Another new build neighbour speed difference

Good news is the new estate has had is cabinet topped up and we're in line to get VDSL. Naturally in my over excitement I've been checking the potential speed. I'm only able to look using the address checker at dslchecker.bt.com, some houses are occupied but others still under construction.

A couple of odd things, one slightly worrying. Speeds appear to increase further along the road from Cabinet 20 at HALFWAY. Speeds for the first few houses are (best case, rounded figures, down speed & up speed) 64 & 17 but then increase to 69 & 20 before dropping down again. Odd, but I'm using an address check only so probably not surprising. But then again the nearest house which I believe occupied only has 50 & 16. Next door (further from the cabinet) is 64 & 17.

But most concerning is one house in a row of 69 & 20 is shown at 55 & 13. Neighbouring houses either side get 69 & 20 so why would one show at 55 & 13? All of these houses are still under construction. I doubt Openreach have been near. So why would one be different? What does the dslchecker think it knows that makes this house slower?

Edited by Woolwich (Tue 23-Jan-18 15:24:23)

Do the homes actually have active copper telephone lines yet?

The checker and estimates are just that and thus you may see some variation depending on what raw data they have, or if its someone looking at a map there may even be a human element for variations.

Until they are live hard to be sure, if close to the cabinet and estimates suggest that and no aluminium involved (new build so not the case) then its a curiosity thing, unless someone wants to pay for a survey to double check things, i.e. to keep costs low estimates are just that.

In reply to a post by MrSaffron:

Do the homes actually have active copper telephone lines yet?

Some do, others are literally just shells. I'm sure the house with 50 & 16 is occupied and I'm assuming (!) has a phone line (no Virgin so it would have to be BT/Openreach) as is the next door with 64 &17. I don't have phone numbers to confirm but when I check a speed at an address I know it shows the result I know to be correct. (My logic is the checker is showing the actual connection speed for the occupied houses.)

its a curiosity thing

I agree but ISPs are using this data so will only offer me a connection at a slower speed. Meanwhile neighbours either side will be offered and guaranteed a faster connection.

And that is one of the problems arising from guarantees and median speeds, providers erring on side of caution.

The checker does I believe take some feedback once a service is live, but this is not the actual connection per se but can help to correct things.

In old days it was order fastest service and downgrade if it does not meet your hopes, now its more order the slower up to 38 or up to 52 Mbps service and if stats show more possible hope checker updates and upgrade then or present data to provider and they'll place an order.

Openreach does no filtering i.e. it offers estimates and what providers order is what they provide, then down to the electronics and physics of the line to determine the speed you will get.

My logic is the checker is showing the actual connection speed for the occupied houses.

That's very much not the case.
My line gets considerably higher than my clean/high estimate (although the estimate is slowly climbing).
2 of my neighbours get considerably less than their clean/high estimates.

I've never seen results from the checker show a perfectly even decline in speeds as distance from the cabinet increases. There's always random properties with higher/lower than expected.

It's also not as simple as which house is closer to the cabinet has the shortest line.
Houses 1,3,5,7 might be nearer the cabinet than 9,11,13,15. The DP/underground chamber they are fed from might mean their lines go a longer route than further away properties. It's just not as simplistic as it initially appears.

They are just estimates, not particularly accurate ones. Add to that most of the properties you are searching aren't even built, the large variations are to be expected.

Edited by j0hn83 (Tue 23-Jan-18 18:37:49)

There is definitely some adjustment, my old line for example used to report 80Mbps as the max, 70 as impacted, but as crosstalk slowly reduced me to a sync of 55Mbps the estimates dropped to show 50Mbps as the max, and 35 as impacted, a fairly significant adjustment.

Similar story at my parents.

At a flat I lived in with extremely poor internal wiring (multiple masters), cables running by electrical mains for 4 floors, my estimate was 40Mbps on move in... Once the wiring was fixed by an engineer and I started syncing at the full 55Mbps with a max rate in the router displaying 85Mbps, the estimates shop up to 70Mbps.

No exact science although I do believe existing user sync feeds into it. Certainly poor internal wiring in my case, which likely existed for years, had enormously reduced the estimates considering the fibre cab was on the sidewall of my flat.