 Tairei
join:2009-07-01
Beaverton, OR
2 edits | Book Reports, Science Reports, Etc It's all about fluff when you're writing reports, isn't it?
Need 2500 words? You know at least a quarter of that paper is useless fluff. Even in college you see it. | |
|
 |  tshirtPremium,MVM
join:2004-07-11
Snohomish, WA
kudos:3
 ·Comcast
| Re: Book Reports, Science Reports, Etc said by Tairei:It's all about fluff when you're writing reports, isn't it? I somewhat agree, gotta have a big headline to sell those $2,000  reports | |
|
| |
| | |
 KrKHeavy Artillery For The Little GuyPremium
join:2000-01-17
Tulsa, OK | They're wrong. No way the "Average" is faster then the best DSL speed of 6.0 mbits or so.
3.9 sounds MUCH more accurate. | |
|
|  BF69Premium
join:2004-07-28
Camden, TN | Re: They're wrong. said by KrK:No way the "Average" is faster then the best DSL speed of 6.0 mbits or so. 3.9 sounds MUCH more accurate. Could be right since they are using AVERAGE and not MEDIAN which is much more accurate. For example 4 houses; 3 with 6 Mbps DSL one with 20 Mbps cable. AVERAGE speed 9.5 Mbps. Even though 75% of those houses can't even get close to that speed. MEDIAN speed would be 6 Mbps which is more accurate. | |
|
| | KrKHeavy Artillery For The Little GuyPremium
join:2000-01-17
Tulsa, OK Reviews:
 ·AT&T DSL Service
| Re: They're wrong. Yes, but even then, seems unlikely because the vast numbers of DSL lines out there. And most DSL lines don't make it all the way to 6.0 anyway--- and then there's the issue of how many people are actually maxxing out their connections. Very few.
--
"Fascism should more properly be called corporatism because it is the merger of state and corporate power." -- Benito Mussolini
| |
|
| n2ubp
join:2007-07-13
Middletown, NY | "Average" does not reflect the real world.
What is the "MEAN" ? | |
|
| |  ThrowDemsOutIf you can't convince 'em, confuse 'emPremium
join:2002-03-03
Mullica Hill, NJ
kudos:4
2 edits | Re: They're wrong. said by n2ubp:"Average" does not reflect the real world. What is the "MEAN" ? »cboard.cprogramming.com/brief-hi···ean.html
Average - the numerical result obtained by dividing the sum of two or more quantities by the number of quantities; an arithmetical mean.
Mean - halfway between extremes; in a middle or intermediate position...a number between the smallest and largest values of a set of quantities, obtained by some prescribed method.
Median: »davidmlane.com/hyperstat/A27533.html
The median is the middle of a distribution: half the scores are above the median and half are below the median. The median is less sensitive to extreme scores than the mean and this makes it a better measure than the mean for highly skewed distributions. | |
|
| | | | | Re: They're wrong. You left out the other measure of central tendency: the mode. In statistics, the mode is the value that occurs the most frequently in a data set or a probability distribution. | |
|
| | | | Tairei
join:2009-07-01
Beaverton, OR | Re: They're wrong. Yeah but everybody still thinks of average being the only relevant one. Ask random people on the street what the mode of a given set of numbers is and they'll likely either quote the average or stare at you all confused. | |
|
| | | |  jester121Premium
join:2003-08-09
Lake Zurich, IL
 ·voip.ms
| said by jjeffeory:You left out the other measure of central tendency: the mode. In statistics, the mode is the value that occurs the most frequently in a data set or a probability distribution. There's a small hairy German man screaming something about "standard deviations" in my brain but I think that's just a flashback to statistics class in college.... the alcohol makes him go away. | |
|
| |
| | | | Re: They're wrong. Caps do not affect speed in any way, shape, or form.
--
Jay: What the @#$% is the internet??? | |
|
| | | |
| | | | | | Re: They're wrong. Your reasoning is flawed. Specifically, your statement that the only accurate way to gauge their network speeds is to adjust for caps. By this reasoning, a provider that doesn't have caps will NEVER be subject to congestion or slower than advertised throughput, and that simply isn't true. Networks are not designed to allow all users to utilize all of their bandwidth at the same time.
Conversley, there are networks with absurdly low caps. These low caps are often not in place because the network cannot handle the speed; instead it is to enable overage charging to augment the providers bottom line.
--
Jay: What the @#$% is the internet??? | |
|
| | | | | |
tshirtPremium,MVM
join:2004-07-11
Snohomish, WA
kudos:3 Reviews:
·Comcast
| Progress, don't stop now I think the 2 surveys are measuring different things OR reporting the same thing differently.
Instat is looking at BROADBAND speeds (ignoring dialup very low speed connection.
Akamai looks at ALL connections which drops the average quite a bit.
The rate of INCREASE is not the same as higher per media.
Cable HSI got a new big bump with D3 deployment, FTTH started higher but hasn't doubled (or more) it's speeds.
