Network Vision/LTE - West Michigan Market (Grand Rapids/Lansing/Kalamazoo)

[nmartine3]

http://newsroom.sprint.com/news-releases/no-one-beats-sprint-for-call-and-text-performance-in-grand-rapids-and-lansing-michigan.htm

Sprint sure is rocking it these days.

[Shima]

Good times coming for Sprint. I have noticed that the Sprint network is continually getting better here in W. Michigan. More B41 coming online as time goes. Let's hope all of the financial problems Sprint faces can be worked out.

Edited May 22, 2015 by Shima

[mittenmitten]

I don't get, Verizon has almost a 10mbps lead on T-Mobile in download speed, with a similar upload and far superior reliability, yet T-Mobile beats them on the speed index? How the hell do they figure that?

 

Beside that, Sprint is now basically in a dead heat with AT&T here. This is really exciting. When B41 is denser, they're going to pull ahead in speed. I'm just blown away by the value proposition: fast downloads, the best calling and texting, great reliability, all for the lowest price of any carrier, AND with unlimited data. Now if they can buy out all the garbage 'all star wireless' stores, and make them corporate owned, they would do well here again.

[maxsilver]

I don't get, Verizon has almost a 10mbps lead on T-Mobile in download speed, with a similar upload and far superior reliability, yet T-Mobile beats them on the speed index? How the hell do they figure that?

 

I don't think they actually said that.

 

They said Verizon's median download was 31mbps, and that T-Mobile + Verizon's median upload was 18mbps. But that doesn't imply that Verizon leads T-Mobile by 10mbps on download speeds -- I don't think they ever actually said what T-Mobile's median download speed was at all.

 

 

The difference between T-Mobile and Verizon is only 0.2 points. Their data speeds are effectively identical.

[mittenmitten]

I can't post pictures for some reason, but if you open "how rootmetrics scores are calculated" it shows Verizon's median download was 31.26 (18.13 upload) while T-Mobile's was 24.06 (17.82 upload). Sprint's is 11.6 (6.34 upload) while AT&T's is 11.48 (5.90 upload).

 

Verizon has faster median downloads, faster uploads, and better data availability, by a large margin, yet they are tied..

 

Same thing for Lansing, and Flint, and im assuming all other markets. Verizon has faster speeds, yet T-Mobile flat out 'won' in their 'speed index'.

[maxsilver]

Oh, I see what your referring to.

 

I'm guessing it's a median vs mean vs average situation. Hard to know for sure without seeing the actual data.

 

Subjectively, Verizon probably should win #1 in Grand Rapids for data (their speeds are faster than the report claims. And Verizon has full backhaul for their AWS LTE here, T-Mobile doesn't). But (still subjectively) Sprint and AT&T are both noticeably slower than RootMetrics's report claims in Grand Rapids.

 

RootMetrics in general does very little indoor testing, so that might be part of the discrepancy. But otherwise I'm not sure why their report looks like that.

 

I'm not seeing anything here that seems worth disputing, particularly since I'm not aware of anyone who does more thorough testing than they do.

[mittenmitten]

There's 1 beefy tower on the north side of town, near Belknap park, that has reliable 50mbps B41, there's the one on McKay tower with B41, and there's one out in Wyoming near 32nd st by my work that tests over 30mbps regularly that shows on their map.

 

I'm guessing token B41 sites helped them eek out higher numbers. That's how it works though. T-Mobile had spotty coverage before B12. Sprint just went the opposite way, far reaching B26 first, now speedy high frequency.

 

As far as new coverage, I now get solid B41 all the way down Jefferson on my way to work. There's a smoke stack with antennas on it, can't remember the exact factory name, but now the Madison neighborhood where I always dropped coverage has solid B41.

 

It's getting better. A LOT better compared to what it was in GR.

 

 

I question the T-Mobile win because in the last report in Q2 2014, they had this to say about Verizon: "Notably, 33.4 mbps is the fastest median download speed Verizon has recorded in any market we've tested in 2014 to date". I get that a lot changes in half a year, but..I can't imagine T-Mobile can trump Verizon's fastest recorded speed in half a year. Especially while coming dead last in every other metric.

[maxsilver]

I'm guessing token B41 sites helped them eek out higher numbers. That's how it works though. 

It's getting better. A LOT better compared to what it was in GR.

Yep. "token B41 sites" is also how I would describe it. And they're great when your near one (Metro Health is a B41 site. 44th and Kenowa, and 28th and Breton are too). 

 

Service is getting better, absolutely. But I don't think service is "good" yet.  There's nowhere in the city that I would describe as "solid B41 coverage". They simply don't have enough site density right now.

 

In fact, I wouldn't describe GR as having "solid B25/B26 coverage" today either. There's still a couple of large chunks of Grand Rapids with no LTE data at all. (Wealthy Street Business District, Easttown / East Hills, SE Kentwood + Broadmoor, etc).

 

 

T-Mobile had spotty coverage before B12

 

T-Mobile didn't get reliable just because of B12.

 

They added a lot of new sites and hardware in West Michigan from MetroPCS.  A few specific examples are Knapps Corner and 44th and Kenowa (ex MetroPCS sites, converted). Along with areas like Muskegon / Grand Haven (in which 100% of market coverage is ex MetroPCS sites). 

