Fraud Scores and IP Reputation (2026): What They Mean for OFM

You've bought a proxy, and someone tells you to "check the fraud score." Then someone else says "0 fraud score doesn't matter." Both partially true. This guide is what fraud scores actually measure, why they matter, and when they don't.

1. What a fraud score is

A fraud score (or IP reputation score) is a composite number, typically 0-100, where 0 is clean, maintained by IP-reputation databases. The score reflects how likely an IP is to be a proxy, VPN, bot, or abusive traffic source.

Who maintains fraud scores

• IPQualityScore (IPQS), most-referenced in OFM.
• Scamalytics, free lookup, community-standard.
• MaxMind, enterprise.
• IPHub, proxy detection.
• Pixelscan, includes broader fingerprint checks.
• Spamhaus, AbuseIPDB, blocklist-focused.

What they measure

• IP has history of botting / scraping.
• IP is known commercial proxy/VPN exit.
• IP is from datacenter ASN.
• IP has recent abuse reports.
• IP participates in anomalous traffic patterns.
• IP is in a heavily-flagged subnet.

---

2. Why platforms care

Most social / dating platforms check IP reputation on login. High-fraud IPs trigger:

• Additional verification (email, phone, ID).
• Shadowban on new accounts.
• Instant rate-limit.
• Outright rejection.

From the community:

"When purchasing a static residential proxy from Iproyal, should I check for fraud scores? to use on Reddit"

→ Yes.

"What proxy provider are people using for DA these days that have low fraud scores?"

Platform staff reviewing signup batches look at fraud score. Good score ≠ guaranteed pass, but bad score ≈ guaranteed fail.

---

3. The "0 fraud score" myth

From the community:

"Anyone here have experience requesting 0 fraud score IPs from IProyal? I did and it is 0 fraudscore but idk what to think cause it looks too good to be true and everyone hates on IProyal lmao"

"Is zero fraud score static residential proxy from iproyal good?"

Honest truth:

• 0 fraud score is GOOD.
• 0 fraud score is NOT GUARANTEED PASS.

Platforms have their own internal IP-reputation systems beyond public databases. Your IP can show 0 on IPQS but still be flagged internally by:

• Platform's ML detection of proxy patterns.
• Subnet reputation.
• Per-platform custom blocklists.
• Device fingerprint issues unrelated to IP.

Treat 0 fraud score as "removed one red flag" not "cleared for takeoff."

---

4. How to check IP reputation

Free tools (check before assigning proxy to account):

IPQualityScore

• ipqualityscore.com/free-ip-lookup-proxy-vpn-test
• Composite 0-100 score.
• Free limited daily queries.

Scamalytics

• scamalytics.com
• Simple risk tier.
• Free.

Pixelscan

• pixelscan.net
• Broader fingerprint check, includes IP reputation.
• Free.

Whoer

• whoer.net
• Trust score + fingerprint.
• Free.

Community workflow:

1. Buy proxy.
2. Check at 2-3 services.
3. If any flag it, rotate to fresh IP or different provider.

---

5. Why fraud scores drift

An IP that's "0 fraud score" today may be flagged next month. Reasons:

• Provider pool exposure, as IPs get used for proxy purposes, reputation databases update.
• Other operators' activity, if someone else used this IP for scraping, reputation degrades.
• Subnet contagion, if neighboring IPs get flagged, yours gets downgraded.
• Database refresh cycles, reputation services update data continuously.

Implication: fraud score at time of purchase isn't permanent. Long-running accounts may see their IP reputation degrade without any action on their part.

---

6. Subnet-level reputation

Even if your specific IP is "clean," neighboring IPs in the same subnet matter:

• Subnet = IP range (e.g., 192.168.1.0/24 = 256 IPs).
• If 10 IPs in that subnet are known proxies, your IP in the same subnet is guilt-by-association.
• Platforms often flag at subnet level, not just IP level.

From the community:

"Cause with bartproxy you have the Problem that you have 25 Proxys in the Same sub net"

Buying 25 IPs from same subnet = all will get linked / flagged together.

Mitigation: request subnet-diverse IPs.

---

7. ISP / ASN reputation

Beyond subnet, platforms check ASN (Autonomous System Number), the ISP-level organization of IPs:

• Real consumer ISPs, Comcast, AT&T, BT, Vodafone → clean ASN.
• Known proxy providers, some ASNs blanket-flagged.
• Datacenter ASNs, AWS, GCP, Azure → universally flagged.

Why this matters for residential: if provider's residential IPs route through one of their own ASNs (rather than the actual consumer's ISP), platforms detect the ASN and flag.

Ask providers about ASN diversity for premium residential purchases.

---

8. Fraud score per proxy type

Mobile 4G/5G

• Generally low fraud score (0-10).
• Carrier ASN is clean.
• NAT cover makes individual-IP reputation less relevant.
• Lowest-fraud-score category by default.

