ECHODEPTHSPORTS
Book Demo →
Alternative Data · Sports Betting · Quant Trading

Pre-Match Manager Score:
the signal the market doesn't price

Bookmaker pricing models use form, injuries, H2H, weather, and referee patterns. None of them use the manager's emotional state 24–48 hours before kick-off. We do. It's orthogonal to every signal in your model.

Request Data Briefing →See hypothesis & methodology

Why it matters

Four reasons the signal is commercially viable

01

Orthogonal to existing signals

Manager facial AU patterns have zero correlation with GPS load, injury reports, historical H2H, or weather. It adds independent signal — not noise on top of what you already have.

02

Involuntary signal

A manager briefing media cannot control which specific AU combinations they produce. Unlike team news or tactics (carefully managed), the facial signal cannot be gamed for public consumption.

03

24–48hr pre-match availability

Press conferences typically occur the day before or morning of the match — after main markets open. The signal arrives in the window between generation and market movement.

04

Scalable to 40+ competitions

Any league with publicly available press conference video can be scored. Premier League first. Bundesliga, La Liga, Serie A, Champions League are natural extensions.

The Data Product

What you receive, when, and how

Delivered as JSON via REST API or CSV push. Target: 2–6 hours post press conference, before match-day open.

Data fields per match

→match_id, competition, date
→manager_a_genuine_confidence (0–1)
→manager_a_instability_index (0–1)
→manager_a_net_confidence (−1 to 1)
→manager_b_genuine_confidence (0–1)
→manager_b_instability_index (0–1)
→manager_b_net_confidence (−1 to 1)
→divergence_flag (boolean)
→market_implied_prob_home / away
→confidence_vs_market_gap (delta)
FormatJSON via REST API or CSV push
Timing2–6hrs post press conference
Phase 1 coveragePremier League + Championship
Phase 2 coverageBundesliga, La Liga, Serie A, UCL
History2023/24–present backtesting dataset
ExclusivityNegotiable at launch
PartnershipAnnual feed or rev-share on alpha

Validation dataset in development. Partnership pricing set once statistical significance is established.

Built for

Trading desks, quant funds, and alt data aggregators

Bookmaker Trading Desks

Pricing models use form, injuries, H2H. Manager confidence is orthogonal to all of these. Even marginal predictive signal at scale is commercially significant.

Sports Betting Exchanges

Exchange matching engines can't see non-priced signals. If manager confidence divergence predicts sharp money movement, it's actionable pre-movement.

Quant Sports Funds

Alternative data with no correlation to existing factors is the only kind worth adding. Manager emotional state has zero overlap with standard sports analytics datasets.

Alt Data Aggregators

Exclusive dataset at launch. Early distribution partnership creates first-mover advantage across your existing client base.

Questions from quant teams

Is the signal statistically validated?+
Not yet at publishable significance — we have directional signal from a 312-match backtesting dataset. The data partnership is structured to co-develop validation. Partners get access to raw scoring data and can run their own backtesting. Pricing is set when the signal proves out, not before.
Is analysing public press conferences legally compliant?+
EchoDepth operates under a Legitimate Interests basis (Article 6(1)(f) UK GDPR). Press conferences are staged, publicly broadcast media events — managers acting in their professional public capacity have a reduced privacy expectation in this specific context. Processing is for analytical and editorial purposes. AU analysis measures emotional state, not identity — EchoDepth does not engage Article 9 biometric identification processing. A documented LIA and DPIA are on file. All data feed subscribers operate under a data use agreement that prohibits automated decisions affecting individuals and requires their own legal basis assessment for any downstream processing. ICO registered ZB915633.
What is the go-to-market if the signal validates?+
Phase 1: annual data feed subscription per competition. Phase 2: API with per-event pricing. Phase 3: integration with existing sports data platforms (Sportradar, Genius Sports, Stats Perform). Revenue share on demonstrated alpha available for early partners.

Data & Legal

Lawful basis
Article 6(1)(f) UK GDPR — Legitimate Interests. LIA ref: LIA-ED-SPORT-001.
Article 9 position
AU analysis measures emotional state, not identity. Not biometric identification processing. DPIA ref: DPIA-ED-SPORT-001.
Data subjects
Professional public figures in their professional capacity. Reduced privacy expectation in staged broadcast contexts.
Subscriber obligations
Data feed subscribers must assess their own legal basis for downstream processing. Automated decisions affecting individuals prohibited under subscriber agreement.
ICO registration
Cavefish Ltd: ZB915633. Verifiable at ico.org.uk.
Object to processing
Data subjects may object under Article 21 UK GDPR. Contact: hello@cavefish.co.uk.

Request a data briefing

30-minute call covering methodology, backtesting dataset, and partnership structure. No commitment. No pitch. If the signal doesn't interest you after 30 minutes, you've lost nothing.

Email us directly at hello@cavefish.ai with your organisation and role, or complete the form on the Manager Intelligence page →

Include: organisation, your role (trading desk / quant / data partnerships / fund), and what would make this signal commercially relevant to you.

How does emotion analytics support sports trading desk decisions?