Remember this isn't highest available speed, but average speed tier that sells, ComCast automaticlly doubled most users speeds, and offers a 8.5 times faster connection to many (6/1>>>50/10) which could greatly skew the results Depending on who/how/ how many, connections/consumers were polled. | |
|
BF69Premium
join:2004-07-28
Camden, TN | Who cares about speed About time they concentrated on getting more people access to broadband that trying to increase speeds. At the end of the day that's where the money is going to be. Going from 5 Mbps to 10 Mbps, sure I can tell a difference. Going from 20 Mbps to 25 Mbps yeah not so much. | |
|
|  en102Canadian, eh?
join:2001-01-26
Valencia, CA | Re: Who cares about speed Exactly. I could run a speed test once, get 20Mbps, run again and hit 30Mbps... I wouldn't notice the difference, even though one has 50% more speed, and the difference is as fast as AT&T's fastest ADSL package.
--
Canada = Hollywood North | |
|
| | BF69Premium
join:2004-07-28
Camden, TN
1 edit | Re: Who cares about speed said by en102:Exactly. I could run a speed test once, get 20Mbps, run again and hit 30Mbps... I wouldn't notice the difference, even though one has 50% more speed, and the difference is as fast as AT&T's fastest ADSL package. Well my point was that if you were downloading say a 2 GB movie from Itunes or Amazon then at 5 Mbps that would take about 54 minutes. At 10 Mbps that would take 27 minutes. Ok so you'll notice 27 minutes difference. With a 20 Mbps conenction it would take 13.5 minutes. With 25 Mbps it will take 11 minutes. You're not going to notice 2.5 minutes. | |
|
| | | en102Canadian, eh?
join:2001-01-26
Valencia, CA | Re: Who cares about speed Exactly. Unless I'm consistantly pulling something that needs the full limit (eg. a stream that 'needs' 25Mbps and would buffer at 20Mbps), to me, anything above 10Mbps isn't noticed.
This is TWC Docsis 1.1 on the 'All the best' 10/1Mbps package (free boost somehow gives me up to 35Mbps).
--
Canada = Hollywood North | |
|
| | | | fiberguyMy views are my own.Premium
join:2005-05-20
kudos:3 | Re: Who cares about speed I dunno about you, but I notice my 30 meg speed over 10... when every I'm having to download software like iTunes, real, etc. I'm loving being able to get it done in a few seconds vs minutes.. but for web surfing.. sure.. no difference.
However, I'm quite surprised you're saying this especially when everyone's pushing for more speeds so they can do more online, such as video, Netflix downloads, etc. | |
|
| | | | | BF69Premium
join:2004-07-28
Camden, TN | Re: Who cares about speed said by fiberguy:I dunno about you, but I notice my 30 meg speed over 10... when every I'm having to download software like iTunes, real, etc. I'm loving being able to get it done in a few seconds vs minutes.. but for web surfing.. sure.. no difference. However, I'm quite surprised you're saying this especially when everyone's pushing for more speeds so they can do more online, such as video, Netflix downloads, etc. Was I comparing 30 to 10? I was comparing 5 to 10 and 20 to 25. See in each case a differnce of 5 Mbps. You can tell a difernc between 5 and 10. Not between 20 and 25. So as I said there's little point to continue to push speeds even higher while many areas don't even have broadband.
Need for MORE access > need for FASTER access | |
|
| | | | | | en102Canadian, eh?
join:2001-01-26
Valencia, CA | Re: Who cares about speed A good part of that is the relative difference. An extra 25% for basic surfing is not noticable at 20Mbps. At 5Mbps, 10Mbps is a 100% increase which may or may not be noticed for basic surfing.
In my case, work related, SSH / remote desktop hasn't seemed much different on TWC at 20-30Mbps than it was on 3Mbps/512kbps ADSL. Latency was actually lower on ADSL (fastpath) with DSL Extreme than it is on TWC and more consistent.
For large downloads, and downloading while running Skype, having sufficient bandwidth (both ways) is a good thing though. | |
|
| Sammer
join:2005-12-22
Canonsburg, PA | The reason why speed is important is the more people who have at least 40 Mbps / 10 Mbps the more pressure there will be on lame providers to get access to true broadband speeds to those that don't have access to true broadband speeds. | |
|
iansltx
join:2007-02-19
Golden, CO
kudos:2 Reviews:
·Comcast
·Verizon Online DSL
| Wonder where they're getting their data InStat's numbers are suspiciously close to those noted by Speedtest.net, which lists the US as ~7.5 Mbps down and ~1.7 Mbps up. The thing with Speedtest.net is there's a significant upward bias to the results:
1) PowerBoost (downloads mostly, uploads some)
2) People with hihger-than-average connection tiers are the ones most likely to be using Speedtest.net
For even more fun check out the results for Fredericksburg, TX. The RR business results have got to be somewhat padded by the city's own 20M symmetric fiber connection, since TWC doesn't offer PowerBoost on uploads and RR business tops out at a little under 2 Mbps. Also, either someone has three or four bonded Qwest-port T1s or the local WISP is using speedtest.net to test their backhauls or uncapped equipment, whether to inflate their score or not I'm unsure.
