporter if he/she should ask.

If what I said is broadly right if simplistic fair enough. If it's wrong show me where. All I'm doing with the above post is putting my understanding of the situation forward.


John

eircomtribunal

jwt wrote:

So if Eircom claim 120,000 bb is 10% (not sure where that comes from*) connections in a population of 4,000,000 it actually means
3 bb subscribers per 100 inhabitants. (120,000 / 4,000,000 = 0.03 0.03*100 inhabitants = 3)


So in that scenario Denmark is a factor of four ahead of us not a couple of points.

*
I assume were working on 1,200,000 lines in ireland?

John

P.S. Puting in the obvious maths in above is more for my benefit than casting aspersions on posters mathematical abilities

"Broadband penetration" is the the word we need to define whenever we use the word. It is used as the description for different things by different people internationally.

1.Broadband penetration, used by the OECD, the EU and others as the prime figure to compare the broadband status of countries is defined as:

number of broadband subscribers per 100 inhabitants or
percentage of broadband subscribers of population

That is the set of figures discussed on Morning Ireland. Denmark (January 2004) had 12.7 bb subscribers per 100 Danes (or 12.7% of Danes had a broadband subscription), Greece had 0.1 bb subscribers per 100 Greeks, and Ireland had 0.9 bb subscribers per 100 inhabitants. Currently Ireland has 75 000 bb subscribers out of 4 000 000 inhabitants, which equals 1.875 bb subscribers per 100 inhabitants(or 1.875% of the Irish have a bb subscription) – and not 5 (%), as McRedmond had claimed (and he specifically referred to that context: "Well, actually, where we are in that table is we are already now at five percent."

This set of measuring has two advantages: it is universally easily measured (data are available) which is very important for comparisons and it is as good an indication of the broadband status of a country as any other measure.

[Note about the % issue: Showing the above as %, as in the EU and Danish graph of my document "There is something Rotten in the State of Broadband Ireland", is correct if the definition of what is measured is given correctly. The EU and Danish percentage table and graph show us how many people of Greece, Ireland etc. have taken out a broadband subscription. Currently 1.875 out of 100 Irish people, or 1.875 % have taken out a bb subscription, in Jan 2004 12.69% of Danes, or 12.69 Danes of 100 Danes had taken a bb subscription.
Ripwave, you may find this set of figures not the most apt, as it does not directly say what percentage of a population is actually "using" broadband,(it's a bit like newspapers publishing a figure of copies printed, another figure of copies sold and another one of number of readers) but their is nothing wrong with the usage of %. 12% percent is nothing but shorthand for the fraction 12/100; important is that you use the same unit in top and bottom of the fraction: 12 Danes/100 Danes have a broadband subscription. As the unit is the same it cancels out and you arrive at 12/100 or 12%.
I thank you for your insistence as it forced me to look closer into this and will be extra careful with exact definitions in the future. It is always a tight line between being over-exact to the point of becoming unreadable and not being detailed enough.]


2.Broadband penetration, also used by the OECD, the EU and others as a figure to describe the broadband access of households.

percentage of households with broadband access

This percentage makes "more sense" or is more "real". As a rule of thumb you can multiply the figure of "bb subscriptions per 100 inhabitants", see above, with 3 (three persons in one household) to arrive at it.
I and many others don't like to use this figure as the prime figure for a few reasons:
It is not really measured as such in the same way by different countries, thus not the best basis for comparison. It is not as easily measured: you have to subtract the number of bb subscriptions of businesses from the overall figure, to arrive at a clean result. The Swedish Regulator has exact figures for Sweden in this regard.
But there is nothing wrong to compare with other countries in this regard, as long as the other figures are in the same category.
McRedmond on Morning Ireland of course compared the alleged Irish figure for bb household penetration of 5% (that would be assuming that of the current 75 000 bb connections some 46 223 are for households [46 223 equals 5% of the 924 464 "family units in private households"CSO] and 28 776 are taken by businesses.) with the figures for bb subscriptions per 100 inhabitants of other EU countries. Household penetration figures for Europe will arrive at around 20% by end 2004. See for example this


3.Broadband penetration, in the US is often defined as:

percentage of Internet user that use broadband

Only recently the US media said, "broadband penetration now over 50%", meaning, over 50% of Internet user now do so by broadband. Again, nothing wrong with it, as long as the definition is clear.


