 AMDUSERPremium
join:2003-05-28
Earth
kudos:1
 ·AT&T U-Verse
 ·RoadRunner Cable
1 edit | The test is inaccurate... I tested my cable internet on it, and it said I had about 30 Meg (even though I only have 7 Meg Road Runner, but with Powerboost(tm).
If they want accurate, they are going to have to do something to mitigate the Powerboost to get the rated service speed. |
|
|
|
 EricthornIt only hurts when I laughPremium
join:2001-08-10
Paragould, AR | Agreed - it's giving me about 2.7m down and any other test anywhere I go gives me less than 2m, usually in the 1.8m range which is accurate.
It wouldn't surprise me a bit if the carriers have some hand in this...
--
Ever try stuffing a melted marshmallow up a wildcat's ass? It can be done, but you have to like your job. - This Is The Way The World Ends by James Morrow - Join a DC club, it can't hurt you! |
|
1 edit | said by Ericthorn:Agreed - it's giving me about 2.7m down and any other test anywhere I go gives me less than 2m, usually in the 1.8m range which is accurate. It wouldn't surprise me a bit if the carriers have some hand in this... I agree. It is pretty inaccurate. It doesn't even match what regular speedtest says:
Here is my result from broadband.gov speedtest:
And here is regular speedtest:
There is no powerboost on my connection. |
|
 funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:6 | reply to AMDUSER
said by AMDUSER:I tested my cable internet on it, and it said I had about 30 Meg (even though I only have 7 Meg Road Runner, but with Powerboost(tm). If they want accurate, they are going to have to do something to mitigate the Powerboost to get the rated service speed. If it measured 30 meg, you want it to dismiss it?
The goal here is to see whether you're getting what you paid for.
If it says 3 Mbps and you're paying for 7 Mbps, then you're not.
If it says 7 Mbps or 30 Mbps and you're paying for 7 Mbps, then you are.
Powerboost is real speed. So some credit should be given to that fact. But it is really temporarily and not assured, and I agree that something should also measure that. But what we don't want measurement tools to do is to discard actual performance.
--
Robb Topolski -= funchords.com =- District of Columbia -- KJ7RL
Tweet! Tweet! -- »twitter.com/funchords |
|
 LinklistPremium
join:2002-03-03
Longport, NJ
kudos:5 | said by funchords:Powerboost is real speed. So some credit should be given to that fact. But it is really temporarily and not assured, and I agree that something should also measure that. But what we don't want measurement tools to do is to discard actual performance. If you want a test that handles distortions caused by powerboost, use this one:
»www.measurementlab.net/measureme···iffprobe |
|
 tschmidtPremium,MVM
join:2000-11-12
Milford, NH
kudos:8
 ·G4 Communications
·Fairpoint Commun..
 ·Hollis Hosting
| reply to AMDUSER
I ran the test twice; the first time jitter and latency were the same, which is impossible. Transfer speed measurement seemed reasonable accurate.
Overall I think having FCC do its own research is laudable but there are several serious caveats:
1) This is by no means a scientifically credible poll of US broadband.
2) We, like many others, have home network. To be an accurate determination of ISP performance no other computer should be using the Internet for the duration of the test. I did not see any mention of that on the test page.
3) Likewise applications on the test PC should not access the Internet for the duration of the test.
4) Performance testing is end-to-end. Issues outside the control of the ISP can degrade performance.
5) As already mentioned, temporary speed increases, will artificially inflate test performance.
/Tom |
|
funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:6 | said by tschmidt:I ran the test twice; the first time jitter and latency were the same, which is impossible. Transfer speed measurement seemed reasonable accurate. Overall I think having FCC do its own research is laudable but there are several serious caveats: 1) This is by no means a scientifically credible poll of US broadband. 2) We, like many others, have home network. To be an accurate determination of ISP performance no other computer should be using the Internet for the duration of the test. I did not see any mention of that on the test page. 3) Likewise applications on the test PC should not access the Internet for the duration of the test. 4) Performance testing is end-to-end. Issues outside the control of the ISP can degrade performance. 5) As already mentioned, temporary speed increases, will artificially inflate test performance. /Tom All true. There are several more issues. That said, these are data points in a set of gathered data. Once analyzed, some test results will be statistical outliers. In the end, analysts should be able to refine and clarify the picture by filtering out much of the noise.
--
Robb Topolski -= funchords.com =- District of Columbia -- KJ7RL
Tweet! Tweet! -- »twitter.com/funchords |
|
 56403739Less than 5 months leftPremium
join:2006-03-08
Naples, FL
kudos:2 | reply to tschmidt
The M-Labs version is trash. The Ookla version tested three different lines in three different locations accurately and more consistently than the tests from this site do.
