WTF? Tried Jehovahs Witnesses to delete Raymond from Wikipeda a couple minutes ago? (page 12)

besty

I know I have gone too far when Randy shows up on my threads :-)

How is Cairns at this time of year anyways jeffro?

Jeffro

I checked with the ISP and the actual location was Norwich - 40 miles from where geoip.co.uk said he was.

So what? You spun the wheel and it was within some arbitrary distance you've decided is acceptable. Who cares? You proudly said "the Westerfield area of Ipswich", and you were wrong.

You said geolocation wasn't accurate beyond country level

No, I didn't say that at all. Nor did I say that my own result was specifically representative of anything, but only that geo-location on its own can't be relied upon to determine a client's exact location. I also explicitly stated, 3 days ago, before any of your extended pontification, that geo-location can be correct to city level. (Remember, the post where you only quoted the more general part of what I said?)

How is Cairns at this time of year anyways jeffro?

Your powers of deduction are much less inspired than you imagine them to be, and you clearly have too much spare time.

besty

I concede I was specifically wrong about Ipswich, but in general IP geolocation for consumers works with enough accuracy to be an extremely valuable commercial proposition, particularly when layered with other services.

http://www.maxmind.com/en/city_accuracy - correctly finds Norwich in this case - they claim city level 72% accuracy in the UK and 81% in the USA, which supports my position of 'reasonable certainty'

You said:

But your assertion was that there was "reasonable certainty" that the person was in Norfolk, when it's actually much less certain .

The experts who make their living with this stuff state 72% accuracy for the UK.

Have you changed your mind, as quoted above, that 'reasonably certain' is a fair position to take, and that the degree of certainty is higher than you initially gave credit for? ie its actually much more certain, rather than much less certain?

I should have used their website to start with instead of geoip.co.uk :-)

besty

Your powers of deduction are much less inspired than you imagine them to be, and you clearly have too much spare time.

You are not in England and I'm not in California :-)

Jeffro

but in general IP geolocation for consumers works with enough accuracy to be an extremely valuable commercial proposition

Never said otherwise.

You are not in England and I'm not in California :-)

Nice work genius. Aside from the fact that there's nowhere in the UK 2000km from anywhere else in the UK. And that I've said several times on this forum that although I was in the UK at one point, I've since been unable to update my location. But, yeah, brilliant.

http://www.maxmind.com/en/city_accuracy - correctly finds Norwich in this case - they claim city level 72% accuracy in the UK and 81% in the USA, which supports my position of 'reasonable certainty'

So that particular company (claims it) will get Norwich right almost 75% of the time. (Well, actually it says it will get within 40km of the correct city 72% of the time in the UK.) Your benchmark for geolocation being 'reasonably certain' is fairly arbitrary. You've only demonstrated that geolocation can be just as hit-and-miss as I suggested.

I concede I was specifically wrong about Ipswich,

Very big of you to finally concede the actual point I made 3 days ago to which you so doggedly objected. Peace out.

besty

Why are you so defensive? I was joking about our respective locations - hence the :-)

Is 'fairly arbitrary' an oxymoron? just sayin...:-) kinda like 'quite random' etc... - no?

What I actually said Page 4 was:

In this case we can say with reasonable certainty the user is located in the Norfolk area, whether thats Ipswich, Bury St Edmunds or wherever.

What you immediately replied was:

We can reasonably conclude that they're in the UK, and they might be somewhere in the Norfolk area. Geolocation of my IP is not even near close. Why should we expect much better for this one?

Your position on my assertion and your counter-position was summarised by yourself Page 5:

But your assertion was that there was "reasonable certainty" that the person was in Norfolk, when it's actually much less certain .

We are arguing about the degree of certainty of geo-IP location services.

if 75% accuracy doesn't meet your criteria for 'reasonable', I'll leave it there. Noted that you didn't answer my question directly. Have you changed your initial position or not? As a reminder, see above, you stated that geo-IP location is 'much less certain' than 'reasonably certain'. I have provided evidence that 75% accuracy is a commercially saleable proposition - are you standing by your original position?

So that particular company (claims it) will get Norwich right almost 75% of the time.

No - that particular company claims 72% accuracy to city level for any UK IP address. Claiming an accuracy level for a single obscure UK town wouldn't really be a sellable business service :-)

Very big of you to finally concede the actual point I made 3 days ago to which you so doggedly objected.

Can you post the bit from 3 days ago where you say 'this IP address is definitely not in Ipswich' - thanks.

(You understand the obvious logical difference between 'definitely not in Ipswich' and 'can't be certain it is in Ipswich', I assume?)

Jeffro

No - that particular company claims 72% accuracy to city level for any UK IP address.

they say "within 40km of" the correct city in the UK 72% of the time

if 75% accuracy doesn't meet your criteria for 'reasonable',

it doesn't meet my criteria for reasonable certainty. And certainly not your initial claim of 'the Westerfield area of Ipswich', which was the only thing I originally objected to, and I was right about it. If you weren't so proud, it would have ended there.

Can you post the bit from 3 days ago where you say 'this IP address is definitely not in Ipswich'

Strawman argument.

I have provided evidence that 75% accuracy is a commercially saleable proposition - are you standing by your original position?

So what? Your original implication was that it is nearly 100% reliable for tracking down JW apologists in Ipswich.

besty

your initial objection did not mention Ipswich - it couldn't possibly have - I had already diluted my estimate to

In this case we can say with reasonable certainty the user is located inthe Norfolk area, whether thats Ipswich, Bury St Edmunds or wherever.

unless you want to re-post the bit where you say 'this user is not in Ipswich'???

besty

its not a strawman:

And certainly not your initial claim of 'the Westerfield area of Ipswich', which was the only thing I originally objected to, and I was right about it

you claim to have originally objected directly to 'Ipswich' - I am asking you to post evidence of this very specific claim.

Jeffro

you claim to have originally objected directly to 'Ipswich'

No, I objected to your claim of accuracy inherent in your statement about 'Westerfield in Ipswich'. But I think you already know that.

WTF? Tried Jehovahs Witnesses to delete Raymond from Wikipeda a couple minutes ago?

Share this