 

And on top of all of that, they did a lot of rural / exurb conversions -- something like 85+% of their 2G is now converted to LTE. (including Zeeland, Hudsonville, Rockford, Sparta + Kent City, etc)

 

T-Mobile did with MetroPCS, what I was hoping Sprint would do with Nextel in GR. And it's paying off for them -- I believe these new sites + upgrades are the reason T-Mobile still ranks competitively with Verizon. If they hadn't just done all of this work, they would have scored much lower in RootMetrics report in every metric.

 

B12 is helpful, but it's not even fully deployed in GR yet. The new sites + upgrades are what's improving reliability more than anything else.

 

 

Especially while coming dead last in every other metric.

 

I wrote about this in another thread, but since it's happened again in this report, I'll restate it. 

 

T-Mobile did come "last" place in voice, but it's not really a reflection of their network quality/coverage. In fact, VoLTE + GSM on T-Mobile is 2x more reliable than Sprint's 1x800 voice is today, according to this RootMetrics report (in the dropped calls metric).

 

T-Mobile's "call blocked" rate is their only problem here, and it's very high. But that's not network coverage -- that's VoLTE call setup issues.  Should T-Mobile's score get penalized for this? Absolutely -- it's a real problem that impacts subscribers. But I don't think it's a reflection of their network -- the problem will get fixed even though no one physically touched any sites / radios / backhaul / etc.

[mittenmitten]

http://www.rootmetrics.com/us/rsr/lansing-mi/2015/1H

 

: (

[nexgencpu]

http://www.rootmetrics.com/us/rsr/lansing-mi/2015/1H

 

: (

Why the sad face? Looks to me like a very close race..

 

Plus more importantly....

 

"Since our testing in September, Sprint’s median download speed increased from 2.7 Mbps to 9.9 Mbps, and Sprint’s median upload speed increased from 2.9 Mbps to 5.1 Mbps."

[mittenmitten]

Why the sad face? Looks to me like a very close race..

 

Plus more importantly....

 

"Since our testing in September, Sprint’s median download speed increased from 2.7 Mbps to 9.9 Mbps, and Sprint’s median upload speed increased from 2.9 Mbps to 5.1 Mbps."

 

I guess you're right, but aren't they a Spark market?

[nexgencpu]

I guess you're right, but aren't they a Spark market?

Its an increase of 3x the performance, and spark is being deployed. But there is still a ton of work to be done...

[maxsilver]

I guess you're right, but aren't they a Spark market?

 

Yes, Lansing is an official Spark market

[mittenmitten]

Yes, Lansing is an official Spark market

 

so..isn't 9mbps median download terrible for a Spark market?

[maxsilver]

so..isn't 9mbps median download terrible for a Spark market?

 

For a Michigan Spark market, it's pretty typical.

 

In fact, it's actually better than average (Ann Arbor and Flint are both official Spark markets, and their median speeds were only 4-5mbps according to RootMetrics)

[lilotimz]

so..isn't 9mbps median download terrible for a Spark market?

Average speeds are pulled down by the slowest which are from band 25/26.

 

Sent from my Nexus 5

[nmartine3]

Grand Rapids got Spark coverage today.

[grapkoski]

Grand Rapids got Spark coverage today.

 

In real life or on a map? Either way, exciting!

[nmartine3]

In real life or on a map? Either way, exciting!

On the coverage map.

 

Sent from my LGLS990 using Tapatalk

[maxsilver]

There's one nice aspect of this that I like. The spark map is now somewhat useful in that most of the areas of the city that have no working data are presently labeled "Spark: Fair".

 

So, if you interpret the Spark areas of the map by reading :

 

- "Spark: Good = Some sort of working data. Might be anything between EVDO/B25/B26/B41" and

- "Spark: Fair = Effectively no usable data coverage. Mostly 1x, some unusable B26/EVDO",

 

then the coverage map as it is today is now (very roughly) accurate.

[S4GRU]

So, if you interpret the Spark areas of the map by reading :

 

- "Spark: Good = Some sort of working data. Might be anything between EVDO/B25/B26/B41" and

- "Spark: Fair = Effectively no usable data coverage. Mostly 1x, some unusable B26/EVDO",

 

then the coverage map as it is today is now (very roughly) accurate.

 

 

I ignore all Fair coverages from Sprint.  I do that on Verizon too.  And Tmo before they flattened their map.  That way I am never disappointed.

[nmartine3]

I ignore all Fair coverages from Sprint. I do that on Verizon too. And Tmo before they flattened their map. That way I am never disappointed.

I find it best to ignore carrier created maps all together. Sensorly is more reliable. FCC should enforce a strict customer based coverage map, completely crowd sourced.

 

Sent from my LGLS990 using Tapatalk

[S4GRU]

I find it best to ignore carrier created maps all together. Sensorly is more reliable. FCC should enforce a strict customer based coverage map, completely crowd sourced.

 

Sent from my LGLS990 using Tapatalk

 

Sensorly is good, but has limitations.  If no one has used Sensorly there before, nothing will be shown.  Also, if someone was using a high frequency band, or stuck on B25/B41, it could show no coverage or weak coverage, when in reality a low frequency band may actually be available.  

[nmartine3]

Sensorly is good, but has limitations. If no one has used Sensorly there before, nothing will be shown. Also, if someone was using a high frequency band, or stuck on B25/B41, it could show no coverage or weak coverage, when in reality a low frequency band may actually be available.

Forgot about that.

 

Sent from my LGLS990 using Tapatalk

[JossMan]

What would be really nice is some sort of mapping tool through the SprintZone app, detailing exactly which Sprint LTE bands are available as well as an average signal strength within a given area.