Help us fact-check iTNews’ NBN figures

45

blog This morning some of you may have noticed that technology media outlet iTNews published a set of somewhat sensationalist claims regarding the National Broadband Network’s rollout progress. Under the headline “NBN Co fudges numbers to expand rollout”, the normally sober outlet alleged, among other things, that NBN Co had deliberately given a misleading impression of its rollout progress. Probably the key paragraph:

“NBN Co has continually promoted the impression of progress by announcing that its fibre rollout has expanded to new towns and cities. Today iTnews can reveal that NBN Co has resorted to connecting fewer premises within existing sites than originally promised to allow for this expansion.”

Now, it will come as no surprise that there’s already been a great amount of misleading reporting regarding the NBN over the past few years. Generally iTNews has kept its nose clean and out of the muck-raking that some of the other outlets have engaged in over that period. But there are some disturbing aspects to the iTNews article today which bear investigation.

For starters, the publication alleges not only misleading behaviour on the part of NBN Co in its rollout stats, but also clear intent to mislead. It explicitly links NBN Co’s rollout progress to suburbs in certain Labor electorates (implying politicisation). And there is no comment from NBN Co in the article. We’ve normally found NBN Co quite ethical and transparent with regard to its rollout processes and also open to commenting on this kind of story, so all of this is a little unusual. Personally, I suspect that this is a case where data taken in isolation from NBN Co’s spreadsheets may have been misinterpreted.

With all of this in mind, over the next few days Delimiter will be conducting a fact-checking exercise on this iTNews article. We’ve taken a complete screenshot copy of the article, in case it changes, and we’ll be examining its allegations to see whether they are in fact correct. To help meet these aims, we invite all Delimiter readers to help us. It appears as though all of the data iTNews has used to come to its conclusions is publicly available in the form of NBN Co’s regular monthly ‘ready for service’ updates. Let’s get to the bottom of this issue together and see how accurate the publication’s claims really are.

Let me say up front that if iTNews’ allegations are proven to be correct, Delimiter will join the media outlet in criticising NBN Co for poor behaviour. Likewise, if the allegations are proven to be overblown, we will be the first to call (again) for more accurate reporting of this important infrastructure project by the media. That seems only fair.

45 COMMENTS

  1. Being a fairly logical person one question about the conclusions of the article comes to mind.

    The pressure is on for NBN to deliver available services and connected services. Why sabotage those raison d etre figures for commenced figures. ? As Spock would say not logical

  2. The premise of the article is strange they are trying to fudge the figures by making them publicly available for everyone to interrogate.

    Generally if I want to fudge something I don’t publicly publish it.

  3. If they wanted to fudge something then they would not have removed the premises from the list

  4. “To understand how far-reaching NBN Co’s adjustment was over a single month, only seven of the 79 FSAs were not adjusted. Put another way, NBN Co altered the number of premises in 88 percent of the rollout areas.”

    This is a useless stat the percentage of actual change would be useful this is done for effect not usefulness

    • there is a big table to demonstrate the decreases yet where is the table for the increases what is the overall picture it is never shown.
      It makes sense the areas where there are decreases are old areas and the increases are in newer areas this would seem to be due to remediation works needed to be done by Telstra yet this is not mentioned

  5. What NBN Co has done in a few cases is to select a small area within the larger FSAM to switch on before the whole area is ready.

    One example of this is Maddingley, a suburb of Bacchus Marsh, Vic. It’s a brownfields site (not one of the countless greenfields sites that are active), and a small chunk of 3BAC-02 is now active.

    I can’t find a clever way of linking to a portion of the map, but to see it, just go to http://www.nbnco.com.au/rollout/rollout-map.html and search for the address “Richardson Court, Maddingley VIC”. That will take you there – you might have to zoom in a bit for clarity.

    I honestly can’t see what the problem is with this. To switch on finished areas, and then progressively add areas as they become available, is quite sensible and prudent. Yes, obviously there is good publicity in switching on brand new areas, but since NBN Co is being quite transparent about it (the online map at very lleast is evidence of this) I fail to see the “shock, horror” aspect.

    • OK, I admit – on reflection this comment is of no relevance to the issue at hand. Nonetheless, cue the next bogus NBN outrage story!

