Marcelo Claure, Town Hall Meetings, New Family Share Pack Plan, Unlimited Individual Plan, Discussion Thread

[NYC126]

Sprint lost the Kansas city market because of weak upload speeds. They will continue to show solid performance across the country, but lose markets because of the weak upload on the TDD LTE band. Until we see Carrier aggregation on the uplink.

[nick]

http://newsroom.sprint.com/blogs/sprint-perspectives/sprint-improves-its-network-as-competition-heats-up-in-kansas-city.htm

 

Great job but they still lost the market on a whole somehow. I hope Legere does not go to twitter with the whole we beat you in your own city bs.

Im sure he will, he loves to bash sprint are every opportunity.  That said that is an impressive download increase although i would rather the speed test be removed, feel free to talk about max peak speeds but the screen shot seems unnecessary in an official newsroom release. just my opinion.  

[WiWavelength]

http://newsroom.sprint.com/blogs/sprint-perspectives/sprint-improves-its-network-as-competition-heats-up-in-kansas-city.htm

 

Great job but they still lost the market on a whole somehow.

 

And that is a RootMetrics mystery, since Sprint ranked ahead of T-Mobile in most metrics.  If this all came down to uplink speeds, that is bullshit.  People barely care about uplink speeds.  Moreover, T-Mobile median uplink speeds are higher than its median downlink speeds.  That network is becoming congested.

 

AJ

[Terrell352]

And that is a RootMetrics mystery, since Sprint ranked ahead of T-Mobile in most metrics. If this all came down to uplink speeds, that is bullshit. People barely care about uplink speeds. Moreover, T-Mobile median uplink speeds are higher than its median downlink speeds. That network is becoming congested.

 

AJ

Preach. This isn't the first time Sprint had the highest speeds but still lost somehow.

 

Sent from my SM-N920P using Tapatalk

[Ascertion]

And that is a RootMetrics mystery, since Sprint ranked ahead of T-Mobile in most metrics. If this all came down to uplink speeds, that is bullshit. People barely care about uplink speeds. Moreover, T-Mobile median uplink speeds are higher than its median downlink speeds. That network is becoming congested.

 

AJ

You're right. Sprint's speeds measured the highest on the download, whereas T-Mobile's was the highest on the upload. But T-Mobile scored a 95 on speed while Sprint scored a 91.3. Sprint's lower speeds were probably a lot worse than T-Mobile's, though. Maybe they score based on average and not median? (Even though they specifically mention median in the reports.)

 

Rootmetrics should keep track of % on LTE/3G/No Signal. I bet those numbers would be interesting to see, especially on T-Mobile.  Ping should also play a part in data performance.

[JossMan]

It's all about that downlink speed and Sprint has that treasure trove of BRS, RootMetrics can shove it.  

[nexgencpu]

It seems they measure the download+upload as one total number. They probably should not count this evenly considering how much more important the user experience is on the download side vs upload.

 

FDD vs TDD also comes into play on congested sites. Whereas TDD favors download (which makes a ton more sense) where as FDD's upload will almost always be on the top end of its limits even on congested sites with crap download.

 

I could see a situation where

 

Example "FDD 2mb download/15mb upload could beat TDD 15mb download/1mb upload according to rootmetrics."

 

If that truly is the case, they really need to reassess how they measure these test.

[avb]

Remember those are median download speeds, not average.  It's possible to be on b41 more than half the time and get more than 20mb/s and median speeds will reflect that.  If nearly the other half the time you're on b26 and getting 3-5mb/s it won't be reflected in those median speeds.

 

***I'm not saying that's the case here, just pointing out median is different from average.

[WiWavelength]

For the record, here is the link to the Kansas City 2015 2H report:

 

http://www.rootmetrics.com/us/rsr/kansas-city-mo/2015/2H

 

AJ

[Fraydog]

Root isn't the best testing methodology, but it's still way better than anything that is crowd sourced. Now should they be open to using the GS6 for all carriers and also weight things a little better and more transparently? Absolutely.

 

 

Sent from my SM-N920V using Tapatalk

[WiWavelength]

You're right. Sprint's speeds measured the highest on the download, whereas T-Mobile's was the highest on the upload. But T-Mobile scored a 95 on speed while Sprint scored a 91.3. Sprint's lower speeds were probably a lot worse than T-Mobile's, though. Maybe they score based on average and not median? (Even though they specifically mention median in the reports.)

Because of its 5 MHz FDD and 20 MHz TDD configurations, Sprint median uplink speeds probably are 10 Mbps or less. Uplink capacity, though, is significantly greater than that of T-Mobile.  Sprint has up to five LTE carriers on air.

 

In speed testing, Sprint also has the disadvantage of inherently highest market share in Kansas City.  But that did not keep it from winning the downlink speed race.

