PC Mag "Fastest" Mobile Networks

[JWMaloney]

What is the difference between 5mbs and 50 from a smart phone? In terms of use? The answer is none.

 

The answer lies within the standard deviation -- how bad are the worst tests to bring a 50 Mbps peak down to a 5 Mbps average?

[utiz4321]

The answer lies within the standard deviation -- how bad are the worst tests to bring a 50 Mbps peak down to a 5 Mbps average?

True enough, that is why this results are uninformative. The don't speak to user experience. Which is why many people here dislike them including myself.

[JonnygATL]

Johnny, they were probably aware of it. But they couldn't postpone the article for the eCSFB issues to be cleared up. The world keeps spinning. My only concern about this article is that it really does make sprint look like it's in for yet another year of "almost there"...when in reality this is supposed to be be "sprints year" for a come back.

 

Edit: I've come to the conclusion this is the only page that needs to be viewed. http://www.pcmag.com/article2/0,2817,2459189,00.asp

I think that is exactly my concern as well.  It colors the results when the network is actually much better than the results would suggest.  More important are the continuing and substantial upgrades going on right now that won't be evident to most readers/users until a year from now when the next PCMag test hopefully shows a greatly improved network.  

 

I just really think it's high time for some positive press on Sprint.  And, despite what another user suggested, going from bad and even awful to usable and even somewhat more than adequate is a MASSIVE change.  I'm happy for Sprint and am glad I stuck around.

[jamisonshaw125]

 

I just really think it's high time for some positive press on Sprint.  And, despite what another user suggested, going from bad and even awful to usable and even somewhat more than adequate is a MASSIVE change.  I'm happy for Sprint and am glad I stuck around.

Agreed. I am definitely glad I stayed around. I enjoy my bill, I enjoy my service, I enjoy the customer service when I need it... Sprint is the perfect representation of Misunderstood Douchebag Guy

[milan03]

I think that is exactly my concern as well.  It colors the results when the network is actually much better than the results would suggest.  More important are the continuing and substantial upgrades going on right now that won't be evident to most readers/users until a year from now when the next PCMag test hopefully shows a greatly improved network.

It's an annual test, done over three weeks period of time, and it's supposed to showcase data performance of all four Tier 1 wireless operators at this point in time. That's exactly what they did.

 

They've also used commercially available user equipment, and travelled to same metro areas year after year to capture the progress.  Although I'm not a fan of these tests, at least they are consistent.

[JonnygATL]

It's an annual test, done over three weeks period of time, and it's supposed to showcase data performance of all four Tier 1 wireless operators at this point in time. That's exactly what they did.

 

They've also used commercially available user equipment, and travelled to same metro areas year after year to capture the progress.  Although I'm not a fan of these tests, at least they are consistent.

Yes, I know, I'm familiar with the methodology.  I did read that part as it's vitally important.

 

However, that is precisely my concern "this point in time."  I am wondering what that "point in time" actually was, especially considering that when scrolling through the (actually very neat and informative) photo sets from each of the 29 cities where Sprint is relevant, one can't help but notice that some photos seem to be shot in spring while others seem to be shot in winter or summer.  

 

So I guess I would need to know if these were all photos the author uploaded during her tours of these cities as she conducted her tests (as would seem to be the case given her excessive use of hashtags) or were some of the photos taken by others and she simply uploaded them to reflect certain aspects of her trip?  If they weren't taken by her then I could understand the apparent vastly different points in time in which they appear to be taken.  But if she took them herself during testing then it would be concerning as it would indicate that these tests were not, in fact, all conducted at the same, all-important "point in time."

 

This is my concern.

[milan03]

"We tested mostly during business hours, during a period from May 1st to May 23rd, 2014; we visited different cities on different days. In each city, our drivers stopped in 10 to 12 locations for at least 15 minutes (five test cycles) each. We also collected data while moving, both within and between cities."

 

http://www.pcmag.com/article2/0,2817,2459188,00.asp

[S4GRU]

Edit: I've come to the conclusion this is the only page that needs to be viewed. http://www.pcmag.com/article2/0,2817,2459189,00.asp

 

I love the Tmo in Ohio part at the bottom of the page you linked.  This was something AJ predicted about Tmo in Cincy for a long time.  Glad to see someone else pick up on it.

