ComReg broadband speed pilot

oscarBravo

http://www.comreg.ie/_fileupload/publications/ComReg1581.pdf

One thing that puzzles me deeply about the report is the lack of comparison between the advertised speed of the product each customer is using versus the actual speed they're seeing.

ItHurtsWhenIP

oscarBravo wrote: »

http://www.comreg.ie/_fileupload/publications/ComReg1581.pdf

One thing that puzzles me deeply about the report is the lack of comparison between the advertised speed of the product each customer is using versus the actual speed they're seeing.

Disappointed that they would test mobile mid-band which possibly skewed some of the results, but there was no FWA in the mix.

Also would they not consider asking Sam Knows for their data?

Ah here:

30. ComReg’s experience from this Pilot is that it can be difficult to recruit a sufficiently large, robust and representative panel of users willing to download a software tool onto their home PCs or laptops and keep it active over a period of time necessary to gather sufficient test results.

All they have to do is come on here and ask us - we'll give them a hell of a panel of users!!!

oscarBravo

The fact that it's Windows-only software running on a single PC makes it less useful also. A single-PC household is fast becoming a rarity in my experience.

ItHurtsWhenIP

oscarBravo wrote: »

The fact that it's Windows-only software running on a single PC makes it less useful also. A single-PC household is fast becoming a rarity in my experience.

They did highlight this in the report, so I hope they do something better next time (if there's a next time).

The other thing that bugs me is the way they compared wi-fi and wired. There are just too many variables that affect wi-fi performance, that they should have simply excluded them and only taken wired connections ... having said that a single-PC wired household is becoming a rarity.

This is why I brought up about the Sam knows data. That would be much more representative.

clohamon

MMFITWGDV wrote: »

Disappointed that they would test mobile mid-band which possibly skewed some of the results, but there was no FWA in the mix.

The mobile wireless users were removed

The remaining 51 panellists are mobile broadband panellists and due to the small sample size, these results are not included here.

clohamon

Some other notes.

This survey has taken over four years to complete. It seems the only thing revealed is how poorly the survey was designed.

• Survey only includes those who actually have broadband.
• Does not reveal location other than self-select Urban/Rural
• Does not reveal provider. (see OP oscarBravo above)

• Measured on a user device rather than network terminal.
• Margin for error +/- 10% due to small sample size
• Network Latency: measured but not discussed
• Web page download speed: measured but not discussed
• DNS look-up: measured but not discussed
• Upload speed measured but not discussed.

"The objective of the Pilot was to measure user experience; to gather information on actual broadband speeds being experienced and to provide ComReg with a national picture or overview of the speeds experienced across a range of platforms.”

but….

"None of the speeds recorded in this trial are statistically significant to the extent that they can be applied to the general population."

"Further detail based on the Pilot undertaken In November 2014 will not be available due to small sample sizes."

"Having considered the options ComReg may consult with respect to further work in this area.”

SpaceTime

Doesn't sound much use then.

There are still certain islands of urban areas with significant broadband access issues due to limited eircom cabinet coverage where all the lines were done with underground splits : A large chunk of Dublin 1, 2 and 7 as well as Cork's central island spring to mind. Many 1950s era housing estates also seem to be In that situation.

UPC also gave areas in urban Ireland that aren't covered again, chunks of Central Dublin feature in this.

Rural broadband is highly variable depending on what level of connectivity the local exchange had to the backbone - still a % of small exchanges only offering very slow speeds and ADSL only.

Really remote places are totally dependent on wireless techs and that varies enormously.

Without locality details this survey isn't going to be useful for planning anything.

watty

You can only do this the way Sam Knows does it. With a customised Router.

This is so garbage it's unreal.

BarryM

MMFITWGDV wrote: »

All they have to do is come on here and ask us - we'll give them a hell of a panel of users!!!

+1, however, that might reveal things they don't want revealed, like the amount of money that they have spent for no significant improvement.