MONOPOLY GO! Proxy: Which Traffic Goes Through the Tunnel?

SOCKS5's UDP relay: what the specification says, what the client does

In the SOCKS5 protocol, UDP transport is called UDP ASSOCIATE. It works like this: the client first establishes a TCP control connection with the proxy, requests a relay point over that connection, the proxy announces a UDP address of its own, and the client wraps its datagrams in a small SOCKS header and sends them there. When the control connection closes, the relay drops too; that is, the UDP flow depends on a session standing on TCP.

Two conditions must be met at once for this method to work. The proxy server must keep UDP relaying open — many providers leave it off because of resource cost and abuse risk. Second, the client has to be able to speak SOCKS5. It is unusual for mobile game apps to contain an embedded SOCKS5 client; and the proxy field in the operating system's wireless network settings is for an HTTP proxy.

The result is that in practice the tunnel remains limited to TCP. The HTTP proxy's CONNECT method also carries TCP only; a transport running over UDP 443, such as QUIC, does not enter this tunnel — it either goes out directly or, if UDP is blocked, falls back to TCP. In a turn-based, HTTPS-heavy game this limit usually goes unnoticed; in games carrying real-time voice and position streams it becomes decisive.

• Confirm from the panel whether your provider has UDP relaying enabled; the default is usually off.
• If a client only has an "HTTP proxy" field, your SOCKS5 exit will not work there.
• If UDP is not carried, that is not a fault but the scope limit of the protocol.

If you want to see the protocol's two options side by side, the SOCKS5 proxy page explains the difference at the transport layer, while the SOCKS5 UDP support article explains step by step how the relay is set up.

How relevant are NAT type and direct peer connections here?

The language of NAT types comes from games in which two clients connect directly to each other. There, an intermediary server exchanges the two sides' candidate addresses, after which the two clients send packets at the same time to try to open a temporary mapping on their own NAT devices; this is called hole punching. When a NAT behaving symmetrically, or carrier-level shared addresses (CGNAT), makes that mapping unpredictable, the connection cannot be established and a warning like "NAT type restricted" appears on screen.

In a server-authoritative, turn-based mobile game there is no need for such a channel between two clients. Every move is written to the server and the other side reads the result from the server; the only path between them goes through the server. For that reason the NAT type indicator on consoles has no direct equivalent here, and looking for a connection problem in NAT settings is usually the wrong door. The place to look is whether the request reaches the server.

NAT does still concern you at one point: the exit you use itself. Mobile operator exits are behind CGNAT, the address is shared with a large number of subscribers, and inbound connections cannot be accepted. Proxies and NAT are often confused with each other, but they are not the same thing: NAT rewrites the address fields of packets, while a proxy re-establishes the connection on its own behalf. The detail of the difference is proxy and NAT comparison article.

One final technical note: hole punching does not work over an HTTP proxy anyway. The CONNECT tunnel carries only the TCP session you initiated; it does not establish a mapping that would deliver an inbound datagram to you. In other words, using a proxy does not make peer connections easier — it adds one more layer in between.

Where is the rule written, and what traffic does it cover?

The coverage question must be answered before setup, because the same access credentials route a completely different set of traffic depending on where they are written. On a phone the options are few and each has a limit; if you are using an Android emulator on the desktop, the number of options increases, because there the operating system setting and app-based routing also come into play.

The steps vary by operating system: Android proxy settings and iPhone proxy settings articles show where the fields are. If you want to route the whole network from a single point, the proxy via the router article compares the gain and the cost; this path provides broad coverage but makes it harder to isolate which device is affected when a problem arises.

When setup is done, do not declare it "working" on the basis of a single screen opening. Try three things separately: the game opening, a piece of seasonal content loading, and the store page coming up. Because all three go to different endpoints, one can work while another remains outside the rule's scope. To confirm the exit is up from the outside, proxy checker tool you can use.

What does the intervening hop add to the time?

With a proxy, the total time of a request consists of three parts: the distance between you and the proxy server, the distance between the proxy and the game server, and the load the proxy is carrying at that moment. Since the third part changes over the course of the day, a one-off measurement is misleading; do not decide without measuring the same exit at different hours.

The conclusion that follows is clear: using a proxy increases latency in most setups, it does not lower your ping. The only exception is the rare case where your default route is convoluted and the proxy sits on a more direct backbone; that is not a rule but an exception that can only be confirmed by measurement. If a provider promises the opposite, measure the promise.

In this game the visibility of added latency is low, because the flow is turn-based and seeing the result of a move a few hundred milliseconds later does not spoil the experience. The difference accumulates in back-to-back requests: while a long event screen loads, dozens of small requests queue up and the latency added to each of them adds up. For the definition of the concept and the measurement method, see proxy latency , and to try your own exit, ping test tool.

Symptom, likely cause and check step

The table below separates the situations most commonly encountered in the field by their source. Following the order from top to bottom lets you make progress without guesswork.

407 The error is almost always related to authentication and has two sources: either the client is not sending the credentials at all, or the provider identifies you by IP authorisation and your exit address has changed. The latter is especially common on mobile lines, because the address can change when the line is renewed.

A timeout, on the other hand, is usually not an identity problem but a port problem. Most corporate networks and campuses limit outbound access to 80 and 443; if the exit your provider gave you listens on another port, the connection is never established. What the numbers mean is proxy port numbers is explained in that article.

Terms of service and cases where routing is not needed

Routing does not change the game's rules. Account ownership, purchase conditions and fair play rules apply regardless of where your connection goes out from. This page was not written for multi-account management, automated play or interfering with platform security measures. Such use both breaches the terms of service and puts your account at risk.

The cases in which routing has a reasonable justification are limited but real: using a fixed, registered address when going out from a corporate network, verifying the interface and campaign appearance in another country, measuring the source of connection problems comparatively against your own line, and separating your exit address for privacy reasons. None of these attempts to change outcomes inside the game.

Recognising when it is not needed matters just as much. If you play from your own country, on your own line, with a single account, there is nothing to be gained from putting a hop in between: latency rises, data allowance is consumed, and when a problem arises there is one more variable. Do not confuse the difference in scope between a proxy and a VPN either; the two do different jobs.