I get the feeling this is just going support the ISP's current claim gamers and file sharers are causing this lag and that bandwidth caps and blocking certain file sharing protocols are necessary to perform as advertised.

I actually just tested my Comcast speed the other day, since they sent me a leaflet saying that they doubled the speed to 12mbit for no additional charge. It actually tested a bit faster than advertised, so I was happy. I have no doubt it has slowdowns from time to time, but overall, it has been very good.

I would really like to know where these numbers came from and how they were obtained. Not to say that I don't believe broadband companies are lying (they are) about speeds, but 50-80%? That seems high. It's worth noting that if you run a speed test from a computer connected to the modem via an ethernet cable, the speed reported will be accurate. If, however, the test is done over wifi, the results are often not representative of the actual speed of the modem. Further, the modem itself can play a part in lower-than-expected speeds. For example, I know someone who purchased a cable modem in order to get a deal on service. The service was 6mbps down, but the modem supplied was only capable of 3mbps. The service was actually as it was advertised, but the modem created a bottleneck.

I'm curious to see how all of this plays out.

@What The Geek: The real question is, do ISPs prioritize traffic to and from known speed test sites?

I know my speed is definitely slower than advertised, and I'm trying to find time to make an inquisitive, yet intimidating phone call about it to Cox. It's been so annoying dealing with speeds that are too slow to stream Hulu. Hulu!

Speaking as a network professional this kind of crap has gone on too long.

The only time I see major change is when network professionals like myself demystify the WHARGARBL (lies) that the ISP's attempt to sell their customers on in forums like DSLreports.com or some other forum similar to that.

It's like they think their customers think what they do is magic and it is definitely not. It's rather obviously that American ISP are more interested in leaving the back haul networks alone and charging more for less. And they keep coming up with new ways to make the "less" look like "more" with creative marketing. (read as MORE LIES.)

If these companies spent just 1/2 the money they do on marketing and applied the other 1/2 to upgrading their infrastructure they wouldn't even have to lie. There would be no need for caps or meters and customers would be happy. On top of that, the US would also not be scored as ridiculously bad as it does in the world rankings.

50-80% lower some of the time and somewhat lower 50-80% of the time are very different things. It seems the title and article disagree on which is being reported.

To quote the article, "...by as much as 50% to 80%." I think those of us who have heard ads promising UP TO anything, realize it's a way to sensationalize.

A great big ol' DUH!! goes out to this post. Many many people have been saying that you do not get what you pay for when it comes to internet speeds. There are any number of reasons why you cannot get the speed you pay for. Namely how fast the other persons upload speed is, how much load there currently is on the server that the website is located on, and if there are any current problems with the net in general. All these play major factors into your download speed. I personally do not see this going anywhere.

@What The Geek:

The article is pretty poorly written. It does state that speeds lagged 50-80%, but then has language to the effect that 50-80% of THE TIME speeds were less.

I have a 3mbit connection. 50% lower speed means it's 1.5mb. 50% of the time means that 15 days a month my speed is less than 3mbit. If they mean OF THE TIME, then getting 2.9mbit 15 days of a month means that 50% of the time time I'm not getting what I pay for.

I get like 0.8mb download at night versus 9mb in the morning. It's really terrible. I have Time Warner/Earthlink.

@What The Geek: Your observations about wi-fi are probably correct, and since many people are using laptops now, I wouldn't be suprised if this is one reason people experience lower speeds. However, I know that my personal internet can experience good and bad periods. I think it averages out overall.

Also people may be misunderstanding that though during a speed test they get 5mbits downstream, they aren't getting download speeds like that.

My Internet cable company Shaw Cable is currently running ads how its cable is "faster then before" it then goes into this dragged out diatribe with numerous zany analogies how its so much better. In the amount of time they were going on and on about analogies about how much faster it is they could have given me actual numbers 10 times by now. I'm serious, they must think we're all idiots.

@Oranges w/ Cheese ontopofBrocolli: actually I think the bigger problem is that people don't understand the difference between a bit and a byte.

@wrjohnston19283: whether it be "of the time" or "of the max speed" there could be a lot of factors effecting the results - I'd really like to know specifically what these numbers mean, and how they were obtained.

