How to Stop Betting Arbitrage and Multi-Accounting in Online Sportsbooks

Betting arbitrage is one of the oldest threats to sportsbook margins, and the internet has made it worse. A bettor who can hold accounts at multiple sportsbooks simultaneously can exploit the small differences in their odds to lock in a guaranteed profit regardless of the outcome. The practice is not illegal in most jurisdictions, but it is deeply corrosive to bookmakers — and when arbitrage bettors are blocked or limited, many turn to multi-accounting, VPNs, proxies, and antidetect browsers to keep doing what they were doing under new identities. This article explains the mechanics of both threats, the signals they leave, and how IPrating's filtering flags the high-risk players behind them — before they place a bet.

How Betting Arbitrage Works

Betting arbitrage — also called "sure bets" — is an opportunity that arises when bookmakers' differing opinions on event outcomes, or simple pricing errors, create a set of odds whose implied probabilities add up to less than 1 (Cortis, 2015). By placing one bet on each outcome with different betting companies, the bettor guarantees a profit no matter what happens. The margins are thin — 98% of arbitrage opportunities return less than 1.2% (Franck, Verbeek & Nuesch, 2012) — but the strategy is volume-driven. An arbitrage bettor placing dozens of sure bets per day across multiple sportsbooks can accumulate meaningful returns, all at the bookmaker's expense.

The problem for sportsbooks is structural: arbitrage requires simultaneous accounts at multiple providers, and it requires the bettor to move quickly when a pricing discrepancy appears. The arbitrageur is not a loyal customer; they are a systematic extractor of margin. Bookmakers typically hold an unfavorable view of it and respond by canceling bets, closing accounts, or reducing the bettor's maximum wager limits to the point where the strategy is no longer viable (Schwartz, 2016).

The Multi-Accounting Escalation

When a sportsbook detects arbitrage behavior and limits or closes an account, the bettor faces a choice: stop, or find a way back in. Most choose the second option. Multi-accounting — creating multiple accounts on the same sportsbook, each appearing to belong to a different person — is the standard response, and it is where the technical evasion begins.

A multi-accounter needs to defeat the sportsbook's identity and device checks, which typically means:

• VPNs and proxies to present a different IP address for each account, masking the fact that the same person is behind all of them. Datacenter proxies are cheap and fast; residential proxies are harder to detect because they route through real ISP IP ranges.
• Antidetect browsers to present a different browser fingerprint for each account — different User-Agent, different screen resolution, different canvas fingerprint, different installed fonts — so the sportsbook's device fingerprinting cannot link the accounts to a single machine.
• Disposable email addresses and phone numbers to complete registration for each account without revealing the operator's real identity.

The combination is designed to make each account look like a unique, legitimate visitor. A sportsbook that relies on IP database lookups and basic User-Agent checks will see a different city, a different browser, and a different device for each account — and will have no way to connect them. This is exactly the gap that real-time traffic analysis closes.

Why IP Databases Miss It

The traditional defense against multi-accounting is to check the visitor's IP against a geolocation database and flag anything from a known VPN or hosting provider. As we documented in our comparison of real-time analysis versus IP databases, this approach has structural weaknesses that multi-accounters exploit directly:

• Databases are stale. Academic research (Gouel et al., 2021) found that 47% of end-user IP addresses move more than 40 km between successive database updates. A proxy IP that was flagged as datacenter last week may have been re-registered or re-routed since.
• Residential proxies are invisible to ASN-only checks. A 2026 study across 175 countries (Nabi et al., 2026) showed that mobile networks exhibit median geolocation errors 10× higher than fixed-line. A multi-accounter using a residential proxy network presents an IP that looks residential to an ASN database, even though the traffic is being tunneled through a commercial proxy.
• Databases cannot see the browser. An IP database tells you where the address was registered. It cannot tell you whether the browser is real, whether the OS matches the User-Agent claim, or whether the TLS handshake looks like a genuine client or a fingerprinting library. Antidetect browsers are specifically designed to defeat the checks that IP databases cannot perform.