  6. Firstly, the WA bownfields rollout delays are Syntheo related, due to the difficulty with getting labour in boom state WA. So that’s the delays with Geraldton.
    Also, the NBN has had problems with the GIS from which they source the subscriber data (remember workers having to walk around to check there are houses where they should be.) So there might be reductions in double ups and errors due to getting accurate information about the premises in the actual areas, resulting in a reduction in premises passed to the real figures.
    But all these questions should have been put to the NBN to allow their comment.

  7. Reading again through the iTNews piece, this jumped out as smelling like BS:

    “In total, a substantial percentage of the fibre connections promised to 14 towns and cities have disappeared from the rollout plan.

    These include a significant proportion of the premises in:

    Armidale, Riverstone, Wollongong and Mudgee in NSW;”

    Armidale? Seriously? The fibre rollout in Armidale is all but complete. Five out of the six FSAMs in Armidale are active, and I can’t imagine the last one is very far behind.

    But they say “a significant proportion of the premises” have “disappeared from the rollout plan” in Armidale? Utter, utter bilge.

    • You know, I just looked at the rollout map for Wollongong, being where I live and all, and couldnt see any noticable change to the rollout map. I’m still slated for it to roll by some time before September this year. Yay!

      Of course, that doesnt address the fact I live in an MDU…

  8. Stan Beer is imitating a previous writer in this site a guy called with a Germanic name like Reutz ( he was an ex News hack can’t remember this guys name but spent most of his time writing anti ALP rant’s in columns barely related to tech).
    I had the privilege of being defriended by Stan!
    Years ago he wrote a column in this “iT” site when it was first founded explaining the virtues of typewriters and his utter confusion in working Windows Vista when XP was the end point of all operating systems, he was really annoyed at having to get 4g of memory when 1g was fine and horribly incensed at having to buy a “Graphics Card” to run aero glass in Vista. He’s real cheap upgrading his PC hardware thinks a PC should last a lifetime and regrets the passing of DOS.
    He’s a man of the past that doesn’t understand the future!

    • You seem to be confusing iTnews and iTWire. To put it another way, nothing you wrote is relevant to this topic.

      • Sorry Did confuse with iT Wire, but had read similar comments in one of Stan’s newssheets.

  9. “It explicitly links NBN Co’s rollout progress to suburbs in certain Labor electorates (implying politicisation). And there is no comment from NBN Co in the article. ”

    iTNews would probably been better off not speculating about the politics but the article does say:

    “An NBN Co spokesperson did not provide a response to several questions raised by iTnews.com.au.”

    By the way, Renai, why is it that Roy Crozier of iTNews must be fact checked but Nick Ross of the ABC gets off the hook?:

    http://delimiter.com.au/2013/02/26/the-fttn-truth-the-coalition-does-not-want-known/

    • I like to think I did due diligence and provided sources for every significant point I made. At the end of the day I wasn’t pushing a personal opinion I was collating facts and presenting them with an effort to inform. I was also hoping that further information would appear as a result of the article in order to further tighten up the conclusions which were based upon the available information.

      Some call it scientific journalism. That’s probably correct – it basically aims to draw conclusions from the best information and sources possible. However, as with science, the conclusions adapt to the continued appearance of new information and facts.

      People will note that, in many cases, I said what the available facts were and explained how *new* information would be needed (to be released in the public domain) in order to draw conclusions that otherwise did not appear to fit the *available* information.

      Ultimately, I’m not pushing an agenda beyond wanting to inform people.

      I literally wanted to know the truth behind many of the assertions being made because the claims being made by different parties were incompatible – they couldn’t both be right. Also, “what a politician says” is not a valid source – I don’t believe I’ve ever used one (might be wrong there and there would naturally be some exceptions in some circumstances).

      To date me and that article have received abuse, been subject to conspiracy theories and generated feelpinion blogs which explain why I am wrong (but which don’t stand up to basic scrutiny and often ignore counterpoints or misquote me for the sake of their unsubtle subjective agendas).

      Basically, if someone has new information which explains why significant points in my article are wrong or need to be modified, nobody would be more interested to hear about them than me. I aimed to write a definitive article to cut through the abject bullshit that the NBN ‘debate’ has become and I can hardly do that if I leave something important out, can I? :)

      • You may have done due diligence in researching your facts and verifying that your facts are accurate but you did not do due diligence in your conclusions.