@milkcake: Maybe everyone else near you is on the interwebtubez downloading at the same time?

A few years ago I worked at a call center that had a contract for Comcast HSI Support, aside from being a horrible place to work to begin with, I was constantly told to remind customers who complained about either slow internet speeds or customers who just plain had no connection at all that if they read their contract that Comcast actually guaranteed a speed of 0.

Whenever I would ask a question about connection speeds being "less than optimal" they would blankly stare at me and repeat "Comcast guarantees a speed of zero" and walk away.

I didnt work there for very much longer after I noticed the pattern.

@axiomatic:

Frankly, the fact that you think that consumer broadband bottlenecks are in the backhaul network makes me question whether you are a network professional.

I guess I'm lucky to have an ISP that I don't want to murder. At peak hours, Verizon is giving me 14-15 of my 15mbps. Off hours, I sometimes get up to 20mbps.

@axiomatic: It's even worse in Canada my friend, at least in Calgary. Way back in the day we were on of the first areas to get cable internet access and ever since then we have been slipping and falling down the ranking to where we're near the bottom of the list especially when compared to US ISP's. Not my comment below how they are running a new radio campaign saying how "blazingly fast" they are but never give actual numbers. Well I know the truth and pretty much answers the question why they want an uninformed consumer population.

@Révolution:

Mine is kinda the same way, Although I only have 2Mbps(Yeah for underground old phone wires!) but pretty much all day I can hit my limit even if its peak time or off hours.

Now if only I could get their 5Mbps or 10Mbps :(.

@milkcake:
I have that same problem with Time Warner, but, it doesn't really correlate to any time frame, but, one time for several hours my download speed was hovering at 300k. And for the most part I get terrible download speeds, and shockingly bad upload. I also am pretty sure they are throttling traffic, especially to iTunes, however, I doubt they would even believe me
But, they are the only game in town where I live, and it sucks.

I suppose it depends on where you live. Ironically I've had both Comcast and Verizon FIOS and have always gotten the maximum speed advertised through my Usenet feed (speed tests have slowly become irrelevant IMHO as providers have slowly been 'skewing' traffic to make those tests appear good, while actual usage can never actually sustain the speed)

@Kamidari: Actually a lot of the cable companies are advertising some sort of "speed burst" which is a temporary speed boost for maybe the first 10 seconds. I can get as high as 20Mb/s (2500kB/s) when a download first starts but then it usually slips down to around 9Mb/s (1200kB/s) which is slower than the advertised 12Mb/s (1500kB/s). I get around 15Mb/s from the speed testing sites.

I can absolutely confirm this.

We have two Internet connections at the house - DSL from AT&T, and cable from Comcast - and I check the throughput occasionally using the usual speed test sites.

AT&T's DSL is just pathetic. The bottom speed for our tier is supposed to be 192 kbps symmetrical. But I just tested it, and it's actually running at 50 kbps symmetrical, which is 26% of what we're paying for. You can get dial-up modems that are faster than that. We have it because the connection is more stable than Comcast, so we use it for some servers that can't go down.

Comcast is nominally 6 Mbps download, or 12 with "speed boost". What that means is they kick up the rate for the start of a download and throttle it back. That's supposed to improve your movie-watching experience or something. But it also has the interesting side effect of improving their results on the speed test sites, whose tests run within the "speed boost" time window. I routinely download multi-hundred-megabyte files from my employer, and the actual throughput is often far less. But I don't have hard numbers on that readily at hand.

I would love to get an "enterprise" quality connection with guaranteed throughput, and a "service level agreement" that guarantees an uptime percentage and fixes within a few hours. Comcast has no throughput guarantee, and their notion of a "service level agreement" for business customers is "we'll roll a truck within 24 hours of when your connection goes down." Not exactly what you want for a mission-critical business system.

This is more of a problem with cable during peak hours or cable with neighbors who torrent shit all day.

I've gotten accurate and consistent speeds with DSL.

@What The Geek: As mentioned below, I'm getting 26% of the throughput I'm paying AT&T for. This was tested over 100-base-T Ethernet connection from an iMac to the cable modem, with no other computers on the network.

