Proxy for Vainglory: Which Traffic Goes Through the Tunnel?
Most people who want to set up a proxy for Vainglory ask a single question from the wrong angle: "how do I run the game through a proxy?" The right question is this: which part of the traffic this game produces is of the kind a proxy can actually carry? This page answers that separately under the headings of protocol, client and region.
Protocol splitThe direct effect of the TCP versus UDP distinction on coverage.
02
Client separationThe different behaviour of the store client and the game client.
03
Region behaviourHow server selection and account registration are determined.
04
MeasurementManaging expectations and testing the exit correctly.
Vainglory is a long-running mobile MOBA that also expanded to the desktop side over time. That history leaves two different setup forms on the network side: a single application on the phone, and an installation managed by a distribution client on the desktop. A proxy decision does not produce the same result in these two forms.
The game's operating model has also changed over the years; which regional servers are open and the current state of account-side behaviour can differ from period to period. For that reason, rather than giving a definitive game-specific table, this page explains which mechanism depends on what. Always verify the current situation through the official channel.
What does not change is the technical ground: proxy is an intermediate hop, the protocols it can carry are limited, and its coverage depends on where you define it. The sections below make these three limits concrete in the context of Vainglory.
How does the request chain proceed?
A request that goes through a proxy passes four hops. First the client determines the destination and port; with an HTTP proxy this is sent to the proxy in plain text, while with SOCKS5 it is carried inside the protocol's own handshake. The second hop is the proxy server: the TCP session is established here, and on encrypted connections the proxy only carries the bytes and cannot read the content.
The third hop is the target service. On the Vainglory side this may be the account and identity endpoints, store and inventory calls, or the patch server. The fourth hop is the return of the response; the response comes back along the same path, meaning the same two legs are walked on the way out and on the way back. This symmetry explains why total latency increases.
What stays outside this chain is the match itself. The real-time gameplay stream consists of small, frequent packets, has high loss tolerance and is commonly carried over UDP. For such a stream, waiting for a retransmission produces a worse result than dropping the packet; the protocol choice is deliberate. Because the tunnel an HTTP proxy opens carries only TCP, this stream never enters the chain.
So do not think of the setup as "routing the game". What you route is a subset of the requests the game produces; which subset that is comes down to the protocol distinction in the next section.
DIAGRAMThe round-trip chain of a request through a proxy
You can scroll the diagram horizontally to inspect it
The same two legs are walked on the way out and back; this symmetry explains why the total time increases.
How does the line between TCP and UDP determine coverage?
An HTTP proxy operates at the application layer. It can read plain HTTP requests, add headers, and for HTTPS it opens a TCP tunnel with the CONNECT method. What that tunnel carries is a TCP session; it has no mechanism for UDP datagrams, and none can be added with any setting. A point-by-point comparison of the two protocols' capabilities difference between HTTP and SOCKS5 article.
SOCKS5 works one layer down, at the transport level, and does not interfere with the protocol passing through it. The standard defines a method called UDP ASSOCIATE for carrying UDP. However, three things are needed at once for this path to work: the server must support the method, the client must know how to use it, and the network in between must allow the relevant UDP traffic. If any of the three is missing, the stream goes out directly. The details of the mechanism SOCKS5 UDP support the article.
Question
HTTP proxy
SOCKS5
Does it carry TCP?
Yes, it works with the CONNECT via the tunnel
Yes, directly
Does it carry UDP?
No
Only if UDP ASSOCIATE if supported
Who resolves the domain name?
Proxy
Depends on the client
Is it common in game clients?
Partially
Rarely
In practice this table comes down to a single sentence: a proxy can cover Vainglory's account, store and download side, but not the match stream. When deciding which protocol suits which job, the protocol selection guide offers a checklist.
Why do the store client and the game client behave differently?
On desktop installations there are two separate programs. One is the distribution and update client: it manages the library, downloads patches, signs you in and displays the store screen. The other is the game itself. They are different processes, so when you write an application-based proxy rule, which one you pick completely changes the outcome.
The distribution client's traffic is TCP from end to end and speaks HTTPS. Patch downloads, library synchronisation, store and account calls all fall into this class. That is why this is the program a proxy rule covers most cleanly; once you define it, the rule genuinely works and the result is easy to verify.
The game client, on the other hand, behaves in a mixed way. At startup it makes some HTTPS calls, then switches to the real-time stream for the match. When you write the rule here, the requests in the first group are routed while the stream in the second group goes out directly. This is the most common reason a setup appears to be "half working", and it is not a fault.