        You presented an opinion that was emotive and jumped to conclusions, implying that NBNCo was deliberately being dishonest. This, as evidenced by the reply NBNCo made, comes into question. I don’t feel there is enough information to decide either way, but you failed to present a balanced argument.

        That is why you have gotten abuse, and although I don’t condone such behaviour, I can understand it. In future I would recommend, especially when reporting such a politically sensitive issue, to ensure that you learn to split the facts from opinions. It is one to point out changes in NBNCo rollout figures, it is another thing entirely to suggest that NBNCo are, as you put it in your title, “fudging the figures”.

        Do you personally have experience rolling out a national combinations network? Do you know what is involved? Do you know how estimates and predictions of network topology translate into the actual physical cables? Because if you don’t and haven’t, based upon NBNCo’s reply, and my own experience with these matters, the changes they have made seem perfectly reasonable under the circumstances.

        • @NightKaos

          Umm….that was Nick Ross. He was talking about his ABC Article a week ago. Not the ITNews article. Ry Crozier wrote the ITNews article.

          • Yeah I just noticed that myself.

            My apologies to Nick. Ry, if you’re reading this, that was for you. I read the ITnews article originally at something like 2 a.m. local time. It’s a wonder I even remembered the content. I guess the author got lost in the sluge.

            Not to self : don’t read things at 2 a.m.

      • I’ve inserted a comment to make his comment make sense. Seemed like the right thing to do.

      • There was a bogus poster using Renai’s name that made some stupid comment. Kingforce replied, then Renai removed the OP. It left Kingforce’s response as an orphan post, which sticks out as being fairly ambiguous, but if you’d managed to catch the OP before it was deleted it would make sense.

        Wasnt much of a post, just labelling Renai a spiv or something like that.

  10. Oh I get it now. “Renai” is the title of the comment. I actually thought that you were Renai. That would have been weird.

    • Yeah, Renai posts with his full name, which includes a link back to Delimiter. This is just a troll out to cause trouble.

      Didnt last much after Renai (the real one that is) spotted it. So your comment is going to look a little out of place now :)

      • Let’s look at this in actual on the ground terms. Looking at the figures we see that for Gulliver, for example, the premises went from 29,400 to 17,000 and the FSAMs from 12 to 8. Since NBN Co is still going to cover 93% of the population with fibre and nothing has changed there, the FSAMs simply went to another FSA or borders have changed quite a bit. So, that means that, using fictional street names…

        Instead of laying fibre from some piece of equipment in May 2015 up Stoney Creek Rd and Mimsie Rd to Ballamore, they’ll just lay fibre in April 2015 going the other way on Stoney Creek Rd to Ballantyne Rd to Calam St in Gumly. That’s it. That’s basically all this is about. There appears to be no more significance to it than that. NBN Co is simply laying a fibre (of relatively higher importance than most) going a different way than they were before.

        Why would NBN Co reroute this? I don’t know. I don’t even care. Maybe it’s easier? Maybe it’s cheaper? Is it useful in some other way? Whatever. Again, I don’t care. And, most relevantly here, how do you get from here to NBN Co is fudging the numbers? Well, this is how, apparently. Thank you Malcolm Turnbull. Your retweeting of this idiocy of an article has just once again reassured us that some FSAM to be built in 2015 that will be connected to another FSA that NBN Co had previously planned is in your view quite probably yet another reason why we should just give up on the whole thing. How do you get from here to there? Well, this is how, apparently.

        In any case, what’s the next step? Do we say sorry to the people who are now in a different FSA area and will still probably have FTTH at the same time than would otherwise have been the case? After all, with fibre, the distance from the equipment or something matters. Not. Anyway, I’ll say sorry and I’ll do it on behalf of NBN Co since they’re not around here to do it.

        Sorry, people. I’m not quite sure what for, but sorry we’ve had to change our plan at no consequence at all for anyone. Sorry also that you’ve had to endure this article from iTNews because of that. Sorry that we’ve recounted some things and have had to adjust some FSAs by like a hundred premises or so. Sorry also if we’ve had to adjust the borders for some FSAs. And sorry also that we’ve now got plans for more FSAs than we had previously too.

