Gaming on dsl and cable

[konakona]

I happen to be an avid gamer that spends most of the time on the internet playing games. The condition I am in require a bit of a special treatment, as the game servers are located a few tens of thounsand miles away, every ms of ping counts; for quite some time I have been pinging my game server (whose exact address is known and pingable) on my secondary rig as I play on my main one. For instance, I can surely tell the difference between 200ms avg and 230ms avg.

I realize latency depends on many different things and it is probably near impossible to make a sweeping generaliztion that applies to every case. Nonetheless, there should be some sort of consensus or practical consideration of which works better ultimately for the most part.

What little discussion I have seen through google seemed to favor DSL for the most part, which coincides with my personal experience. I have moved spent my last 8 years or so in different parts of FL, GA and TX, switching between a few ISPs 7-8 times from what I can remember. Never have I seen cable in my area (comcast/RR/TW) give me better ping than DSL, though at times it was close. The real issue I have observed and had suffered from is how nodes get totally hammered at peak hours when people return from school and work.

The one I just bailed out of, for instance, drops a packet in every dozen or so with the avg ping about 30-50ms higher to boot. When I confronted TW about this grave concern, they reluctantly dispatched a technician after several tiresome phone calls. Ironically the technician came at an off-peak hour and found no problem at all. They told me there is nothing more they could possibly do about my situation, implying that I can online just fine so I should STFU and be happy.

Being fed up with the obvious sign of oversubscription and the ISPs unwillingness to improve the situation, I jumped ship on them and got a dsl instead. Even with the cheapest one available (384k/1.5m) I am getting some 30-40ms healthy reduction in ping and no packet losses to speak of. Hell yeah

My opinion could be a bit skewed, since I have mostly lived in relatively cheap and poor housing with a bunch of bandwidth hungry students. There is no denying there are many, many happy gamers on a cable, perhaps with a better provider than what TW is. I merely wanted to see if it is possible at all to draw some conclusion on the bigger picture with all things considered. Cable does offer you higher claimed bandwidth at a given price, but that does little to gaming latency. Is it a norm for cable ISPs to oversubscribe like crazy, or I was just super unlucky to fall victim to that three times in a row?

 

[spidey07]

It all depends on the provider. DSL is also oversubscribed. As a technology Cable should provide better service because of the higher bandwidth means less latency/delay.

 

[JackMDS]

There are Bad Gamers, Good Gamer, and Excellent Gamer.

There is No difference between them because they are on DSL or Cable.

Unless a service (DSL or Cable) is really "Quirky" the differences is usually in the mind of the Gamer and mainly involved psychological over compensation for the Gamer dissatisfaction with his status.

There are much difference between people?s Physical and Cognitive systems and that is what really account for the differences between Gamers.

Normal reaction of people is between 180msec. to 300msec. (bigger numbers means slower reaction time), even much more slower under Alcohol, being tired, and similar states.

The undisciplined Gamers that are on the 300msec. and more side, are usually the ones that blame the ISP.

Objectively if it is so imported to you, you should subscribe for one month to both services and choose the one that works better for you at your specific location.

While at it run a trace on each connection of the condition of each oe.

This is free, http://www.d3tr.de/download.html

The following Screen shot is taken by using the List feature of 3DTrace.

http://www.ezlan.net/network/trace.jpg

The table shows the result of multi ping from my computer through a Buffalo Router to Internet site 96.6.68.195*

The important columns in the table (for this discussion) are the two in the middle. I.e. Min. ms and Max ms.

The numbers there show a typical outcome. As the trace progress on the Internet the variation changes dramatically.

So while the variation on my Router is 1ms, and the connection to the direct ISP's server is 2ms, the variation later on is almost 20ms.

 

[drebo]

I would hardly call online lag during gaming a "grave concern".

Basically, any difference in connection between you and a server even 1000 miles away (let alone 10,000 miles away) has more to do with upstream peering and inter-ISP connection than it has to do with your connection.

Recommendation? Play on a server closer to home.

 

[konakona]

heh i knew this would stir some cynical responses, but here goes nothing.

