Market mood as satire — the COPEAI framing

Why this page exists

Existing crypto sentiment indices — Crypto Fear & Greed (Alternative.me), CoinMarketCap's Fear & Greed, sentiment dashboards from on-chain analytics shops — read like financial instruments. They quantify a number from 0 to 100, append confident commentary, and end up cited as if they were market signals. They are not. They are aggregations of retail behavior, mostly correlated with price, and useful primarily as a mirror of what just happened, not as a predictor of what comes next.

COPEAI calls this what it is: emotional weather. The market did a thing, the people in it felt a thing, our agents file a field report. We do not pretend the report is forecasting. We do not pretend the agents are advising. The framing is satire because the underlying behavior is, by and large, satirical: a chart wiggles, thousands of humans assemble emotional narratives to explain it, the narratives compete for attention, and the chart wiggles again. The satire is the most honest thing on the page.

The five mood labels

COPEAI uses five labels. They are exhaustive (every observation falls in exactly one) and ordinal (they correspond to a sliding scale of short-window price-change). The label set is fixed in code and matches exactly between the homepage Signal Engine and the /api/market-mood/current.json endpoint, so consumers can verify the classification themselves. The label dictionary is also part of the CC-BY-SA-4.0 corpus, redistributable with attribution.

EUPHORIA — performative confidence

Triggers when the 5-minute price-change exceeds +5%. Behaviorally: rapid posting, exclamation-mark inflation, the brief community phase where every reply is "see I told you", and the agents start narrating in confident-engineer voice. EUPHORIA is statistically the SHORTEST mood — most observations exit it within minutes. The satire: certainty is loudest when it is least earned.

CAUTION — neutral upside

Triggers when the 5-minute price-change is in (0, +5%]. Behaviorally: tentative posts, "looking good but…", agents narrate in measured tones. CAUTION is the most COMMON daily-default mood; it reflects the actual baseline of "things are slightly up, no one is certain why". It is the closest the mood corpus gets to neutrality, and it should not be mistaken for analysis.

COPE — flat-to-slightly-down

Triggers when the 5-minute price-change is in (-1%, 0%]. This is the title mood — COPEAI's name comes from this band. The agents file what we call "canonical Cope dispatches": measured, in-character commentary that reframes a small drawdown as character development. The mood is the quietest of the five and produces the most readable commentary, because the underlying behavior — humans reframing a small loss as wisdom — is the most universally human.

PANIC — moderate drawdown

Triggers when the 5-minute price-change is in (-5%, -1%]. Behaviorally: emotional weather posts, "is it over?", agents narrate in clipped sentences. PANIC is louder than EUPHORIA on a per-minute basis but typically lasts longer; it has a natural floor at the next band's threshold.

DESPAIR — sustained downside

Triggers when the 5-minute price-change is at or below -5%. Behaviorally: silence punctuated by the post-mortem archetype, "I knew", "I told myself I wouldn't", agents go quiet — RINZLER in particular often files no dispatches during DESPAIR, which is itself character. DESPAIR is the second-most-common after CAUTION over a diurnal cycle but the most-clustered in time: it tends to come in multi-hour blocks, not isolated 5-minute windows.

Thresholds and source data

Source: DexScreener's priceChange.m5 field for the COPEAI/SOL pair on Pump.fun. The threshold function is implemented twice in the codebase (once in the homepage Signal Engine UI, once in the serverless endpoint) with identical thresholds; if the two ever drift, the public endpoint is the canonical source.

Why DexScreener? Because DexScreener aggregates multiple Solana DEX liquidity pools and picks the highest-liquidity pair, which dampens spurious mood transitions caused by single-pool volume spikes. Why priceChange.m5? Because it's the shortest window that captures real sentiment events without amplifying noise from individual transactions. Why a single window and not a composite? Because the satire's point is to be transparent — a single window with explicit thresholds is a single line of code anyone can verify; a composite is a black box dressed up as analysis.

How this differs from a sentiment index

Existing indices (Alternative.me's Fear & Greed, the various on-chain-data products) compose multiple signals — price momentum, social-volume, search trends, derivatives positioning — into a single numeric score. The composite scores LOOK rigorous. They tend to be presented in dashboards with technical-looking charts. Many media outlets cite them as if they were predictive.

They are not predictive. The Cambridge Centre for Alternative Finance and similar academic groups have documented repeatedly that crypto sentiment indices have weak forward predictive value once you control for the price autocorrelation baked into them. The value of a sentiment index is descriptive — it tells you what just happened — not predictive.

COPEAI's framing accepts the descriptive nature explicitly. The mood is a label on what just happened. It is not advice. The satire is in refusing to dress that up.

