Not applicable

speed test

Why do different speed tests vary at times up to 50% at the same time of day? For lack of better terms, corporate test seems to give what the ISP says the speed should be 9 (If it should be 700M Mbps that's the result) while independent test give vastly different results. Most of the test I've run rely on Jave or Flash. A few of the independents are starting to write in HTML5 but still give results similar to independents that use Jave or Flash. With the exception of a time of day where there is less Internet traffic, when almost all test agree within ±5%.

 

I like the metaphor of a freeway. It seems the corporate test looks at the posted speed limit while the independent ones look at the prevailing traffic conditions. Which might explain why at 2am CDT all three types come close to being the same.

 

If it is possible, I would appreciate as unbiased an answer as anyone is allowed to give.

 

 

26 REPLIES 26
Valued Contributor

Re: speed test

Try different test sites. I never fully trust speedtest.net. ISPs give priority to speedtest.net.

 

http://testmy.net/

fast.com

http://www.dslreports.com/speedtest

Not applicable

Re: speed test

I have already found several sites that give results more in line with what I see happening.  Of the three, one is derived from HTML protocol that has been around since Babbage cut ten holes in a 2X4 thus creating the first calculator.

 

The problem is that if an ISP will not take results they won't do anything about speed discrepancies. They do admit that there is noise on the line at but won't explain why their speed test shows results twice that of another speed test not connected with the ISP.

 

So why bother with other speed test, if the ISP will only admit a problem if the test run is the one they favor?

That only makes end users like myself frustrated by a knowledge that bears no weight on a solution.

 

.

 

 

 

 

 

Expert

Re: speed test

It's because anything outside the TWC intranet pipeline is out of their control and have often been throttled by AT&T, Tcore, Level 3 and Cogentco

 

Helper

Re: speed test

If you have noise on the line and a local variance, its not hard to prove you have and hsd signal page, right. Well if you have a lot of noise, there is good chance your upstream signal will be hitting the 48 49 threshold(not always, though) but its a darn good indicator.
Not applicable

Re: speed test

Look, I appreciate the replies to my question of varying speed test results. In waiting for an answer, I realized that knowing the answer will not explain why this ISP will do nothing about the disagreements. Then I started looking at other markets serviced by the same 4 or 5 companies and realized that they all have a customer satisfaction rating of around 50%. From the east coast to the west coast of this country that is all a customer expects, an approval rating of 50%. So if the big 5 service providers exists on a level that wouldn't keep any of us employed very long, who is happy?

 

Someone right?

 

 

Expert

Re: speed test

TWC advertises it as "Speeds up to xxx"    They willseldomly respond to anything in the 80-100% of what you pay for

All speeds must be measured to their speedtest site and with a wired connection.

 

Proven Sharer

Re: speed test

Also understand that the TWC number only applies WITHIN the TWC network. As soon as your network traffic is routed over another carrier all bets are off because those transport circuits and equipment may be undersized for the combined peak demands being placed on them.

Observer

Re: speed test

Them TWC should do full disclosure of this possible situation and let the customer know they may be paying for a service that is not possible.

Not applicable

Re: speed test

thanks