How Proxys and/or VPNs Can Reroute Your Data (and Improve It)

 

 

There are many people that logically assume that adding a proxy or VPN to your connection is only going to add latency to your game connection. These people have probably used free proxies or VPNs in the past, got bad results with them, and incorrectly concluded that all these technologies must be the same.

Before pursuing WTFast, I was also skeptical that gaming proxy/VPN services actually did what they advertised. Before launching WTFast there were about 5 to 10 other companies online offering these services. Many of these services were Australian, which I know now is very logical given the fact that they are on a big island and playing games in the US. Their data has to travel very far, giving lots of opportunity to improve on the connection.

This guide will go through some due diligence steps that I completed using proxy technology BEFORE I even considered pursuing WTFast. You can do these same steps yourself to validate that WTFast's GPN can dramatically impact game connections.

Proxy Tests, Traces & Data Samples With WoW

One of the things that I heard quite often is that these services were just a placebo effect and the in-game ping meter was not accurate. People claimed that the in-game ping meter only measured the latency between the proxy server and the game server, i.e. it was only measuring a partial trip, instead of the entire trip = end user + proxy + game server. In essence the claim was the in-game ping meter was being tricked and the gains were not real.

With the initial tests I did, this didn't make sense; I knew that the proxy server I was testing was less than 10ms from the game server. If it was only measuring a partial trip, then the in-game ping meter should be reading 10 ms. Instead I was seeing 100ms, which was 90ms to the proxy, and then 10ms to the server. This already seemed enough to confirm that the in-game ping reading was accurate and not just a false reading.

However, in order to confirm the in-game ping meter was measuring the full trip, I did some additional tests to confirm without a doubt that this is true. I will do some similar tests in this guide.

To keep things simple I will test with a WOW server in LA, and use a single server connection with WTFast, Los Angeles (choose best). If you want, you can use outside tools like traceroute or WinMTR to monitor connections and see where the trace is going. I really like PingPlotter since it shows a visual representation of the connection as well.

Using the trace data, I can see that the default WOW LA connection goes through Telus/Telia to get to Seattle, then it switches to AT&T, who heads over to Portland, then San Francisco and then finally LA. This is expected since AT&T is who Blizzard works with to deliver their game to you.

WOWLA.jpg

Now if I look at the trace data for WTFast CA4, I can see that Telus drops me off in Seattle, then Comcast takes over and sends the data down to LA, and then hands the data to the WOW server, which is only another 1 or 2ms away.

WTFastCA4.jpg

What does this mean? It means that using a single proxy, I can completely avoid any issues AT&T might be experiencing with their network, and use Comcast to get to the game server instead!

To confirm that this is real, below is a screenshot of the in-game ping meter for WOW:

WoWInGameMeter.jpg

The in-game ping meter is quite close to what we are reporting in WTFast (within 1 ms). Because we have to make estimates with our calculations, the reading is rarely so close, but it is nice to see when it is. Using the single WTFast server gave some decent improvements to the connection. Average ping was reduced a small amount, but the largest gain was with spikes and flux. That is the big jumps you see on the graph. Each one of those big jumps is a potential missed move.

Screenshot_2015-03-28_22.27.28.png

Next I will try running a test with our GPN automation/intelligence system, which attempts to find the best server combination for you automatically. I should highlight that this is very complicated, and also very new technology. I know that in many cases the GPN intelligence is failing for some users, and we continue to work on this. If our automated system ever lets you down, you can always just revert to picking a server manually (a server close to the game server, or close to you, picking (choose best). This logic generally should get you good results.

In this test, our system decided to do a chained connection, first to WA8, and then to CA9. With this configuration, I can see that Telus is dropping off in Seattle, and then it is handed over to Telus again, who goes directly to Los Angeles after this. It is important to note that if we were to just do CA9 all by itself, it would have sabotaged the connection, bringing it to 70ms+. But, our system was smart enough to see that if we chain WA8 and CA9 together that we could reduce the ms by another 3-4ms! This is the kind of automated intelligence that we are building into WTFast, our goal is to continue to expand on the intelligence to ensure we can get great results like this for every user all the time.

WTFastWA8.jpg

WTFastCA9.jpg

In addition to reducing the average ping by 3-4ms, the flux/spikes in the connection were also improved.

WoWInGame50ms.jpg

Screenshot_2015-03-28_22.56.20.png

Now, what happens if someone picks the wrong WTFast server? If our automated system picks the wrong server, or an end user picks the wrong server, the results will not be optimal, or they may even be worse.

I am going to see what happens if we use our WTFast server in Hawaii to play WOW LA! Nothing positive will happen with a poorly configured connection.

So the trace to Hawaii is 90ms.

WTFastHW2.jpg

And the data still has to get from Hawaii to Los Angeles, resulting in 142ms total trip time (ouch) and a red alert for a sub-optimal connection.

WoWInGame140ms.jpg

Screenshot_2015-03-28_23.12.08.png

In summary:
1. WTFast can improve routing, and it can improve ping/flux/spikes/loss for game connections. But it can only do so if the network connection is properly configured.
2. If our intelligence/automation system fails, the connection will not be optimal. If this happens, you can always manually pick a WTFast server. Generally a server close to the game server, or close to you (pick choose best) will get good results.
3. If you manually pick a bad server based on your game, the results will not be good.
4. If you live in the same city as the game server you play on, it might be hard to improve your connection further unless you are experiencing some really bad connectivity issues, in which case WTFast can help.
5. Improving the game connection is not just about reducing average ping. We created our connection meter to emphasize flux, spikes and packet loss. When you experience ping spikes and packet loss, each of these events is a potential missed move. Before we created the connection meter, our users said they could "feel" a difference, but it wasn't represented with the in-game ping meter. The WTFast connection meter visually shows all the missed moves that WTFast users only could feel before.
6. If you are having big problems really close to where you live, this is also something we may not be able to help with WTFast. Our traceroute/traffic report in WTFast can help to show if you have cascading problems with your connection. This is helpful information to show to your local ISP so they can fix the problem. I will do another guide later focusing on the traceroute reporting tool in WTFast.

 

 

What is Latency & How Does it Affect Game Response Time? Link

What is the difference between WTFast GPN and VPN. Link

Is WTFast Safe to Use? Link

Don't forget to upgrade to WTFast

Have more questions? Submit a request

0 Comments

Article is closed for comments.