Receiving feedback is invaluable, but it loses its worth when reviews are merely spam for credits.

Receiving honest, constructive feedback on your dating profile is invaluable. It’s the core purpose of this site. But that value evaporates instantly if the feedback isn't genuine – if reviews are just low-effort spam submitted purely to earn credits.

We understand how crucial trust and authenticity are to this entire process. That's why we want to be transparent about how we approach detecting and dealing with potentially fake or unhelpful reviews. However, this transparency has limits; explaining every single detail would unfortunately make it easier for bad actors to game the system. Our goal is to give you confidence in the process without handing over the keys.

No detection system is perfect, and determined individuals might find ways around safeguards. We can only strongly discourage this: attempting to spam reviews or use bots harms the experience for everyone – those giving genuine feedback and those hoping to receive it. We truly hope it never comes to this, but we will take action, including banning accounts, to protect the integrity of the platform.

Our Approach to Detecting Inauthentic Reviews

We employ a variety of automated and manual methods to identify reviews that don't meet quality standards. Here are some of the key approaches:

• Time Spent Per Review: Genuine feedback takes time. Reviewers need to actually look at the photos and read the profile text. We track how long each review takes overall, and how long a reviewer spends looking at each picture (including if they viewed all the images). Reviews completed faster than a reasonable minimum threshold – indicating they likely weren't given proper attention – won't earn credit and their input won't affect the reviewed profile's aggregated results. Excessively long times could also be flagged for review, though this is less common for simple spam.
• Profile Score Deviation: We expect some variation in scores – attraction and perception are subjective! However, a review score wildly out of sync with the consensus for a profile can be a red flag (either overly positive or negative spam). To measure this, we look at how far a user's score deviates from the running average score for that specific profile. We use the squared difference, which heavily penalizes outlier scores more than those that are just slightly different, and then average these squared differences for the reviewer. We also apply weights so that profiles with fewer reviews initially have less impact on this measure, allowing a consensus to build naturally.
• Profile Score Distribution: Simply spamming the same score (like "5") over and over is an obvious tactic. Some might try to get clever and spam different random numbers. However, humans are surprisingly bad at faking randomness – we often avoid realistic streaks or patterns. While we won't detail the statistical checks we run here (as that could help evade detection), we analyze the pattern of scores submitted by a reviewer over time to look for anomalies inconsistent with genuine human input.
• Behavioral Pattern Analysis: Beyond individual review metrics, we monitor for broader suspicious patterns. This could include reviewing profiles at an unusually high velocity, interacting with the site in ways characteristic of automated scripts, or other behavioral signals that deviate significantly from typical user engagement.
• User Reporting: Automated systems are powerful, but human oversight is invaluable. We rely on you, our community, as well. If you receive feedback that seems nonsensical, extremely low-effort (e.g., "good" on every aspect), abusive, or otherwise suspicious, please use the reporting feature. These reports help us investigate specific cases and also refine our automated detection models.

What Happens When Fakes Are Detected?

Our goal is to maintain quality, not just punish users. The actions taken depend on the severity and frequency of the issue:

1. Flagged Review: At a minimum, a review flagged by our systems will not earn credits for the reviewer and its data (scores, tags, time) will be excluded from the results shown to the profile owner.
2. Warning: For isolated or borderline cases, a user might receive a warning about review quality.
3. Temporary Suspension: Repeated low-quality reviews may lead to a temporary suspension of reviewing privileges.
4. Permanent Ban: Users engaging in blatant, persistent spamming, bot usage, or abusive behavior will have their accounts permanently banned.

We strive to be fair, but protecting the platform's integrity and the value of feedback for everyone is our top priority.

An Ongoing Effort

Combating spam and fake reviews is not a one-time task; it's a continuous process. Spammers adapt their techniques, and we constantly review and update our detection methods in response. It's an ongoing "cat-and-mouse" game, and we are committed to staying vigilant.

Conclusion

We are committed to building and maintaining a platform where you can trust the feedback you receive. This requires a multi-layered approach to detecting and handling reviews that don't meet quality standards. We appreciate your understanding of the need to balance transparency with security, and we thank you for participating constructively – both in giving thoughtful reviews and helping us identify suspicious activity. Together, we can ensure this remains a genuinely helpful resource for navigating the world of online dating.

No system is perfect and you may be able to get around it if you want to spam reviews or bot. Please don't. It hurts everyone and I'd rather not have to ban you.