Anyway, I agree that the data could mean any number of things; if they have a decent sample size of non-tech-savvy users then I'd give credibility to the report more than if they set up some speedtest site that nobody really goes to. Also, as others said averages are easily biased by a relative handful of super high speed connections (40-60M DOCSIS or fiber, or even 100M fiber). Quartile distribution would be much more useful. As the saying goes, there are lies, damn lies and statistics.
As for the DOCSIS speed increases by 100% versus less on fiber, as others have said cable was lower to begin with, and fiber higher. Verizon went from 10/2 this time last year to 15/5 as their base FiOS tier, and from 20/20 to 25/25 for their midrange to higher end tier. In the same time, Comcast went from 6/1 and 8/2 to 12/2 and 16/2, and added 22/5 and 50/10 tiers. What's really interesting is that now Comcast's standalone price on 12/2, including modem rental, is around $60 per month, $5 more expensive than FiOS 15/5. Comcast's 16/2 tier costs the same with modem rental as FiOS 25/25, and it's a valid price comparison; Verizon gives you all the equipment you need to connect.
Another thing to note: other than Comcast no cable provider has drastically increased speeds in the wake of DOCSIS 3, though if Instat is getting PowerBoosted numbers TWC may well have kicked things up a notch due to inclusion of PowerBoost on their Standard package in many areas, versus a year ago.
I'm betting that the 50th percentile on speeds is in the standard cable/DSL range, between 3 and 7 Mbps. But I'm not going to pay two grand for the report to find out. | |
|
|  NerdtalkerWorking Hard, Or Hardly Working?Premium,MVM
join:2003-02-18
Tucson, AZ | Re: Wonder where they're getting their data Exactly, couldn't have said it better myself.
There's a definite statistically advantage approach to using PowerBoost, both in user perception and benchmark speed. I've noticed that they've almost perfectly tweaked the boost to coencide almost exactly with Speedtest.net over the course of the whole roll-outs.
I mean, does this:
Really mean anything anymore?
--
"Some people never see the light till it shines thru bullet holes." -Bruce Cockburn
I'm testing Gmail's spam filters: Broadbandreports1@gmail.com
Spam: 12900+ messages currently using 406 MB. | |
|
 VanPremium
join:2009-07-08
New Orleans, LA | Sad that I am below the average Sigh....DSL blows | |
|
| | SPEEDS up 28% CAPCITY DOWN 1000% thanks to caps and throttling
YUP cant wait for the one gigabit /sec connection and the 1GB CAP thas throttled to 5 Kbytes/sec 99% of the day | |
|
| BF69Premium
join:2004-07-28
Camden, TN | Re: SPEEDS up 28% CAPCITY DOWN 1000% Think of it this way a dial-up connection can use 16.5 GB a month. So anyone offering broadband better be offering SIGNIFICANTLY more than that.Charter's 10 Mbps tier has 100 GB monthly cap. So in otherwords only 6 times more than dial-up. Might as well just offer a 384 kbps connection and no cap. But wait no one is going to pay $55 a month for 384 kbps. | |
|
|  dvd536as Mr. Pink as they comePremium
join:2001-04-27
Phoenix, AZ
kudos:4 | nothing wrong with capacity.
caps are all about preventing cord cutting via internet video!
--
When I gez aju zavateh na nalechoo more new yonooz tonigh molinigh - Ken Lee | |
|
dvd536as Mr. Pink as they comePremium
join:2001-04-27
Phoenix, AZ
kudos:4 | Want to see an even more pathetic report? Do one for UPLOAD! | |
|
2 edits | Vetting the data You could also use the data DSLR has in /archive to vet these reports. Granted the user audience may be different, but the weighted average (# tests * average speed) shows about 5.5Mbps in the US with 1.5Mbps in the upload.
If you break it down further between the cable and telco icons you get
Cable @ 6.7M/1.6M
Telco @ 3.5M/1.3M*
* Most telcos are low on the upstream with VZ holding this average up with 5M down and 3M up. | |
|
 joebarnhartPaxio evangelist
join:2005-12-15
Santa Clara, CA | Darn... they didn't include my connection in their survey Or did they? Naaa... Otherwise they wouldn't have slammed FTTH so hard.
 | |
|
| | Yay! I'm average! My blazing fast Verizon connection is right at the average then!! | |
|
| | Just a note in passing... In-Stat is currently a wholly owned subsidiary of Reed Business Information, who was also (until VERY recently) the publisher of Broadcasting & Cable, Multichannel News, and Twice (consumer-electronic magazine for the trade). So, U.S. broadband speeds are faster? Well, OF COURSE they are. | |
|
|
|
·
·