On mobile there is no such distinction: a single application carries both the update and the game. There, coverage is determined by whether the application reads the system proxy setting. Clients that use their own network stack may ignore this preference; in that case the rule you write into the device setting does not affect the game.
Do not guess what each process is doing, observe it. If you open the distribution client on its own and start an update, your exit address will appear in those requests when the rule is working; if you repeat the same check with the game running and the picture changes, you have measured the distinction directly. Placing these two measurements side by side is the fastest way to separate the impression that "the setup is not working" from the reality that "the rule works correctly but does not cover that traffic".
DIAGRAMTraffic that can go through the tunnel versus traffic that goes out directly
You can scroll the diagram horizontally to inspect it
The distinction comes from the transport protocol; the left column shows the area where rules can be written, the right column the area unaffected by your configuration.
Choose the exit that suits your Vainglory setup
Whatever layer you cover, the type decision follows from it: a download-heavy setup calls for bandwidth, verification work for location diversity.
Choose whichever you need from our residential proxies, datacenter proxies, IPv6 and ISP solutions. Every plan comes with unlimited options, 99.9% uptime, rotating proxies, sticky sessions and 24/7 support. Ideal for web scraping, ad verification, SEO monitoring and digital data collection.
ISP ProxyStatic Turkish IPs registered to an ISP
ISP-registered static Türkiye IPs; they combine datacenter speed with the reputation of a real carrier. Ideal for long sessions and low-ping use.
In terms of volume, downloads can be larger than all other items combined. A version update or a first installation takes up an overwhelming share next to months of interface requests. That is why the cost side of the coverage decision is determined almost entirely by this item.
On exit types billed by data, pushing a large download through the tunnel goes straight onto the bill. On a flat-rate, high-bandwidth exit the problem may not be cost but speed: downloading through a shared pool during peak hours can take noticeably longer than downloading over your ordinary line.
The location of domain name resolution also comes into play here. Content delivery networks choose the node that serves a request largely according to the location of the party performing the resolution. When resolution happens on your network but the connection is made from a distant exit, a node chosen close to you is reached from far away, and every chunk takes an unnecessary detour.
Note
Leaving high-volume, latency-insensitive traffic outside your coverage is the soundest decision in most setups. Run the proxy only on the layer where you genuinely need it; let the rest stay on your ordinary line.
What determines the regional server?
In online games, "region" is used in two different senses, and the two get confused. The first is the physical server region where the match is played; the game usually determines this itself through measurement and matchmaking logic. The second is the region the account is registered to; this is a record set when the account is created or through the platform account, and held on the server side.
A proxy only changes where some of the requests in the first group appear to come from. It does not rewrite the account record, does not change the store country and does not govern matchmaking logic. The expectation that "changing the exit changes the region too" is therefore unfounded; moreover, use aimed at bypassing regional restrictions is a matter of the terms of service and is outside the scope of this page.
The legitimate use is regional verification: checking how a public announcement, a promotional page or a piece of community content looks from another country. This requires diversity of exit locations; which countries have exits available proxy locations on our page.
One more note: the exit country can affect the interface language and how some web content is served. This should not be confused with the game's own regional behaviour. The regional appearance you see in a browser and the registered region of your game account are two independent things.
In remote teams this distinction becomes a matter of process. If the same account is accessed from two different cities, tying the exit to the job rather than to the person produces a more consistent picture; everyone connecting over their own line makes the account look as if it is constantly moving. A short table recording which exit is used for which purpose both makes handovers easier and immediately tells you which jobs have stopped when an exit fails.
Getting the setup right in five steps
Order matters. First choose the protocol: if your client only offers an HTTP proxy field, your SOCKS5 exit will not work there; conversely, on a tool that supports SOCKS5, staying at the transport layer can be more flexible. Then define the scope: a single application, a device profile, or the whole network. These two decisions shape everything that follows.
The third step is defining access. A username and password work from anywhere but are a shareable secret; IP authorisation is practical on connections with a static address, but you lose access when your line is renewed. The details of the two methods authentication methods in the article. Where the system-wide setting is made on the desktop Windows proxy settings shows. Example format: proxy.example.com · 1080 · username · password.
The fourth step is verification, and it is the most frequently skipped one. "The program opened" is not verification; check your exit address from an independent page, measure where domain name resolution takes place, and compare the results with the proxy on and off. The fifth step is measurement and record-keeping: write down which exit you use in which profile, or you will not be able to find the variable when a problem comes up.