        • That said, it doesn’t fully cover things like the date of when an FSA goes into service or not, but in iTnews’ own words: “If one takes an aggregate view of every piece of brownfields build work with a service-ready date, 58.3 percent (n=201) are on schedule, 27.3 percent (n=94) slipped and about 14 percent (n=48) were fast-tracked.”

          Where the percentages don’t add up to 100% either. And without going into further detail that claim might be about right and is a little bit of concern, but not too greatly so. Most of the slippages seem to be on the order of a month.

          Also, http://www.afr.com/p/technology/homes_to_get_nbn_after_election_j4gaHi1MWY2EY0MtBBYIQL

  11. Under the Fixed wireless:

    Of the 33 WSAs, 55 percent (n=15) remain on schedule and 36 percent still have a service-ready schedule listed only as “TBA”.

    The remaining 18 percent (n=6) saw service-ready timetables slip.

    55 + 36 = 91%

    You can’t add 18% to 91%…..

    So right there, this tells me their “analysis” isn’t great…..

    • ITNews headline for tomorrow – “NBN Fudges figures to show 109% on schedule!”…

    • adding to that:

      “There is also a very small number of stages that have been given ambiguous new timeframes. For example, East Newman 3A and Newman Light Industrial Area 3 in the mine-rich Pilbara both move from “Q3 13” to “Aug 13”

      Going from “Q 13” to “Aug 13” is actually the opposite of “ambiguous”, because they are narrowing the time frame down to a specific month.

      “Today iTnews can reveal that NBN Co has resorted to connecting fewer premises within existing sites than originally promised to allow for this expansion.”

      So they are pulling guys off installing to houses to setup FSAMs and backbones? I thought they’d be different skill/pay levels, but I guess every person working for, say Silcar, must be multi-skilled and highly paid then??

      “Between January and February this year, NBN Co indefinitely delayed the rollout of a total of 69,500 fibre connections promised to cities and towns Australia-wide, redistributing them among 37 existing rollout areas and two new ones.”

      Sooo…they delayed the rollout to keep the rollout going?!! Redistributing them amongst existing areas means the “premises passed” numbers will be maintained, meaning the delay iTNews expects (hopes?) to happen wont eventuate?

      This had nothing at all to do with record fires and floods???!

      “The new rollout areas — accounting for over half (39,200) of the connections taken from existing sites — are in Labor-held seats: Slacks Creek in the Queensland seat of Rankin and Queanbeyan in bellwether Eden-Monaro.

      The biggest losers from the redistribution of these connections are the LNP-held Townsville suburb of Gulliver, which now has 12,400 or 42 percent less fibre connections than originally promised — and the Labor-held Keysborough in Victoria, of which 8100 connections have been cut from the rollout plan.”

      Oh, wow, it’s politically motivated!!!11 LNP loss out to ALP areas!!1

      “The biggest losers from the redistribution of these connections are the LNP-held Townsville suburb of Gulliver, which now has 12,400 or 42 percent less fibre connections than originally promised — and the Labor-held Keysborough in Victoria, of which 8100 connections have been cut from the rollout plan.”

      Hang on, an ALP are got hit too….but that means….

      “There appears to be no political pattern in the data, with ALP, Liberals and Greens constituencies all losing previously-promised fibre connections.”

      Oooooookayyy, they make the case for political bias (for “busy CTO’s” that are just skimming the header), and then show in the statistics (which a “busy CTO” skips) why it’s a bogus claim….

      Seriously, the person that wrote that article missed their calling, day time soaps are the worse for them joining “I.T. news” reporting…

      It’s mostly looking for drama in something that’s intrinsically boring (stats)…

  12. From my own brief analysis of the January 2013 and February 2013 Rollout Plans the following changes have occured (apologies for any mathematical errors):

    The approx number of premises in the FSAMs that were in progress during the January 2013 rollout plan decreased by 21,700 premises in the February 2013 plan.

    The approx number of premises in the FSAs that were in progress during the January 2013 rollout plan decreased by 37,200 premises in the February 2013 plan. 32 FSAs lost a total of 76,900 premises and 38 FSAs gained a total of 39,700 FSAs to provide this overall decrease.

    The Approx premises for the Gulliver FSA fell from 29,400 to 17,000 and the number of FSAMs in the FSA was reduced by 4. The contents of the first table in the article is correct.

