There is ping, then there is jitter. Ping is the delay between you and your opponent. Jitter is the range of your ping. For example, if your ping ranges from 30ms to 50ms, then the jitter is 20ms.
Use http://www.pingtest.net/ , pick a local server to test your Ping and Jitter. If your Jitter is above 50ms, then that means it’s your ISP that is lagging the packets. In terms of game play, it means your input packets will be lagged behind up to 50ms.
So before you make complains about online lag, check to see if your ISP’s fault first.
Cheapest internet connection packet you can buy. If only I get more than one person per hour. My schools internet connection is insane though, it’s connected directly on the hardline or whatever, fiberoptics and stuff. If I can find a smaller tv, I might bring my xbox over for some afternoon rounds.
The pingtest website is pretty picky about what browser plugins you have for the packet loss test.
My recommendation (for windows users) is to test with http://winmtr.net/ . It’ll even tell you But you’ll have to figure out the target server your self. Check out universities near you, or you can try the address of one of the pingtest servers.
Here’s a sample of WinMTR to a local university webserver:
As you can see, the pings are quite a lot higher than pingtest showed. That’s because I used 1KB ping packets, which is a lot larger than usual. This can be used to test whether there’s some problem of large packets getting dropped.
The pings for the intermediate route nodes don’t actually tell much here, they’re probably not responding to pings promptly. Usually you should see the ping (Best/Avg/Worst) steadily increasing along the route from top to bottom. For some reason the Finland backbone network isn’t responding to pings promptly, so you see excessive pings along the routing nodes, even though packets are relayed quickly.
At the end node my average ping was 48ms, which is good considering the packet size. Also, there is 0 packet loss, which is important.
This tool is good, because you can just set it up and leave it running for 30min, so it can detect even intermittent problems. And it will show you which node is dropping the packets / causing latency to spike.
58ms with 24ms jitter is actually very very bad. Modern broadband should be in the 10-20ms range (if testing to server in same town) with roughly 0 ms jitter.
I would check what is wrong with your connection, is someone leeching all the bandwidth with downloads, or maybe you’re on bad wireless connection?
I hope all of you guys realize that choosing the server closest to you will only help you determine if you are gonna have good matches with people within that range.
The lag generally comes from playing people who are further out, so that’s what you actually have to check. You are not gonna have the same ping with someone who’s halfway across the country as someone who’s within the same city.
I believe these games are p2p online so only testing your own connection will never do any justice. Your opponent could have a 90ms connection to you but it could of been a wifi signal with 20ms in jitter. So it will eventually bounce into underworld land sometimes
There needs to be additional functions added to online games.
constant jitter test for both opponents prior and during games.
identify wireless connections. The xbox system has got to recognize someone is running wifi. Pull from that info from that section and display it for online mode
jitter test is really good to see if someone is running a wifi\shared router. Wifi bounces quite high and it always has packet loss associated with it. A shared router can be great when no one is on it but those connections can be random because @ any given moment, someone could jump onto their network and start surfing\downloading.
When you see a 60ms connection but its in underworld land, chances are someone is on a wifi\shared connection and its being abused. There is no way to tell that someone is going to do that before going into matches and its quite troublesome.
Well, you can only fix your own connection, so that’s what you want to be testing. It’ll play much better already if at least one player in the game has their cables on straight.
As a simple rule of thumb, one can do this: add up your pingtest latency and jitter, then multiply the sum by two. That’s the worst case ping you’d get in-game when playing against someone with the same connection as you in the same town.
Taking myself as an example, thats (10+0) = 10, times two is 20ms. Very good. Now take Oerba’s connection with 54ms ping, 14ms latency. Add up 54+14 = 68, then multiply by two = 136ms. So he’d be playing with 136ms at worst against someone in the same town with similar connection. For comparison, that’s the kind of ping I get from northeast Europe all the way to USA east coast. It may be somewhat playable at GGPO Delay 8, but nowhere near smooth experience anymore.
This is why your latency matters, if even one player has eliminated their latency, the chances of getting terrible games lessens by a lot. Because if it was me vs Oerba, that’d be 54+14+10+0 = 78ms worst case (Assuming we lived in the same town of course). That’s not so terrible, it can be playable, just need to up GGPO Delay a little.
And just a friendly PSA, because somebody was posting about connection speeds in GD thread: connection speed has pretty much nothing to do with gaming performance, 1Mbit/s is plenty for any game, even having less shouldn’t negatively affect games as long as its some kind of broadband. You’ll want to focus on the latency. Larger bandwidth is essentially like adding more lanes to a highway. If you don’t upgrade your car (latency) from Lada (wireless) to Ferrari (wired fiber optic), you won’t get to your destination any faster.
Disclaimer
Spoiler
I know I’m oversimplifying a lot of things, just trying to explain a complex issue in simple terms.