The speedtest.net site seems to do the best of all, though, since it looks for the best server for your location. |
|
| reply to funchords
said by funchords:The goal here is to see whether you're getting what you paid for. I have to disagree. The goal for the Gov't speed test is to sample what speeds are available where.
If powerboost causes a test to show 30Mbps then it is a false result. |
|
 tshirtPremium,MVM
join:2004-07-11
Snohomish, WA
kudos:3
·Comcast
| reply to fifty nine
There are 2 tests available at BB.gov the Ookla (short test, see maximum powerboost effect) and the Mlab version (larger/more complex test, PB has little effect, (but jitter measuremnet is unrealistic) )
Which test you see first is randomized, everyone should be sure to take BOTH tests, each time they visit the site, so any data collected is closer to reality.
That said it is a beta, and the location data is probably more important than the speed data at this point (if you don't feel comfortable using your own address, please use one nearby, so as not to skew the location HAVE/HAVE NOT part. It's not like you really are hidding using a fake/distant location, they can trace your IP if the really want too anyway.) |
|
| reply to funchords
Rob,
With all due respect, any real honest statictician would have a hard time using any of this data, because it violates basic rules of sample selection bias. There are ways to correct for that, but nothing that could render anything worthwhile from these self-selected tests.
The FCC should take a cue from Offcom, and do a SamKnows style testing, which corrects for selection bias as well as all sorts of other intervening factors. |
|
| reply to tshirt
The problem is that the mlab test doesn't work with Firefox on Linux, which is what I'm using.
It also doesn't work with chrome on windows 7 which is what I also used.
If this test is IE only they're going to get a lot of skewed results. Not everyone wants to use IE. |
|
LinklistPremium
join:2002-03-03
Longport, NJ
kudos:5
1 edit | said by fifty nine:The problem is that the mlab test doesn't work with Firefox on Linux, which is what I'm using. It also doesn't work with chrome on windows 7 which is what I also used. If this test is IE only they're going to get a lot of skewed results. Not everyone wants to use IE. From FCC:
Note: the M-Lab application currently does not work with Safari, Chrome, and Opera web browsers.
The ookla test should work:
Ookla also provides free services at their public web site Speedtest.net.
It is the same test as used at speedtest.net |
|
 jlivingoodPremium,VIP
join:2007-10-28
Philadelphia, PA
kudos:1 | reply to funchords
Robb - You are the resident M-Labs expert... Do you know how many TCP connections that the M-Labs tool uses?
Also - do you have any idea how the test decides to use one test or the other (M-Labs or Ookla)?
--
JL
Comcast |
|
LinklistPremium
join:2002-03-03
Longport, NJ
kudos:5
2 edits | said by jlivingood:Robb - You are the resident M-Labs expert... Do you know how many TCP connections that the M-Labs tool uses? Also - do you have any idea how the test decides to use one test or the other (M-Labs or Ookla)? The FCC web site says it picks it at random. But when the test ends, you can then run the other test.
»www.broadband.gov/qualitytest/about/
Users will be assigned to one of the two chosen testing tools: Ookla or M-Lab, or they can directly choose one of these tools at the links at the bottom of this page. But you can pick which one you want from the above link or use the links I found below:
Ookla:
»www.broadband.gov/qualitytest/?sCode=O
M-Lab:
»www.broadband.gov/qualitytest/?sCode=M |
|
EricthornIt only hurts when I laughPremium
join:2001-08-10
Paragould, AR | reply to Linklist
If it's the same then why are the results so skewed? Ookla had inaccurate d/l data while speedtest was accurate. |
|
LinklistPremium
join:2002-03-03
Longport, NJ
kudos:5 | said by Ericthorn:If it's the same then why are the results so skewed? Ookla had inaccurate d/l data while speedtest was accurate. Are they? I get similar results:
Not identical, but I wouldn't expect them to be. Which speedtest server did FCC choose for ookla test? It may not be the same. And how long the test runs for the FCC vs native speedtest.net may not be the same. |
|
| They are way off.
Speedtest.net is more along the line of what I'd expect, whereas broadband.gov's test showed about one third of our total download speed and one fortieth of our total upload speed. |
|
jlivingoodPremium,VIP
join:2007-10-28
Philadelphia, PA
kudos:1 | reply to Linklist
Thanks!
From the office the Ookla test returns: 35M/3M
and M-Lab: 21M/4M
--
JL
Comcast |
|
| reply to AMDUSER
Way off here also, my home DS3 test is way lower then what it should be. Of course I suppose I should have entered an accurate address, as usual government fail. |
|