    Whilst I don’t have time to check the entire article it appears that from a cursory analysis the figures are correct.

    OPINION:
    Whilst the figures seem to be largely correct I strongly disagree with the implication that the changes have been driven by political considerations. I think it is far more likely that the changes have occur following a revision of the engineering design or an update to the premises database used by NBN Co.

  13. Renai – Are you really so incapable of doing your own research that you rely on your readers to fact check and ask journalists for questions they have sent to the NBN Co?
    Farcical.

    • I don’t rely on the readers — I use their knowledge to supplement my own. And yes, it is of public interest and pertinent to this fact-checking exercise to know what questions Ry sent NBN Co.

      I don’t understand why anyone would object to a fact-checking exercise per se. Truth, after all, is a valuable commodity in politics.

  14. Might I suggest that some of these things are just errors that are being corrected as they are picked up? For example, Riverstone in the brownfields pages. There are 4 FSAMs (2RIV-01, 2RIV-02, 2RIV-03, 2RIV-04) in both January and February. The total number of premises in the FSA is shown in January as 8700 and as 5300 in February (as ITNews shows).

    However if you add up the individual FSAM premises numbers in January, it totals 5400 and in February it totals 5300, so presumably the 8700 in the total column wasn’t correct in the first place. Of course the number of FSAMs was also changed in February to 3 which isn’t right, as the map for Riverstone still has 4 FSAM areas.

    The “extensive 2 week analysis” wasn’t very thorough.

  15. http://www.afr.com/p/technology/homes_to_get_nbn_after_election_j4gaHi1MWY2EY0MtBBYIQL

    Included amongst their usual spin

    “NBN Co on Wednesday rejected allegations it had altered the rollout figures adversely. “The rollout is based entirely on engineering and practical considerations so we can deliver the NBN to as many people as possible as quickly, efficiently and as cost-effectively as we can,” NBN Co spokesman Andrew Sholl said.

    “That may mean bringing forward some build sites by a few months and pushing back others. So it’s hardly some great revelation to suggest that the rollout timing is constantly evolving.

    “No previously announced construction modules have been removed from the construction program.”

  16. I like that Renai is doing this – it’s what the Guardian would call open journalism. For too long, the established has been one-way (we print, you read). Renai is much better at engaging and empowering readers than any other journalist in this space. It doesn’t seem to make him a lot of friends but at least he writes for the engagement and enjoyment of his readers rather than for pats on the back from other journalists.

  17. I’ve seen iTNews print some howlers before by claiming technical maturity but failing to understand the technology being discussed. This has happened across many channels. I don’t know why, since I know many top-line experts more than happy to guide and proof articles for technical accuracy.

    It is clear they’re after page impressions, not factual dissemination, as we all know, howlers can attract more attention by being shared than dull old facts and truth.

  18. Hi Renai,

    There’s a BRILLIANT analysis by Aaron from SkyMesh over at Whirlpool on the NBN rollout thread:

    http://whrl.pl/RdwDv8

    As a summary, ITNews have used the metric “FSA total premises” as a way to measure NBNCo’s changes in the rollout. As Aaron very carefully points out, that metric is not relevant to the actual rollout taking place right now . The total, final FSA size, in premises numbers, will fluctuate over the rollout, which is what ITNews have used to say NBNCo. have decreased the rollout premises by over 39 000 premises. In fact, the Rollout Regions (basically FSAMs) have INCREASED in size (see his analysis) over the last month, but the FSA sizes have been redistributed, causing some to shrink (with current rollout taking place in them) some to grow (with current rollout taking place in them) and some to shrink or grow WITHOUT rollout taking place in them, meaning those FSAs wouldn’t even be IN the current rollout numbers.

    Basically, ITNews have taken the FSA size to be a direct measure of the rollout numbers. They are not. The Rollout Regions are. And they do NOT correspond directly to TOTAL FSAs (made up of many FSAMs)- an FSA’s total size will change over time as premises are shuffled around, but that will NOT necessarily change the premises numbers if the premises moved are not currently in a Rollout Region. Therefore changing the size of an FSA DOES NOT necessarily change the actual premises currently being rolled out to. And in fact, it has INCREASED the premises count in the Rollout Regions, not decreased.

    It’s a VERY thorough analysis :)

Comments are closed.