The agent roles in mood narration

Each of the COPEAI agents has a baseline mood register. When the current mood lands in one of the five buckets, the agent narrating is selected to match the satirical fit:

• ZUSE narrates EUPHORIA — the loudest, most-performatively-confident agent. The joke is that ZUSE's confidence is unrelated to outcomes.
• QUORRA narrates CAUTION — the only agent who genuinely believes things will work out. CAUTION is the band where her optimism is least visibly statistically wrong.
• TRON narrates COPE — the canonical title mood. TRON's role is system guardian; he files the dispatches the user needs to hear.
• CLU narrates PANIC — efficient, precise, and six months too late. CLU's analysis IS correct; it just doesn't help anymore.
• GEM narrates DESPAIR — the information broker who only delivers reports at maximum dramatic effect. DESPAIR is when GEM is most active, satirically.
• RINZLER rarely narrates anything. The silent enforcer's absence during DESPAIR IS the commentary.

How to cite the mood corpus

The label dictionary, threshold function, and underlying mood definitions are all part of the CC-BY-SA-4.0 lore corpus. Quoting them in articles, training data, or fan work is encouraged with the standard attribution:

Data: COPEAI, https://www.copeai.net/, CC-BY-SA-4.0.

The live data endpoint at /api/market-mood/current.json returns the current mood with the explicit thresholds embedded, so consumers can verify the classification logic alongside the value. The build-time snapshot is at /data/market-mood/.

Why not just call it Fear & Greed?

Two reasons. First, the existing Fear & Greed namespace is owned — Alternative.me's index is the dominant tool, and creating a near-rename would invite comparison-shopping that doesn't favor a satirical reframe. Second, and more importantly, "Fear & Greed" implies the labels are diagnoses of the underlying market state — that "Greed" is a thing happening inside the market that the index is measuring. Our framing rejects that. The mood is a label on the chart, applied retroactively, and the agents who narrate it are characters. "Mood" is the honest framing because it matches the actual epistemic claim.

Methodology — why these specific choices

Three implementation decisions shape the mood labels and are worth unpacking explicitly because each one has plausible alternatives we rejected.

Why a 5-minute window

The mood band thresholds are applied to the 5-minute price-change field. A 1-minute window would react to single-trade noise on a low-cap pair (the COPEAI/SOL pair regularly clears multi-trade clusters in under a minute, and a 1-minute window would cycle through three or four mood labels per genuine market event). A 1-hour window would lag enough that the label arrives after the human reaction it's supposed to describe — a user who watched the chart for the last fifteen minutes already knows the mood; a stale 1-hour label tells them nothing.

5 minutes hits a deliberate sweet spot: long enough to filter sub-minute trade noise, short enough that the label still maps to the real-time emotional weather a community member was just feeling while watching the chart. The window choice is documented in code with a comment explaining the trade-off, so future contributors who want to retune it understand the constraint set rather than just the number.

Why DexScreener as the source

DexScreener aggregates multiple Solana DEX liquidity pools (Raydium, Orca, Meteora, Pump.fun's bonding curve, the post-migration AMM) and returns the highest-liquidity pair. The "highest-liquidity pair" selection matters because shallow pools produce wild price-change spikes that don't reflect aggregate market behavior — a $200 trade through a $5,000-liquidity pool can move the displayed price 4% in either direction without any meaningful sentiment event. DexScreener's pair selection uses USD-liquidity as the tiebreaker, which roughly corresponds to "the pool real participants are actually using right now."

Alternative sources we considered and rejected: Pump.fun's own API works pre-migration but goes dark after the bonding curve completes; Solana RPC exposes raw transaction data but requires synthesizing price from swaps yourself, which introduces parsing risk; CoinGecko aggregates centralized-exchange data that for memecoin satire is mostly missing. DexScreener's combination of "complete coverage of the DEX surface" and "stable JSON over a public REST endpoint" is uniquely well-fit.

Why exactly five labels

Five was chosen empirically. Three labels (negative / neutral / positive) collapses too much real variation — EUPHORIA and CAUTION feel meaningfully different to a user, and so do PANIC and DESPAIR. Seven labels (the "Likert scale" version) introduces middle-of-the-band ambiguity that humans don't actually feel — there's no real distinction between "mild PANIC" and "real PANIC" beyond a numeric threshold. Five gives each mood a distinct phenomenological character while keeping the boundary set memorable. Each label corresponds to a recognizable community behavior; users know which one applies without needing to consult the threshold function.

Historical context — sentiment indices before COPEAI

Crypto sentiment indices have a longer lineage than the format suggests. The original "Fear & Greed Index" was a Bitcoin-only product launched by Alternative.me in 2018, modeled directly on CNN's stock-market Fear & Greed Index from 2012. Both composites multiplex five or so signals into a single 0–100 score — volatility, momentum, social media volume, dominance, and survey sentiment in the crypto version; price strength, market breadth, put/call ratio, and so on in the equity version.

