X1 has a discovery problem. There are dozens of projects live on the chain — explorers, DEXes, launchpads, games, wallets, validator tools, identity protocols — and no single place to find out which ones exist, let alone which ones are any good. The information lives in Telegram threads that scroll away, in X replies, and in the heads of the twenty people who happen to have tried everything.

That is the gap Spanked was built to close. It is our community review directory for X1, and it now lists 83 projects across fourteen categories.

Why we built it

Three reasons, in order of how much they mattered to us.

To give X1 developers somewhere to be found. Building something good on a young chain and having nobody discover it is the default outcome, not the exception. A developer shipping a genuinely useful tool competes for attention with whatever is loudest in the group chat that week. A permanent, indexed page per project — with a description, a link, and whatever the community has said about it — is infrastructure that a Telegram message cannot be.

To let users see what actually exists. Most people arriving on X1 have no map. They know about the two or three projects someone mentioned to them. A directory answers "what is there?" before it answers "is it good?", and for a chain at this stage the first question is the more useful one.

To make the evaluation fair and honest. This is the hard part, and it is the reason the rest of this article exists.

The problem with review scores

Any public rating system in crypto faces the same attack on day one. A project wants a good score. Getting one is trivially cheap: post in your own community, ask twenty supportive people to leave five stars, and the average moves. Nothing about that is illegal or even unusual — it is simply what happens when you average unweighted opinions from self-selecting groups.

The result is a scoreboard where the number reflects how organised a project's community is, not how good the project is. Everyone learns to discount it, and the whole exercise becomes decorative.

So the scoring had to be built to resist that from the start, rather than bolted on after the first incident.

What Spanked does differently

We can describe the principles without publishing the parameters. Here is what the system does:

  • Every review is moderated by a human before it appears. Nothing auto-publishes. That alone stops most low-effort manipulation, at the cost of not being instant.
  • Reviewing requires a real X account. Not an email address, not a wallet — an account with history and a following, which is expensive to fabricate at scale. One review per account per project.
  • The score is not an average. Reviews are weighted, and not all of them count the same. A rating from an established account with a long history is not equivalent to one from an account created last week.
  • A small number of reviews cannot produce a confident-looking extreme. Scores from a thin sample are pulled toward the ecosystem average, and every score carries a visible confidence label. Most projects on the site currently read TOO EARLY TO CALL, and that is the honest description.
  • Reviews have to earn agreement beyond one group. A review that only people on one side of a disagreement find useful is treated differently from one that people who disagree with each other both find useful. This is the single most important defence in the system, and it is why a review can be rated helpful by a hundred people and still not count.
  • There is a delay before a new review affects the score. A coordinated burst does not move the number in real time.

The rule we will not break

Reviewers can optionally link a wallet, and having demonstrable on-chain history with a project earns a review additional weight.

What it never does is subtract weight from anyone who does not.

This was a deliberate decision and it is worth explaining, because the opposite design is tempting and would be a serious mistake. If unverified reviews counted for less, the system would systematically silence two groups: people who got scammed and no longer want to connect a wallet to anything, and people sensible enough not to connect a wallet to a review site in the first place. Those are precisely the reviews most worth having. Verification is a capped bonus, never a penalty.

The same logic applies to projects. A project with no on-chain program to check — an explorer, a wiki, a Telegram community, a news site — shows no verification state at all, rather than a failed one. "Cannot be checked" and "failed a check" are different claims, and most X1 projects are the former. Conflating them would quietly punish entire categories of legitimate work.

Why the algorithm stays private

Here is the uncomfortable part. We are not publishing the weighting.

This is a genuine cost. The most persuasive thing a review site can say is "check our arithmetic yourself", and we are declining to say it. We would rather be transparent. We have concluded we cannot be, for one reason:

The weighting is the anti-manipulation model. There is no separate defence sitting behind it. Publishing exactly how reviews are weighted — which signals raise a reviewer's influence, what thresholds a review must clear, how the cross-group agreement test is computed — would be publishing a step-by-step guide to defeating it. Every parameter we disclosed would be a parameter someone optimised against, and the people most motivated to do that optimisation are precisely the ones the system exists to stop.

Transparency about a security mechanism and the effectiveness of that mechanism are in direct conflict here. We have chosen effectiveness, and we would rather say so plainly than pretend the trade-off does not exist.

What we will commit to instead: the principles above are accurate and complete as a description of intent. Every review is human-reviewed. Scores are visibly labelled with their confidence. And the absence of verification never counts against anyone.

What it looks like today

83 projects listed, 11 reviews published, 8 projects with enough reviews to carry a score at all. The index across those sits at 7.1 out of 10.

That is a small dataset and we are not going to dress it up as more. Most projects on the site have no reviews yet, which is why most of them show no score rather than a fabricated one. The directory is useful today; the ratings will take time and volume to become useful, and pretending otherwise would undermine the point of building it carefully.

If you build on X1 and your project is not listed, submit it. If you have used something on this chain, leave a review — including the negative ones, which are the ones the system was designed to protect.

Spanked is at x1report.com/reviews. Figures current as of September 10, 2026.