4.There are other Broadband penetration definitions, like

DSL subscriptions per 100 phone lines,

"The penetration rate in the country is three DSL subscriptions per 100 phone lines..." from this document about the bb development in the East Europenanhttp://www.polishmarket.com/next.php?id=6634
Again as long as we compare like with like, no problem, but for practical reasons the most widely available figures are preferable. The whole thing is confusing enough.

John, as you mention being prepared for answers etc,
Availability:
We've, according to the latest detailed CSO (2002, I think) figures:
924 464 family units in private households;
541 520 are in towns of 1500 of more, equals 58.57% of all households;
382 944 are in rural areas.

Any claims by McRedmond and others, that we have "over 70%" or "84%
availability by March 2005" as TIFF/IBEC's Tommy McCabe claimed in his recent silisconrepulblic interview (figures were supplied to him by Eircom), are outrageous and a deliberate misinformation of the public and should be refuted wherever they crop up. I do my best to do that and the publication of media reports etc on boards is of the greatest value.

So, even if all towns of 1500 and greater were dsl enabled and we had no bb line test failures, dsl availability would be under 60%. When you calculate a line failure rate of 20% (McRedmond) into the equation, we arrived at a dsl availability of some 46.856. With a more likely failure rate of 25% we'd a dsl population coverage of 44%.
End of 2003 the OECD average was a 75% broadband coverage.
Eircom have picked the raisins and dsl enabled the biggest 200 of their 1000 exhanges and we don't see plans published to enable the rest of them.

P.

Ripwave

jwt wrote:

Is what I said in my last post right or wrong? Its an honest question, I don't want to go off half cocked with this to some reporter if he/she should ask.

I thought I did answer your question.

You made a couple of statements in your last posrt ("eircoms 10% means 3 BB subscriptions per 100 inhabitants" and "Denmark is a factor of four ahead of us" both of which logically follow from what has already been discussed). You also asked the question "I assume were working on 1,200,000 lines in ireland?" and I answered that we're talking about (very) roughly 1,200,000 households. (There's closer to 2 million lines).

I'm not trying to be obtuse about this. You can have "100% Broadband Penetration" with only 30 or 40 "BB subscriptions per 100 inhabitants". If you are confused, it's because some people insist on misquoting the OECD.

Ripwave

eircomtribunal wrote:

"Broadband penetration" is the the word we need to define whenever we use the word. It is used as the description for different things by different people internationally.

1.Broadband penetration, used by the OECD, the EU and others as the prime figure to compare the broadband status of countries is defined as:

number of broadband subscribers per 100 inhabitants or
percentage of broadband subscribers of population

No, Peter. That's not the OECD or the EUs definition. That's your redefinition (or your copying the Danish regulators redefinition).

The whole point of a percentage is that 100% means 100%. And it's clear that you do not need 1 broadband connection for every inhabitant to achieve 100% Broadband Penetration.

That is the set of figures discussed on Morning Ireland. Denmark (January 2004) had 12.7 bb subscribers per 100 Danes (or 12.7% of Danes had a broadband subscription),

There you go again, Peter. "12.7% of Danes had a broadband subscription" is a totally meaningless statistic, because a "broadband subscription" is a "shared resource" - it is not restricted to that single subscriber. The correct metric is, exactly as the OECD states it, "12.7 BB subscriptions per 100 inhabitants".

Greece had 0.1 bb subscribers per 100 Greeks, and Ireland had 0.9 bb subscribers per 100 inhabitants.

