What is CORE3? Quantifying why 15M crypto projects failed

Since 2021, 13.4 million cryptocurrencies have died. Add failed protocols and leftover ICO tokens, and the number climbs to roughly 15 million projects that no longer operate.

At CORE3, we first asked why they fail. Then we asked a harder question: what are the odds the next coin someone sees will go to zero?

Roughly 95% never found a market or a use. That group includes memecoins, “utility” tokens tied to empty products, rug pulls, and honeypots. CoinGecko reports that 2025 set records for both new token launches and shutdowns, driven by one-click launchpads and fair-launch sites like pump.fun.

But it wasn’t just low-effort coins; serious projects collapsed too, many during the October 2025 flash crash. When volatility hit, there was no depth. Prices fell straight to zero, sometimes in seconds, leaving tokens with dozens of decimals and no bids.

This points to a basic flaw. Crypto assets are priced mainly on speculation and the signals around it. When extreme events hit, even projects in the top 200 by market cap can fail.

Extreme events happen, and projects fail under their force, or even without it. But CORE3 emerged to quantify how the project is exposed to the risks, and how well it’s prepared to adverse events, when they happen. This quantification comes from our flagship metric.

Probability of Loss (PoL) is CORE3's unbiased, data-driven metric that quantifies a project's risk exposure on a scale from 0 (Exceptional) to 100 (Critical risk).

The methodology evaluates 98 assessments across six risk domains: operational, financial, security, dependency, regulatory, and reputational — to create a unified risk language for investors, builders, and institutions.

Currently, the market measures trust through tokenomics, TVL, security audits, and high-tier backers standing alongside the project. This approach opens doors to partnerships and liquidity. It's also why crypto remains a grey market.

PoL changes that. It assesses beyond audits and compliance to capture the full risk picture: multisig configuration and key rotation frequency, insurance coverage, continuous incident monitoring, custody architecture, CCSS certification.

Code audits might catch 14% of the risk. Operational failures, access control breaches, custody misconfigurations, team impersonation, and dependency compromises, cause 58% of all Web3 losses, totaling over $2 billion. Traditional risk due diligence stops at the audit report. PoL dives deeper.

The probability of loss brings to light the most common reasons projects actually fail — not low TVL or weak trading volume, but centralized custody control, outdated audits, and unprotected attack surfaces that backers and influencers never flag.

Today, assessing a crypto project means stitching together 15 different signals. Check the audit, GitHub commits, multisig design, governance votes, Discord activity, price history, on-chain data, community sentiment, licensing, past exploits.

After three hours, one question still goes unanswered: is this more or less risky than the project reviewed last week?

Institutions try to solve this with internal spreadsheets. They don’t scale. They can’t be monitored over time. They aren’t comparable across teams. When an incident hits, there’s no clean way to reassess every other project with similar traits.

Builders end up competing on hype because there’s no reward for quiet execution. With $20,000 and decent marketing, a malicious actor can game every credibility signal in use today. Teams that invest in real security get buried by whichever community shouts the loudest.

Retail users face two bad options. Either trust influencers who sell opinions. Or try to combine dozens of data points they can’t reliably collect. Both paths lead to losses that could have been avoided.

The probability of loss score changes the frame. It replaces crowd belief with math. Signals from 98 separate assessments consolidate into one number, with a clear breakdown by risk type. Due diligence becomes infrastructure: consistent, measurable, repeatable, and ready for review.

For builders, the probability of loss score becomes a roadmap, where a high-risk score isn’t a verdict. Instead, it shows what’s missing: add a bug bounty, rotate custody keys, write an incident plan. Each fix lowers the score and makes the project less prone to incidents. The dynamics of lowering the risk show the team can be trusted by all parties.

For institutions, PoL becomes a standard that works at scale. Listing teams set cutoffs, funds shape portfolios by risk exposure, compliance teams can explain decisions without guesswork. The metric fits into daily workflows because it is actionable and covers a wide area of overall risk.

For retail power users, the shared risk score becomes a fast filter. Firstly, the projects are now comparable by risk. While probability of loss is not a price change indicator, it helps as a part of a broader investment strategy. When there is a speculative opportunity, investors can know which risks they take. 

Proof of opinion (PoO) is an unbiased review from certified expert, shows projects as they are, not as founders want them to be seen. Each review is authored and linked to an SoulBound DYOR-certificate credential. 

PoL measures mechanics. But metrics alone can't answer: Is this project actually used? Does it have relevance? Is the ecosystem alive?

PoO was created for cases, when a project can score well on security, documentation, and governance, yet remain stagnant with no user adoption, no product-market fit, and no real-world traction.

The PoO contributors represent how the average user see the project, their analysis focuses on what project does and it’s strengths, weaknesses, and perspectives. 

The SoulBound Token makes every opinion attributable. The credential system makes every analyst accountable.

By design, PoO is separated from PoL. The metric stays objective. The PoO context stays human.

PoL must remain independent from opinions to be trusted by institutions and regulators. If reviews changed the score, the methodology loses its credibility. 

Separation lets experts speak honestly, critically but fairly, without being accused of manipulating ratings. It also shields PoL from brigading, reviewer coercion, and coordinated campaigns.

Since 2021, 15 million crypto projects have failed. The pattern keeps repeating because the market pays for noise, not clarity.

CORE3 argues for full transparency and competition built on it. That idea comes from crypto’s original ethos: trust the code, not the third parties behind it. If risk can be measured consistently, failure stops being a surprise; instead, it becomes something you can estimate in advance.

Our ultimate mission is to measure crypto risk so trust can exist.

CORE3 is currently in pilot. We’ve scored 50 projects, each reviewed through multiple Proof of Opinion assessments. The method is being refined, the next group is in progress, and the system is ready to run at scale.

What you can do now if our mission resonates to you: explore live PoL examples. See which traits separate low-risk protocols from high-risk ones. Submit a project to CORE3, expose your own gaps, or share a common risk score with people who need a clearer way to judge tradeoffs.

The open question is timing: do we put this standard in place now, or wait for the next round of losses that didn’t have to happen?

Dmytro Zaporozhchenko, CORE3 content lead, has a background in public relations for cybersecurity firms, centralized exchanges, and DeFi projects.