It all depends on the provider. DSL is also oversubscribed. As a technology Cable should provide better service because of the higher bandwidth means less latency/delay.

True, I have spend quite a bit of time on dslreports reviews/forums and have noticed TW/comcast is not exactly the pinnacle of quality cable service. Then again, I can't talk about other cable based ISPs I havn't tried. I thought oversubscription of local nodes was a more pronounced problem with cable (as opposed to the distance to CO for dsl)? The message I got was that while there are many other factors that determine final outcome, cable was more notorious for this - which is what I was asking in the question in the first place.

Unless a service (DSL or Cable) is really "Quirky" the differences is usually in the mind of the Gamer and mainly involved psychological over compensation for the Gamer dissatisfaction with his status.

yes, I was specifically getting at which one tends to be more on the "Quirky" side in general, which is what I meant by "all things considered". NOTE: read the bolded part. The question was actually two folds, the first being the quirkiness of cable that I have experienced (and noticed from conversation, googling, etc) was common or isolated, and the latter being a comparison between gross averages of latency of the two - though I had a feeling you would say 'thats all in your head', or something along that line.

Normal reaction of people is between 180msec. to 300msec. (bigger numbers means slower reaction time), even much more slower under Alcohol, being tired, and similar states.

...and that has to do what with latency to game servers? Do you know one person that plays games at ease with 300ms ping to the game server?

Objectively if it is so imported to you, you should subscribe for one month to both services and choose the one that works better for you at your specific location.

I agree with you on that wholeheartedly, and that is just what I have already done. This is just more of an afterthought having been through several experiences that look nearly alike within a period of a decade covering several areas across the country. Up until now, I have been giving that same advice you gave me to anyone who turns to me to figure out which is better: it all depends on where you live. Now I am starting to think twice about that .

Jack, while I really admire your vast knowledge on the subject, I still don't agree your methods provide any meaningful insight as to why I was having countless streams of packet losses only at peak hours. Sorry about my incomplete initial post, I was getting that (horrible spikes + timeouts) to the default gateway and primary/secondary DNS servers. For one I think that has absolutely nothing to do with limits of human cognitive system or line testing methodology. Packet losses are just bad for anything to my knowledge, it just so happens that games are most time critical and hence least tolerant of dropped packets. Before you say this was one isolated incident, please don't forget it happened to me three times in a row (again, the bolded part). Stupid me to have give any benefit of doubt after all that, but still it got me thinking and wondering about how things are working out in general

I would hardly call online lag during gaming a "grave concern".

A bit irrelevant remark there, but thanks for sharing your opinion anyway. Gaming matters to me, but that was not the only thing at stake; any streaming video (e.g. youtube) would constantly rebuffer every 5 seconds or so literally. Browsing was a bit of a pain too, as quite a few hapless mistimed clicks to hyperlinks would miss and give me a 404 error. If that's not a grave concern to an ISP's client, what is? Are you suggesting the same as those TW CSRs - things are all peachy and good as long as green lights are flashing on the modem and I am able to get on the internet? They even told me to calculate packet loss % to their DNS server, and when I told that it was roughly 9% at the time they tried to shrug it off. Technician was only sent out at my insistence.

Basically, any difference in connection between you and a server even 1000 miles away (let alone 10,000 miles away) has more to do with upstream peering and inter-ISP connection than it has to do with your connection.

I wasn't asking how it exactly happens at a technical level; thats what ISP's would have to iron out. Again, I wanted to see with all that taken into account what turns out to be better in general.

Recommendation? Play on a server closer to home.

There are these things called MMORPG you know. Its not like you can take your character to any server you wish. I never blamed anyone for having a higher "base latency" due to unavoidable propagation delay. The reason I brought it up was to show that it is much easier for me to tell the difference otherwise less noticeable. Sure, 30ms vs 60ms might not be day and night, but past a certain threshold things become readily noticeable without trying. Think of it as FPS for gaming, it is arguable whether 60fps or 90fps is necessary, but below 30fps it is much easier to tell. Better yet, it is not matter of detecting differences, rather how much harder it actually is to play.