@Kamidari: What time during the day or night you tested it, how many people in your neighborhood use the same cable, and how far you are from the dispatch or routing center all factor in as well. Try running the same test after school is out at around 3pm and try at around 6 when people are home from work. Then compare that to a later time at night to see if you are actually getting that speed all of the time. The problem I see is that the cable companies advertise highway mileage if as though that's how the vehicle always performs.

This is accurate for Charter Comm at my house. I pay for Charter's 20Mbps service, which usually gets the 15 - 18 Mbps route. I don't complain, 15 - 18 is plenty usually.. but it's def. less than what's advertised.

@bobert: Back when I was still on Comcrap, I was only getting about 12mbps out of my 20 mbps plan from them. Comcast just sucks. They told me they only guarantee 60% of what you pay for. :\

Honestly, 50-80% of the time is significant enough for me to consider this fraud.

If you can't deliver what you promise even 50% of the time, you should be promising a lot less.

Aren't there laws about this kind of thing? I mean, isn't there a regulation that would state that if an advertised promise can not be kept more than X% of the time (15, 25, even 30) that it is considered false advertising?

I just got into a argument with a salesman at Office Max over this... He's got it in his head that Verizon is going to waltz into a ILEC run by AT&T and sell FIOS.

And run 3Gbt speeds too.

Heh.

I set his rudder straight on how telecom business are run and who can do what and still make a profit, especially in the field of broadband.

And, occasionally, they're faster than advertised. Says me who "accidentally" ended up with a 6-mbit ADSL line while signed up for 5-mbit service. :)

@MichaelBrazell: I wonder if they would complain if you paid them 20% less on their bill?

My provider, Bresnan, can only seem to offer me 2Mbps of the advertised 8Mbps. Using speed tests on their network I only attain 2 to 2.5Mbps. I have lived in several areas of Wyoming and have always had this issue with Bresnan. It's a shame I don't live near Powell... Bresnan tries to blame wiring as the issue, I keep telling them everything is brand new and you are the ones that crimped my cable. When you compare Qwest as the alternative, it's the lesser of two evils.

@pecan 3.14159265: I too have Cox and I am told I get "high-speed" economy, purportedly 1.5mbps, which I often don't get. And even at that slow speed, I still get charged $30 a month. It's a ripoff. That speed I can watch Hulu but still it's pretty darn slow and I think $19 a month max is a fair price. Seriously, they already have all the infrastructure installed.

@madog: It was reasonably prime-time - a little while after I got home from work in the evening, and had checked the mail. I have no idea about the other factors, though... I don't plan to poll my neighbors. ;)

@anonymousryan: Yeah, I think speedboost is why it ended up higher than expected, but I don't know of a good way to factor that out.

Maybe I'll try snagging a large file from EasyNews this evening to see how it goes. It was usually pretty close to advertised speed when I had 6mbit cable (800kB/s), but I haven't downloaded anything since the speed increase.

@NightSteel:

I've often wondered that. I've gone to those sites a few times and tested my connection speed. The result is always a hell of a lot faster than anything I ever experience normally.

I recently moved to a new house, and our only option for internet is 3Mbps At&t. At first it was ok, it stayed around 2-2.5... then for about 3 weeks straight it fluctuated between 200-500bps not even Kilobits mind you. finally after 3 hours on the phone with Je~ff they scheduled a tech to come out. The Tech was out the next day on-time mind you, and "fixed" the problem within 20minutes(side note, the tech said himself he didn't know how we were ever getting internet because nothing was hooked up right at the pole). I say fixed in quotes because a week later it was back down to 1Mbps to about 1.2Mbps, so I called again. This time I was told that was within "targeted" speed for my package and area. I have found if I do a hard reset on the modem, it will get 1500-1800Kbps long enough to watch some Netflix or play some COD, but as soon as I'm done, it's back down to 1Mbit.