 

Rootmetrics should keep track of % on LTE/3G/No Signal. I bet those numbers would be interesting to see, especially on T-Mobile.  Ping should also play a part in data performance.

I would think that the Network Reliability metric -- in which Sprint beat T-Mobile -- would factor in those criteria.

 

AJ

[bmoses]

i've been noticing that a lot in the reports where sprint did really well yet didn't win as many awards. the only think i could come up with is rootmetrics is putting a lot of weight on the upload speed and so thats what is preventing sprint from wining the awards

[WiWavelength]

i've been noticing that a lot in the reports where sprint did really well yet didn't win as many awards. the only think i could come up with is rootmetrics is putting a lot of weight on the upload speed and so thats what is preventing sprint from wining the awards

 

Because it is Sprint.  Everyone knows that Sprint sucks.  Sprint cannot win anything.

 

 

AJ

[Terrell352]

i've been noticing that a lot in the reports where sprint did really well yet didn't win as many awards. the only think i could come up with is rootmetrics is putting a lot of weight on the upload speed and so thats what is preventing sprint from wining the awards

That or Sprint doesn't have enough triband sites up. There is still too many sites without it.

 

Sent from my SM-N920P using Tapatalk

[clbowens]

I think it's kinda funny that sprint under-exaggerates their coverage considering pretty much every one else completely over exaggerates theirs. It makes they map look really bad when compared to the others.

I'd rather they under-exaggerate than over-exaggerate.

[RAvirani]

I'd rather they under-exaggerate than over-exaggerate.

I'd rather they over exaggerate so that the coverage doesn't look as shitty next to Verizon's/Att's/T-Mobile in ads

[Ascertion]

I'd rather they over exaggerate so that the coverage doesn't look as shitty next to Verizon's/Att's/T-Mobile in ads

Actually, I've seen Sprint exaggerate their coverage much more than they've under-exaggerate it.  At least, these are my findings in my area.  Sprint's LTE just does not reach as far as the maps say it does.  Especially Spark coverage.

[cortney]

Root isn't the best testing methodology, but it's still way better than anything that is crowd sourced. Now should they be open to using the GS6 for all carriers and also weight things a little better and more transparently? Absolutely.

 

I can't disagree with the argument I've seen that they should've used Nexus 6s or something better for this quarter. Reusing S5s for AT&T and Verizon and two different phones for Sprint and T-Mobile isn't fully unbiased. All phones should be the same, and IMO I wish they'd use more than one e.g. including an iPhone for far more inclusive results.

 

They have a great platform but it feels like they can't go that extra mile. 

[sbolen]

The Verge: Sprint's companies are stealing T-Mobile's idea to mess with net neutrality

 

Just like T-Mobile did with music, and then video, Boost Mobile and Virgin Mobile have decided arbitrarily that music is a more important form of speech than everything else that travels over its network. That's clearly discriminatory behavior, even if consumers benefit from it in the short term. It appears companies are willing to follow T-Mobile's example, which sets a worrying trend considering FCC Chairman Tom Wheeler has praised similar schemes as being "highly innovative and highly competitive."

 

[derrph]

Oh well Papa John started it. In all seriousness though now we are starting to see people somewhat worry more about this. I guess you can say if one is doing it it's ok but we will be watching but If more start to do it it's going to become a problem.

 

 

Sent from my iPhone using Tapatalk

[Ascertion]

Oh well Papa John started it. In all seriousness though now we are starting to see people somewhat worry more about this. I guess you can say if one is doing it it's ok but we will be watching but If more start to do it it's going to become a problem.

 

 

Sent from my iPhone using Tapatalk

I predicted this would happen.  I fully expect it to get worse once the FCC supports it.  Even if it's against net-neutrality, ISPs and carriers will have to follow or they'll lose customers.

[Fraydog]

I hate this, but once the FCC OK'ed BingeOn and Music Freedom for T-Mobile, it greenlighted everyone copying those programs. My ultimate fear is that this precedent will be used in a court case by Verizon that will lead to Title II being thrown out altogether.

 

Sent from my SM-N920V using Tapatalk

[Dkoellerwx]

That or Sprint doesn't have enough triband sites up. There is still too many sites without it.

 

Sent from my SM-N920P using Tapatalk

 

In Kansas City at least, I can tell you from experience that this isn't the case. Essentially every site is either full Sprint tri-band, or co-located/right next to a Clear B41 site with 2xCA enabled. 

[Terrell352]

In Kansas City at least, I can tell you from experience that this isn't the case. Essentially every site is either full Sprint tri-band, or co-located/right next to a Clear B41 site with 2xCA enabled.

Well I guess we can rule that out at least for Kansas.

[Dkoellerwx]

Well I guess we can rule that out at least for Kansas.

 

And the speeds reflect that. If memory serves, I believe this is the first time that Sprint has had an average download greater than 20Mbps. The ONLY reason for them not to tie/win according to the report would be the upload speed, which is ridiculous.