How IPrating Flags High-Risk Players

IPrating does not rely on a static database. It classifies every visitor in real time, at the moment they arrive, using the multi-layer detection we described in our real-time detection and VPN configuration articles. For a sportsbook, the filtering controls map directly onto the signals that arbitrage bettors and multi-accounters leave behind:

VPN and proxy detection — Set the VPN, Datacenter Proxy, Public Proxy, and Residential Proxy dropdowns to Block or Challenge, and every account created through an anonymizing layer is flagged before registration completes. A multi-accounter rotating through residential proxies will still be caught by the cross-layer correlation that compares the proxy suspicion against the browser's network-stack behavior — the same detection that catches antidetect browsers, because the browser identity and the network identity do not agree.

Antidetect browser detection — IPrating does not detect antidetect browsers by their User-Agent, which they forge. It detects them by checking whether the browser's claimed identity is consistent with what the network and TLS handshake reveal. When a browser claims to be Chrome on Windows but the TCP fingerprint shows Linux, or when the browser claims protocol support that the handshake does not confirm, the classification captures that inconsistency. Each such account is flagged as high-risk before it can be used for arbitrage.

Bot and automation detection — Arbitrage opportunities are often identified by automated scanners that monitor odds across multiple sportsbooks in real time. Set the Automation Tools category to Block and the scanner traffic is filtered before it reaches your odds pages — meaning the arbitrageur never sees the discrepancy in time to exploit it.

Geo compliance — Many arbitrage bettors operate from jurisdictions where online gambling is restricted, using VPNs to appear as if they are in a permitted region. IPrating's Country Whitelist, Country Blacklist, and VPN detection work together to enforce jurisdictional compliance at the connection level, before the account is created.

Custom signatures — When a specific arbitrage tool or multi-accounting operation uses a recognizable User-Agent pattern, a custom signature catches it on every request and labels it in your analytics — so you can see how many high-risk signups you stopped and from which source.

From Detection to Decision

The key difference between IPrating and a post-hoc fraud review is timing. Traditional sportsbook fraud teams review accounts after they have been created, after bonuses have been claimed, and after arbitrage bets have been placed. The damage is already done; the team is chasing refunds and closing accounts that should never have been opened. IPrating classifies the visitor before they reach the registration form — so the high-risk signup is blocked, challenged, or flagged at the door, not discovered in a report three days later.

And because the configuration is the same four-action dropdowns we described throughout this series — Allow, Block, Challenge, Redirect — the sportsbook's risk team does not need to build detection logic or tune thresholds. They set the network types they want to block, choose the action, and save. The detection, the cross-layer correlation, and the browser reality checks happen on every request, in under 5 milliseconds of compute, with no impact on the signup or login experience for legitimate players.

Protect Your Margins

Betting arbitrage and multi-accounting are not going away — the economics are too attractive and the evasion tools are too accessible. But the signals they leave are identifiable, and they are identifiable before the damage is done, if you are looking at the right layers. IPrating looks at all of them — network, browser, OS, TLS, and behavioral — and turns the conclusion into a simple action. For a sportsbook, that means margins protected, bonuses reserved for real players, and compliance enforced at the connection level. See our iGaming solutions page for the full capability set, or check pricing to see how protection scales with your player volume.

Sources

• Cortis, D. (2015). Expected Values and variance in bookmaker payouts: A Theoretical Approach towards setting limits on odds. Journal of Prediction Markets, 9(1), 1-14. doi:10.5750/jpm.v9i1.987. (Definition of arbitrage in betting markets.)
• Franck, E., Verbeek, E., & Nuesch, S. (2012). Inter-market Arbitrage in Sports Betting. Economica. (98% of arbitrage opportunities return less than 1.2%.)
• Schwartz, A. J. (2016). Arbitrage in the European Soccer Betting Market. Yale University. (Bookmaker responses: account closure, wager limit reduction.)
• Gouel, M. et al. (2021). Longitudinal Study of an IP Geolocation Database. arXiv:2107.03988. (47% of IPs move >40km between updates — database staleness.)
• Nabi, S. T. et al. (2026). Lost in the Prefix: Revisiting IP Geolocation Accuracy Across Networks and Geographies. arXiv:2605.21937. (Mobile networks exhibit median errors 10× higher than fixed-line; residential proxy detection gaps.)