According to the press release from the FCC (linked in the Washington Post article), the services aren't slower 50-80% *of the time*, but rather that "actual broadband speeds lag advertised speeds by as much as 50% to 80%." It goes on to point out that high traffic during peak times (cited as 7 p.m. to 10 p.m.) create congestion, causing the loss of speed. So, while this is still appalling, there's a significant difference in what the FCC actually reported and what the Washington Post article stated ("[M]ore than half the time, and sometimes as much as eight out of ten times, consumers are paying for slower Internet access speed....").
IMO, poor reporting by the Washington Post.

@Hooray4Zoidberg: Those are still customers and while Comcast is certainly not above tearing them down, I think they would prefer to target the third party services first.

eg 'It's not our fault our customers can't get 5MB/s download speeds at website.com, they just can't send you data that fast. Because they didn't get Comcast."

I have had Cox Cable, ATT DSL, Comcast Cable, Verizon DSL, and I always had the speeds that were promised (minus any overhead) except with Insight Cable who oversold their nodes. Except for Insight I never noticed any slow downs at any "peak" times and I used to download 24/7 as much as I could.

This study has to be BS.

How about having an independent agency (government or private) that establishes a standardized method for calculating the average connection speed per each tier offered by an ISP. Then require that ISPs can only advertise speeds obtained using these methods.

And there you have it! A real-world, apples-to-apples comparable value a consumer can use to go shopping for an ISP. Yes there will need to be a margin of error for this, but as long as the same methodology was used throughout, relative comparisons would at least be more accurate than they are now.

@axiomatic: Well there are three very good reasons the USA lags behind other countries in terms of broadband penetration. 1) It's really big. 2) There's already an extensive legacy infrastructure to upgrade, which is a lot harder to do than simply building out a new one. 3) This stuff ain't free.

@satoru: It depends on where you live, who your provider is, and in some cases, what your neighbors are doing on their connections. I use Verizon FIOS and Clear WiMax and both services have always delivered 95-100% of their advertised maximum rates.

People don't realize that it's not the ISP fault all the time, maybe they do cap your bandwidth, but it's also on the other end.
It goes according to how many people are on any server at any giving time. Some companies cap they bandwidth on their server so people don't use more then anyone else, and the more people on that one server, the slower it's gonna be.

When you downloading something, even if you're are just reading online, all those pages that you go to, you still downloading something.

Downloads really has nothing to do with it, you got take a look at the servers upload speeds. If you on a 10Mbs connection, and their server only has a 1.5Mbs upload. You think you are gonna be downloading at 10Mbs? Of the 1.5Mbs, that could possibly be capped at say like 100k. And the more people on there, the slower it can get.

Also if you're on cable, you're sharing your bandwidth with your neighbors, so if they're doing some downloading, that would also impact your speeds.

@Oranges w/ Cheese ontopofBrocolli: Taking it to the other extreme on WiFi, what happens when you have a 25MBbs or 100Mbs download connection (Verizon Fios is offering this). In the real world, I think you would be hard pressed to find a webserver that is connected to that large a pipe where you would ever see such speeds (Amazon and Google may be the exception), so when you hit a web server with a 100Mbs pipe they are going to trickle the data to you at a few Mbs upload speed.

I personally have a 25Mbs down/15Mbs up from Verizon that I recently upgraded from a 5Mbs down provision (part of a tripple play package...sure, I'll pay less money for faster service). I have noticed no significant difference in speed in my daily surfing, a moderate increase in speed when downloading files (nowhere near 5 times as fast), but I have seen a huge improvement in upload speeds. When I go to speed testing sites, I am usually at 23Mbs down and 14.5Mbs up.

@liquidnumb: Except that countries such as Korea have a huge broadband market at speeds well above ours and they also have the largest online gaming market in the world. WoW is a small fry when Korean games are thrown in the mix. Those titles are 99% ignored when the Western world talks subscriber base for games.

@Kamidari:

That is a temporary speed. It slows down over time, and the power boost only lasts like 30 seconds anyhow.

Most of the ads I see avoid mentioning actual speed other than saying it's "blazing fast" and how it's better than dial-up (?!). I bet half of subscribers don't even know what speed they have. If they did they'd probably realize that they are paying for more internet than they need.

@Telekinesis123:

That happens to early adopters. Internet speeds in Japan sucked until their ISPs got their shit together and stopped trying to sell ISDN (which they'd deployed really heavily early on).