 

Robert

[Steve Dean]

I think one part of the equation that I believe is missing with the speed tests that were done.  Verizon, AT&T and TMO are have caps set on the amount of data a person can use without being charged extra.

 

If you know that you are limited on the amount of data that can be used in a billing cycle, are you going to keep your wireless data turned on 24/7?  The answer is NO!!!  You will turn off your data and use it sparingly.

 

With Sprint, there is not a data cap.  You can leave it on all the time and have the same invoice.  So why disconnect?  I believe they should have taken into account how many much traffic is going through particular tower and factor it in.  I bet if you took the same traffic factors and divided the numbers equally amungst the carriers, the speed tests would be very similar.

[cletus]

Sprint would still be the slowest.

 

Anyways, this is a weird thread to be honest. We all know Sprint has the slowest network out of all 4 major carriers. I didn't think the subject was up for much debate. What I took away from the article is T-Mobile has had a meteoric rise in terms of speed and Sprint has marginally improved, 

 

I somewhat agree with WiWavelength in that the use cases for faster than 10Mbps mobile internet are pretty limited right now. My concern is that when this is the average that means there are way too many cases in which someone will have a connection slower than that which quickly gets into the realm of annoyingly slow. In Austin the results are pretty close to my experience in that the 3G network is abysmal at .8 Mbps and nearly half second latency with 4G quick for certain tasks but latency being an issue as well. For my personal use I find current speeds here to feel slower due to latency. I often have double the LTE latency of my coworkers T-Mobile phones and about 1/4 the speed.

 

TL:DR this is unsurprising and Sprint has fallen behind the performance of all other carriers.

[orangeblue]

Nobody noticed that Republic Wireless was ranked No. 2, which happens to be a Sprint MVNO. Okay you offload a lot to Wifi but still WTF.

[Destroyallcubes]

Not a car guy, are ya Jeremy.

I don't have to be a car guy to know that a 450+ HP car is a show off car or a Racing car. And racing cars you show off to others how fast you can go. But in the end no matter what you think 50mbps on smart phones is not what every day users can see. They see the small things like Facebook, news, web browsing, and YouTube. And all of which can be handles on a 5mbps connection.

[cletus]

Nobody noticed that Republic Wireless was ranked No. 2, which happens to be a Sprint MVNO. Okay you offload a lot to Wifi but still WTF.

That ranking was for customer satisfaction. I am on Ting which is also an MVNO and I would probably rate my actual experience with them 9/10. However, these MVNOs score highly because the network isn't their fault and I think their customers realize it.

[Tajustin95]

I understand the tests and where I'm coming from Speint is horrible with drop calls, lte speeds being only fast when one across a td site and loads of ecsfb issues still in my area. So most of the time stuck on slow 3G speeds. Sprint said that this year was a comeback by mid2014. And it has been mid2014 with nothing changed much. Even on my travels to LTE/Spark announced cities we're just on with still loads of dead spots and slow speeds indoors.

[jamisonshaw125]

I understand the tests and where I'm coming from Speint is horrible with drop calls, lte speeds being only fast when one across a td site and loads of ecsfb issues still in my area. So most of the time stuck on slow 3G speeds. Sprint said that this year was a comeback by mid2014. And it has been mid2014 with nothing changed much. Even on my travels to LTE/Spark announced cities we're just on with still loads of dead spots and slow speeds indoors.

eCSFB issues can clear up pretty fast. Your basing decision on a network that is faulty at the moment.

 

 

Sent from my iPhone using Tapatalk

[Tajustin95]

eCSFB issues can clear up pretty fast. Your basing decision on a network that is faulty at the moment.

 

 

Sent from my iPhone using Tapatalk

I am talking about how long it is taking for the network, and also I been having ecsfb issues for more than a couple months now.

[jamisonshaw125]

I am talking about how long it is taking for the network, and also I been having ecsfb issues for more than a couple months now.

I said they "can", not will. Some areas take longer than others. And if you are un satisfied, why not switch providers? Is that not an option or do you just prefer to stick it out.

