Home  ›  News  ›

4G Networks Tested: WiMAX vs. HSPA+

Intro The Numbers Equipment Summary Comments  

all discussions

show all 13 replies

Dumb conclusions

tbacba

Jun 4, 2010, 1:31 PM
By including location F, the overall average was greatly skewed. PS admits there was little 4G coverage in the burbs. The 4G signal there appeared almost non-existent, and for some reason the HSPA was the strongest. Drop location F from the test and WiMax ends up 50% faster on download, although still much slower on upload.
...
tbacba

Jun 4, 2010, 1:58 PM
🤭
...
Rich Brome

Jun 4, 2010, 2:22 PM
We thought it was very important to have different kinds of locations, including the 'burbs. If anything, we would have liked to include more locations outside of downtown. People need to use 4G not just in city centers, but other places, as well. At least that's the promise Sprint is advertising, so it's only fair to test and rate them on performance in all of the places they advertise coverage.

Location F was in a well-populated area, well within what Sprint's coverage maps showed as 4G coverage. It wasn't at the edge of coverage or anything like that.

Every carrier is going to have strong and weak locations in their network. So naturally, excluding any one location will change the overall average. But excluding one location b...
(continues)
...
Kayslay34

Jun 4, 2010, 3:42 PM
Did you even read the pcmag article, when they showed the test results it all showed sprint 4g being faster in almost every area... So i think pcmag doesnt know how to read there own data. They kept giving AT&T the award when the test results clearly show sprint 4g being faster LOL.... http://www.pcmag.com/article2/0,2817,2364263, 00.asp

For the lazy people here are some examples.

Northeast they gave to tmobile but tmobiles speed was 1.42 avg 2.01 max, when sprints wimax was 3.09 avg and 3.09 max

Southeast they gave to ATT but there speed was 1.79 avg and 1.95 max, when sprints 4g was 2.27 avg 3.14 max.

Is today april 1st or are people blind?
...
mingkee

Jun 4, 2010, 5:52 PM
This test is based on advanced 3G (3.75G vs WiMax).
ATT doesn't have 3.75G yet, so does LTE with VZW.
...
Kayslay34

Jun 8, 2010, 2:32 PM
How is it sprints fault that att doesnt have that..... Its in the article data, sprint wins.
...
tbacba

Jun 4, 2010, 9:30 PM
Fair enough, but using the average is probably not the best tool for this comparison. The extremes greatly affect the result. I think the median value would be more representative of actual performance.
...
Iknownothing

Jun 5, 2010, 1:35 AM
Begging your pardon but,
1. When making this comparison network technology and network penetration should be two separate and distinct values. Conflating them, as you did, certainly does skew the numbers.

2. Comparing an area where one network has advanced experimental coverage directly to an area where the other network has poor coverage without including an opposite representation simply because you would have had to travel outside the city limits to find it is just silly. And yes misleading.
...
Rich Brome

Jun 5, 2010, 10:49 AM
Iknownothing said:
Begging your pardon but,
1. When making this comparison network technology and network penetration should be two separate and distinct values. Conflating them, as you did, certainly does skew the numbers.

I disagree. The goal wasn't to test what the technologies are capable of in some theoretical world. The goal was to test these technologies in the real world, in places people might actually need to use them.

2. Comparing an area where one network has advanced experimental coverage directly to an area where the other network has poor coverage without including an opposite representation simply because you would have had to travel outside the city limits to find it is ju
...
(continues)
...
Chippewas86

Jun 5, 2010, 9:54 AM
Rich Borne,

I agree with you that in order to perform your "tests" in a "fair" manner you must include all of your data. However, I believe that your "empirical" conclusions are still biased. Your outlier, in this case location F, significantly affects your data which arguably flaws your conclusion. In order to reduce the significance of these anomalies, your sample should be much larger. Because of your small sample producing skewed results the only conclusions people can draw will be either false, or that your data is misleading. It is unfortunate that a usually credible news source is, and will, wrongly influence readers with this misleading comparison. I hope that your testing techniques will improve going forward.

Sincerely,

...
(continues)
...
P-TRIZZ2931

Jun 5, 2010, 10:25 AM
Danggggg Rich!!! you got burnt! 😲 😲 Lack-Luster misrepresentation!!!!!
...
Rich Brome

Jun 5, 2010, 10:40 AM
I completely fail to understand how the conclusions are "false" or "misleading". Our conclusion wasn't "T-Mobile beats Sprint". Our conclusion was that they're both fast.

If you removed location F, WiMAX would come out a bit faster on the average, on download speed, but I would still reach the same conclusion that HSPA+ is roughly comparable in speed and T-Mobile's claim of "4G speeds" is valid. That was the main point of the article.

The upload and latency numbers were quite consistent. That wasn't an "outlier". I highly doubt that more testing in more locations would have changed the solid conclusion that HSPA+ is faster in those regards.
...
Kayslay34

Jun 8, 2010, 2:33 PM
How is it not Misleading...
Did you even read the pcmag article, when they showed the test results it all showed sprint 4g being faster in every area... So i think pcmag doesnt know how to read there own data. They kept giving AT&T the award when the test results clearly show sprint 4g being faster LOL.... http://www.pcmag.com/article2/0,2817,2364263, 00.asp

For the lazy people here are some examples.

Northeast they gave to tmobile but tmobiles speed was 1.42 avg 2.01 max, when sprints wimax was 3.09 avg and 3.09 max

Southeast they gave to ATT but there speed was 1.79 avg and 1.95 max, when sprints 4g was 2.27 avg 3.14 max.
...
cellphonesaretools

Jun 6, 2010, 11:50 PM
Check out this PC World article giving data from five cities. It pretty much corroborates the data that was given in this PhoneScoop article.

table of data:
http://www.pcworld.com/zoom?id=198054&page=1&zoomIdx=1 »

full article:
http://www.pcworld.com/article/198054/sprints_evo_ph ... »

To say the author of the PhoneScoop article should have discarded the "F" data just because it lowers the average, is total BS. That suggestion is totally in conflict with the scientific method and is counter to the incredibly important matter of integrity of the data.

Read the real-world data and weep, or ignore it if you personally don't like it, but don't EVER discard data points just because they conflict with your...
(continues)
...

This forum is closed.

Please log in to report a message to the moderator.

This forum is closed.


all discussions

Subscribe to Phone Scoop News with RSS Follow @phonescoop on Threads Follow @phonescoop on Mastodon Phone Scoop on Facebook Follow on Instagram

 

Playwire

All content Copyright 2001-2024 Phone Factor, LLC. All Rights Reserved.
Content on this site may not be copied or republished without formal permission.