ISP proxies

• Variable (0-30).
• Fresh ISP IPs usually clean.
• Long-resold ISP IPs accumulate fraud reputation.
• Check before purchase.

Residential rotating

• Variable per IP (0-50).
• Pool quality varies wildly.
• You may get clean IPs one session, dirty the next.

Datacenter

• High fraud score (70-100).
• ASN universally flagged.
• No remediation, avoid for account work.

---

9. Per-platform sensitivity

Platforms weight fraud score differently:

Very high sensitivity

• Tinder, high fraud = instant shadowban.
• Bumble, similar.
• Instagram new account, very strict.

High sensitivity

• Instagram established accounts, moderate tolerance once aged.
• TikTok, strict on creation + video upload.
• Facebook, strict, KYC overlay.

Moderate sensitivity

• Twitter / X, more permissive than IG.
• Snapchat, moderate.

Lower sensitivity

• Reddit, less proxy-detection investment.
• Telegram, minimal.

Match fraud-score threshold to platform. Reddit accounts tolerate 10-25 score. Tinder demands 0-5.

---

10. Operator strategies for fraud-score management

Pre-assignment checking

Before moving proxy to account, verify score. Reject IPs above threshold.

Provider guarantees

Some providers (premium mobile, premium ISP) offer "0 fraud score" guarantees. Test the claim.

Rotation off flagged IPs

If fraud score degrades on an IP, rotate account to fresh IP before ban.

Tier-matching

Use premium proxies for platforms that demand low fraud. Use budget proxies where tolerance is higher.

Multi-provider stack

If provider A's pool is flagged this week, provider B's might be fine.

---

11. The "I bought 0-fraud IPs and accounts still banned" pattern

From the community:

"I bought 0-fraud-score IPs from IPRoyal and accounts still banned"

When this happens, it's NOT the fraud score. Usually:

Fingerprint issues

AD browser fingerprint inconsistent with IP location (timezone, language).

Account content

Profile or posts triggered platform AI review.

Behavior patterns

Account acting bot-like (too consistent timing, too mechanical).

Subnet reputation

IP clean at individual level, subnet flagged at platform level.

Account history

Purchased account has prior flags that surface on login.

Fraud score is ONE signal among many. Clearing it is necessary but not sufficient.

---

12. VPN detection vs proxy detection

Most platforms detect both via similar systems:

• VPN exit ranges, commercial VPN providers' IP blocks are published.
• Proxy exit ranges, commercial proxy providers' blocks similarly trackable.

Using VPN stacked on proxy doesn't hide anything, it adds more detection signals, not fewer.

From the community:

"Has anyone tested do vpn's work for ig/fb? (nordvpn...) or proxy is a must"

→ Proxy is a must. Commercial VPNs fail on strict platforms.

---

13. The 15-minute lock / verification trigger

From the community:

"Which proxy provider yall suggest to have the lowest fraud score as possible and not get 15min lock?"

The "15-minute lock" is a platform-specific soft ban triggered when fraud score is borderline. Lock resolves but indicates IP reputation is marginal.

Fix:

• Rotate to fresh IP.
• Check fraud score of replacement first.
• Warm up new IP slowly before high-stakes action.

---

14. When fraud score doesn't matter

Platforms that ignore it

• Telegram, near-zero proxy checking.
• Some niche platforms.

Scenarios where fraud score is secondary

• Legitimate heavy-proxy use-cases, scraping public data with legal right.
• Backend automation, not user-facing.

When to not obsess

• Score is 5 vs 0, rarely matters.
• Score degrades from 2 to 8 over time, still fine.
• Score is 12 on a budget provider for a budget use case.

---

15. Common mistakes

Obsessing over 0 vs 5

Platforms don't threshold that precisely. Chasing perfection wastes money.

Ignoring fraud score entirely

Then wondering why Tinder bans all signups. Budget-provider dirty IPs cause this.

Checking only IPQS

Single source. Cross-verify with Scamalytics at minimum.

Thinking VPN is cleaner than proxy

Usually opposite. Commercial VPNs are heavily flagged.

Assuming fraud score stays static

It drifts. Monitor periodically on long-running accounts.

---

16. Frequently asked questions

What's a good fraud score for Tinder?

0-5 ideally. <10 acceptable. >25 will ban fast.

What about Reddit?

<25 is safe. Reddit is less proxy-sensitive than dating apps.

Can I check fraud score for free?

Yes, IPQS free tier, Scamalytics, Pixelscan all free.

Does 0 fraud score guarantee account approval?

No. One signal among many.

Why does my fraud score keep changing?

Databases refresh continuously. IP reputation drifts with usage.

---

Related guides

• Guide 4, Evaluating Providers
• Guide 8, Setting Up Proxies
• Guide 9, Proxy Troubleshooting

---

Built from a corpus of ~405 real operator discussions across 11 OFM Telegram communities (2024-2026). Usernames anonymized.