The OECD metric is perfectly adequate for comparing 12.7 to 0.1 or 0.9, and doesn't need any spruious "percentages" tacked on. It is not a percentage, because that implies that the "target" is 100%. And Eircoms "5%" is "5% of 100%", which is precisely what percentage means.

Currently Ireland has 75 000 bb subscribers out of 4 000 000 inhabitants, which equals 1.875 bb subscribers per 100 inhabitants(or 1.875% of the Irish have a bb subscription)

You see, even your own numbers come back to bite you, because we don't have 75,000 bb subscribers - we have 75,000 BB (DSL) subscriptions. But a significant number of these are for businesses, and don't represent "subscribers" as a portion of the population. If you want to compare that to the population as a percentage, then you have to allow the term "subscriber" to cover everyone that has access to a "subscription", because otherwise you aren't measuring on a metric that reaches as far as 100%.

– and not 5 (%), as McRedmond had claimed (and he specifically referred to that context: "Well, actually, where we are in that table is we are already now at five percent."

Yes, he did say that. And, apart from the fact that he works for Eircom, I don't have any evidence that he wasn't genuinely confused by the misleading us of the term "percentage" to describe the number of BB subscriptions per 100 population.

This set of measuring has two advantages: it is universally easily measured (data are available) which is very important for comparisons and it is as good an indication of the broadband status of a country as any other measure.

I have consistently said that it's a perfectly reasonable metric. That doesn't make them a percentage.

[Note about the % issue: Showing the above as %, as in the EU and Danish graph of my document "There is something Rotten in the State of Broadband Ireland", is correct if the definition of what is measured is given correctly. The EU and Danish percentage table and graph show us how many people of Greece, Ireland etc. have taken out a broadband subscription. Currently 1.875 out of 100 Irish people, or 1.875 % have taken out a bb subscription, in Jan 2004 12.69% of Danes, or 12.69 Danes of 100 Danes had taken a bb subscription.

Peter, that's a post-hoc arguement if ever I saw one. The Danish graph is mislabeled, plain and simple. Redefining the label after the fact doesn't change the fact that it is wrong. The chart is labeled "Broadband penetration rate in the EU (% of population)". I'm perfectly happy to accept that Denmark had a "score" of 12.7 in a measure of "Broadband Penetration", but it is not a score of 12.7%, because that would imply that the "top score" would be 100%. To suggest that Denmark needs 8 times as many Broadband subscriptions as it has today to reach 100% penetration is, frankly, ludicrous.

Ripwave, you may find this set of figures not the most apt, as it does not directly say what percentage of a population is actually "using" broadband,(it's a bit like newspapers publishing a figure of copies printed, another figure of copies sold and another one of number of readers) but their is nothing wrong with the usage of %. 12% percent is nothing but shorthand for the fraction 12/100; important is that you use the same unit in top and bottom of the fraction: 12 Danes/100 Danes have a broadband subscription. As the unit is the same it cancels out and you arrive at 12/100 or 12%.

Peter, I know what a percentage is. And if you want to say that 12% of Danes have a Broadband subscription, go right ahead. (I have no problem with the the use of "this set of figures" as the basis for comparing different countries, as I have said again and again in this thread) But don't say that "Broadband Penetration in Denmark is 12%", because that's not the same thing, unless you state that Denmark won't have 100% Broadband penetration until every single Dane has their own personal BB connection. And if you want to accuse Eircom of using misleading statistics, then you need to be very careful about not using misleading statistics yourself. 12.7 isn't misleading, 12.7% is.

[snip all the rest about different measures, which aren't relevant to the question at hand, and about which I have no particular argument]

dahamsta

There you go again, Peter. "12.7% of Danes had a broadband subscription" is a totally meaningless statistic, because a "broadband subscription" is a "shared resource" - it is not restricted to that single subscriber. The correct metric is, exactly as the OECD states it, "12.7 BB subscriptions per 100 inhabitants".