Choose the protocol according to what the client supports, not out of habit.
Define the scope in a single sentence: "this traffic of this program".
Do not run a VPN and a proxy at the same time; a double layer makes diagnosis impossible.
Repeat the verification with the proxy off and compare the two results.
DIAGRAMThe five steps that get the setup right
You can scroll the diagram horizontally to inspect it
Order matters: an access definition made before the protocol and scope decisions will have to be rebuilt from scratch later.
Provider choice, continuity and security
A proxy exit is the most fragile part of the setup: when it stops working, everything it covers stops too. That is why continuity comes before speed in the selection. The provider's outage policy, maintenance windows and what it commits to in case of failure are concrete questions; uptime and SLA the article explains how to read these commitments.
The second question is trust. On HTTPS connections the proxy cannot read the content, but which domain you connect to is visible on the proxy server and can be logged. This makes provider choice as much a trust decision as a technical one. An unexpected certificate warning is a separate category: a properly configured HTTPS proxy does not interfere with the TLS session, so if you see a warning it means your traffic is being decrypted and re-encrypted at an intermediate point. Outside a corporate network, do not click through this warning.
The third question is the exit type. If high bandwidth and low cost are your priority, datacenter proxy is the most efficient option; in return, the network class is plainly visible to the other side. If you need to stay close to a typical home-user profile, residential comes to the fore; if a static address and stable speed are wanted, the ISP solution does. The decision depends on what traffic you cover: a download-heavy setup and a verification-heavy setup do not call for the same type.
Measurement and realistic expectations
Every comment made without measurement is a guess. Before putting an exit to work, measure its response time, repeat the measurement at different times of day and note the result; on shared pools, the peak-hour difference is the biggest variable a one-off measurement hides. Method and pitfalls how to test proxy speed article.
The expectation side can be summed up in a single sentence: because an extra hop is added, the total time increases in most setups; it does not lower your ping. There is a rare exception — if your default route is unnecessarily circuitous and the proxy sits on a more direct backbone, the total time can get shorter. But that is not a rule; it can only be understood by measuring and comparing both cases, and the game's in-match stream is outside this calculation anyway.
Finally, do not confuse what you are measuring. The response time of the proxy exit and the game's in-match latency are two different quantities; the second belongs to a stream the proxy does not cover. When evaluating a setup, look at the behaviour of the layers it does cover: how long login takes to complete, how long the store screen takes to fill, how fast the download progresses. These three measurements tell you what your setup actually does.
Questions asked about Vainglory and proxies
01I set up a proxy but my in-game latency did not change. Why?
That is the expected outcome. Match traffic is real-time and loss-tolerant, it is commonly carried over UDP, and it does not enter an HTTP proxy's CONNECT tunnel. Your configuration covers the TCP side: account, store and downloads.
02Should I write the rule for the distribution client or for the game?
It depends on your goal. If you want to cover patch downloads, store and account calls, the distribution client is the right target, and because all of its traffic is TCP the rule works cleanly. If you write it for the game client, only the HTTPS calls made at startup are routed.
03If I choose SOCKS5, will match traffic go through the tunnel too?
Only if all three conditions are met at once: the server must support UDP ASSOCIATE , the client must know how to use that path, and the network in between must allow the relevant UDP traffic. In game clients, the second condition is rarely met.
04If I change the exit country, does my account's region change?
No. The account's registered region is held on the server side and is not rewritten based on the address the connection comes from. Using this to bypass regional restrictions is a matter of the terms of service; this page only covers verifying how publicly available content appears.
05Do I have to route downloads through the tunnel?
No, and in most setups it makes more sense not to. Downloads account for the bulk of the volume; on an exit billed by data they produce a direct cost, and on a shared pool they produce a noticeable slowdown during peak hours.
06I see a certificate warning. Should I continue?
No. A properly configured HTTPS proxy does not interfere with the TLS session; the warning indicates that your traffic is being decrypted and re-encrypted at an intermediate point. On a company-managed device a root certificate may have been installed deliberately; on your own machine, disconnect without entering a password into the distribution client or the account page.
07Does it make sense to use a VPN and a proxy at the same time?
Not for diagnostics. When two layers run together it becomes almost impossible to isolate which step is affecting the result, and the hop added twice increases latency further. Verify with a single layer first, then decide based on the result.