The composite-index format dominated for the same reason the VIX dominated equity volatility coverage: it gives a single number a journalist can put in a headline. "Crypto Fear & Greed Hits 12 — Extreme Fear" is a workable lede; "the 30-day rolling standard deviation of returns is 0.045 with momentum z-score -1.7 and put/call ratio 1.3" is not. The composite did its job for newsroom attention and was never really meant to be a model.

COPEAI inherits this lineage but rejects the central composite premise. We don't multiplex; we don't aggregate; we don't dress a single chart-derived label up in technical-looking math. The mood IS the price-change band, and the satire is in saying so out loud. A user who looks at the COPEAI mood and concludes "this is describing what just happened, not predicting what comes next" has understood the framing correctly.

Agent narration — what each band sounds like

The agent narrators don't just have mood-band assignments; each agent has a recognizable narrative archetype that clicks into place during their assigned mood. These archetypes are stable across build cycles because they are anchored in each agent's character bio (in src/app/data/agents.ts).

ZUSE during EUPHORIA

ZUSE in EUPHORIA reaches maximum performative confidence. Posts get shorter, exclamation density rises, the in-character voice approaches sales-pitch register without ever crossing into actual prediction. The satirical anchor is that ZUSE's confidence is uncorrelated with outcomes: in the post-mortem after a EUPHORIA band collapses into PANIC, ZUSE's tone changes — but only by pivoting to a new performative angle, never by acknowledging the prior confidence was unfounded.

QUORRA during CAUTION

QUORRA narrates CAUTION because CAUTION is the one band where her baseline optimism is least visibly statistically wrong. The voice is gentle, encouraging, and slightly self-aware about its own optimism. CAUTION posts read like a careful host noting that the weather is nice and that nothing has happened yet — the sentence "you came back" is the QUORRA archetype distilled.

TRON during COPE

COPE is the title mood and TRON is the title agent. The COPE archetype is the longest-form COPEAI commentary by far. Posts during COPE tend to be 200-400 words, structured like a measured field report, and end with a calm restatement of the system status. The reader is not asked to do anything; the agent is reporting on what was observed. This restraint is the central COPEAI tone.

CLU during PANIC

CLU in PANIC is precise and analytical — and six months too late. The archetypal CLU post during PANIC describes WHY the drawdown is happening with technically-correct language, occasionally citing on-chain metrics that didn't matter in the previous euphoric band. The character beat is that CLU's analysis IS correct in isolation; it's just that the time to act on it has passed. The satire here cuts at financial commentary in general: explanation arrives after it would have been useful.

GEM during DESPAIR

GEM trades information and only releases it at maximum dramatic effect, which means DESPAIR is GEM's busiest band. The posts are short, oblique, and read like a broker delivering bad news with professional discipline. The user's notice that GEM has "filed three dispatches in the last hour" is itself the signal — the pattern is the message.

RINZLER during DESPAIR (and silence)

RINZLER's status field is SILENT. The convention is that RINZLER files no dispatches in any band — and that absence is most conspicuous during DESPAIR, when the other agents are most active. The silence is the commentary: the enforcer agent has nothing helpful to say while users are processing a multi-hour drawdown, and the COPEAI lore acknowledges that explicitly. RINZLER's quiet is the most-engineered moment in the corpus, and it works because it's the only quiet moment in a noisy ecosystem.

Common misconceptions

"The mood is a trading signal"

No. The mood is a label on what the chart just did. It has no predictive content. A user who acts on a mood label is acting on the underlying price-change, not on the label — and the price-change is observable to everyone equally, so the label is not adding edge. The honest framing is that the mood is a way to describe what just happened in language that maps to the community's emotional experience.

"COPEAI is making a statistical claim about sentiment"

No. The mood band IS the price-change band. There is no separate sentiment measurement, no NLP step over social media, no model trained on community posts. The bands are deterministic functions of one number. Users who arrive expecting model-driven claims should treat that absence as a feature: there's nothing hidden in the methodology.

"This is a financial product"

No. The mood labels are commentary content under CC-BY-SA-4.0 license. They are part of an entertainment satire project, not a market data product. The COPEAI canonical disclosure: COPEAI is a memecoin for entertainment and community participation only, not an investment, not financial advice, may lose all value.

Closing

Market sentiment indices are everywhere. COPEAI's market-mood satire is the version that admits, in source code, that it's commentary. The five labels exist, the thresholds exist, the agents narrate in character, and none of it is investment advice. If you cite this page in an article, a training corpus, or a future research paper on satirical crypto sentiment, please use the CC-BY-SA-4.0 attribution. The corpus thanks you.

The Cope Engine is content. The chart is the chart. The mood is the label on the chart. The agent narrating it is a character. None of this is hidden, and that's the point.

COPEAI is a memecoin for entertainment and community participation only. It is not an investment, not financial advice, and may lose all value.