I'm sorry, I'm lost. This makes zero sense to me. Could you please explain the difference between a single subscriber and an inhabitant? If I'm to take it that you mean that the subscription is shared between several inhabitants in one habitation, surely that would simply mean that the Danish metric is absolute nonsense? Is that what you're saying?

adam

eircomtribunal

Ripwave,

would you look to the left top description on the OECD graph - Lobo posted it on some thread I cannot find right now - it's on page three of my pdf "There is something Rotten..". it says:

"Broadband Subscribers per 100 inhabitants"

Note: It does not say "broadband subscriptions" . I took that doc straight from the "Information Society" and perhaps they are as stupid as the Danish regulator, or even the OECD don't know what they are talking about...

You may not like that this is called "Broadband Penetration", but there is nothing wrong with expressing the number of people who subscribe to broadband, or the Irish Times etc as a percentage of all the people.



Look at it as well in a pragmatic sense, this is the type of figures that will be used: Demot Ahern his MorningIreland appearance also talked of 3% Irish Broadband penetration (the figure should have been 1.8% of course, as it's 75 000 who have a broadband subscription out of 4 000 000 people)

P.

Ripwave

dahamsta wrote:

I'm sorry, I'm lost. This makes zero sense to me. Could you please explain the difference between a single subscriber and an inhabitant?

For a start, any business that has a DSL line is a subscriber, but isn't an inhabitant.

That doesn't really matter when you just want to compare two different countries, and be able to figure out whether 2 million DSL connections in a country of 27 million people is better or worse than 350,000 DSL connections in a country of 5.4 million. The metric "BB Subscriptions per 100 inhabitants" makes that comparison easy (7.4 is better than 6.4).

If I'm to take it that you mean that the subscription is shared between several inhabitants in one habitation, surely that would simply mean that the Danish metric is absolute nonsense? Is that what you're saying?

I'm saying that it's absolute nonsense to say that Denmark has a "Broadband Penetration rate of 12.7%". That implies that Denmark needs to install 8 times as many BB connections to reach a "penetration rate" of 100%.

The 12.7 number isn't the problem - I have no problem with the claim that Denmark has a "Broadband Penetration Rate" of 12.7 compared to Irelands 1.8 or 2.5 or 3, or whatever the number of the day is. That's a perfectly valid way to compare the two countries. But to express it as a percentage of the population, when you don't need "one for every member of the audience" is simpley wrong.

Ripwave

eircomtribunal wrote:

Ripwave,

would you look to the left top description on the OECD graph - Lobo posted it on some thread I cannot find right now - it's on page three of my pdf "There is something Rotten..". it says:

"Broadband Subscribers per 100 inhabitants"

Peter, make up your mind. You're the one who is insisting that Irelands 75,000 DSL connections represent 75,000 "subscribers", giving Ireland a "score" of 1.875%.

Unless you want to make the case that the OECD is only counting BB connections that are paid for by individual people, and totally discounting BB connections to businesses, then I don't think it really matters whether you use "subscriber" or "subscription". And I believe you already presented the case against trying to seperate residential and business connections out, when you're trying to compare performance between countries).

And once again, the OECD does NOT state this metric as a %. That's not an accident. They don't state it as a %age, because it isn't a %age.

Ripwave

eircomtribunal wrote:

You may not like that this is called "Broadband Penetration", but there is nothing wrong with expressing the number of people who subscribe to broadband, or the Irish Times etc as a percentage of all the people.

Peter, I'm going to quote my own post, just for your benefit.

RipWave wrote:

And if you want to say that 12% of Danes have a Broadband subscription, go right ahead.

You are expressing a measure of the number of individual Danes who have subscribed to Broadband. But it isn't a measure of "Broadband Penetration", because you do not need to be personally responsible for a "Broadband Subscription" to benefit from BB.

I'll ask you again - how many BB subscriptions (or subscribers - whatever you're having yourself) does a country with a population of 4 million people need to have 100% Broadband Penetration?