 

 

Sent from my iPhone using Tapatalk

[Hurrcayne_dre]

HAHHAHA No Speed test ran on North Ave in Baltimore huh? 

[tybo31316]

I am talking about how long it is taking for the network, and also I been having ecsfb issues for more than a couple months now.

Sprint isn't the only carrier that has problems. They just don't have sites out there like s4gru to point them out to you.

 

Sent from my Nexus 5 using Tapatalk

[milan03]

I'm seeing lots of posts about "usefulness" of very high peak data rates on our smartphones, and without a doubt pulling 50Mbps to a single device definitely isn't necessary in order to achieve a great user experience. However, what many avoid to point out is that for a great user experience, capacity needs to be readily available on the network.

 

For instance, in case of Verizon's network, their initial 700MHz network performed very well up until about a year ago when it started slowing down significantly, which resulted in rapid deployment of AWS capacity layer. Today they seem to be the fastest once again. T-Mobile's strategy was to deploy as wide as possible regardless of relatively low amount of LTE subscribers on their network. They've had to rip and replace most of their sites as well. This obviously resulted in more available capacity and a solid user experience in LTE coverage areas. It all requires strategic planning and execution.

 

So once again, let's not avoid the purpose of these performance benchmarks, and that is to showcase network capacity, not whether excessively high data rates on smartphones are needed or not.

Speedtesting is just one of the ways for a consumer to understand network capacity and performance in any given location.

[LuisBOSS302]

I said they "can", not will. Some areas take longer than others. And if you are un satisfied, why not switch providers? Is that not an option or do you just prefer to stick it out.

Sent from my iPhone using Tapatalk

That's why some of us have other lines on other carriers. The amount of drop calls I get on my Sprint S5 everyday compared to my Nexus 5 on t mobile is pretty sad. Sometimes I can be on a fully upgraded area and my texts won't send or I won't receive then. It's not just something new that has happened since I got the S5. It has happened with couple of single bands lte phones too. As well with other people that I know that have Sprint.

 

Ecsfb issues in south San Jose were a nightmare. I wouldn't be surprised with the low market share Sprint has in the area. Due to the recent cluster , things have gotten a lot better.

 

On the other hand sprint's lte speeds are good enough with me. I wish I had 5-7 Mbps connection instead of 0.05-0.20 Mbps speeds...

 

Ps: people on this site should stop telling users to switch carriers. People are smart enough when to switch. A little rant shouldn't lead to automatically replying to switch carriers.

[utiz4321]

I'm seeing lots of posts about "usefulness" of very high peak data rates on our smartphones, and without a doubt pulling 50Mbps to a single device definitely isn't necessary in order to achieve a great user experience. However, what many avoid to point out is that for a great user experience, capacity needs to be readily available on the network.

 

For instance, in case of Verizon's network, their initial 700MHz network performed very well up until about a year ago when it started slowing down significantly, which resulted in rapid deployment of AWS capacity layer. Today they seem to be the fastest once again. T-Mobile's strategy was to deploy as wide as possible regardless of relatively low amount of LTE subscribers on their network. They've had to rip and replace most of their sites as well. This obviously resulted in more available capacity and a solid user experience in LTE coverage areas. It all requires strategic planning and execution.

 

So once again, let's not avoid the purpose of these performance benchmarks, and that is to showcase network capacity, not whether excessively high data rates on smartphones are needed or not.

Speedtesting is just one of the ways for a consumer to understand network capacity and performance in any given location.

It shows excess capacity not overall capacity.

[milan03]

It shows excess capacity not overall capacity.

You're joking, right?

[Nickel]

I could have sworn we've already discussed these kind of things before. Let's find out:

[JonnygATL]

"We tested mostly during business hours, during a period from May 1st to May 23rd, 2014; we visited different cities on different days. In each city, our drivers stopped in 10 to 12 locations for at least 15 minutes (five test cycles) each. We also collected data while moving, both within and between cities."

 

http://www.pcmag.com/article2/0,2817,2459188,00.asp

Crap. I dunno how I overlooked that. Thanks!