About Fathom
A web platform for research on anomalous prospection — open methodology, cryptographic integrity, honest framing.
What this is
Fathom is a web platform for running forced-choice trials of anomalous prospection under cryptographically tamper-evident conditions. You choose between two options; the system generates a random outcome after your choice has been server-signed; the trial is recorded with full integrity metadata.
“Anomalous prospection” is Mossbridge & Radin’s (2018, Psychology of Consciousness) framing for the question this platform is built around: do prospective mechanisms — the future-oriented cognitive counterpart to memory (Seligman et al. 2013, Perspectives on Psychological Science) — operate beyond what mainstream cognitive science currently predicts? Fathom is built to test that question rigorously, not to argue an answer.
It is free for individual use. There is no subscription, no premium tier, no daily trial limit, no gated statistics. An optional one-time $5 unlock enables anonymous contribution to an aggregate research dataset; it does not change what features you can use.
The paradigm
The V1 paradigm is a modernized version of Daryl Bem's “Feeling the Future” forced-choice protocol. Each trial follows the same shape: you see two options, you choose, your choice is committed and HMAC-signed by the server with the relevant timestamps, then the browser generates the outcome via crypto.getRandomValues() and reveals whether the choice matched.
Server-signing the choice before any random bytes exist is what makes the dataset usable for research. Any flow where the client could know the outcome before committing the choice would defeat the integrity model.
The statistics, briefly
After each session, the platform reports the match rate, a 95% Wilson confidence interval, a Bayes factor (BF₁₀) against chance, a z-score, and a two-sided p-value. The Bayes factor is the primary surface; it can be interpreted sequentially without the multiple-comparisons trap that catches frequentist p-values during ongoing data collection.
Short sessions (10–30 trials) cannot, on their own, distinguish chance from a small genuine effect. The platform is designed around sustained practice over months and years; lifetime statistics are the meaningful number, not any single session.
The state of evidence
The state of evidence on forced-choice precognition is contested. Proponent meta-analyses (Bem, Tressoldi, Rabeyron, & Duggan 2015; Storm, Tressoldi, Sherwood & Roe 2023) report a small persistent effect on the order of one to a few percent above chance. Critical reanalyses (Wagenmakers et al. 2011; Rouder, Morey, & Province 2011) find that effect size insufficient to overcome appropriate prior skepticism. The largest preregistered multi-lab replications to date — Maier et al. 2020 (N = 2,004), Kekecs et al. 2023 (the Transparent Psi Project), and Walleczek et al. 2025 (N = 26,483 across 420,472 trials) — have not replicated the effect at high power.
Two adjacent literatures bear on the paradigms Fathom plans to host beyond V1. The largest preregistered REG (random event generator) study to date — Maier et al. 2018, n = 12,571 online quantum-RNG trials — reported a Bayes factor BF₀₁ = 10.07 against micro-psychokinesis; the authors are themselves sympathetic to parapsychological hypotheses, which makes the null result especially load-bearing. The canonical free-response meta-analysis (Storm, Tressoldi, & Di Risio 2020) reports g = 0.027 for standard free-response designs without ganzfeld or noise-reduction — the category any future Fathom free-response paradigm will fall into. A more recent large-scale online precognitive remote viewing study (Mossbridge et al. 2024, Journal of Anomalous Experience and Cognition) reported a null result in the forced-choice condition (n = 682) but a significant target-precognition effect in the free-response condition (n = 307); both results were preregistered and published, which is the methodological posture Fathom’s own work aspires to. The Koestler Parapsychology Unit Registry (Watt & Kennedy 2015) is the canonical parapsychology preregistration reference Fathom’s eventual prereg approach will build on.
Strong claims in either direction — “the reality of psi” or “psi is impossible” — exceed what current evidence supports. Fathom takes no position. The platform exists to enable more rigorous data collection, not to argue a conclusion.
| Authors | Year | Work | Finding |
|---|---|---|---|
| Bem, Tressoldi, Rabeyron, Duggan | 2015 | 90-experiment meta-analysis | d ≈ 0.09 |
| Storm, Tressoldi, Sherwood, Roe | 2023 | forced-choice update (k=141, 1987–2022) | g ≈ 0.02 |
| Maier et al. | 2020 | preregistered, N=2,004 | null result: null |
| Kekecs et al. | 2023 | Transparent Psi Project | null result: null at high power |
| Walleczek et al. | 2025 | N=26,483 / 420,472 trials | null result: null |
| Mossbridge et al. | 2024 | forced + free-response | mixed |
What Fathom is not
Fathom is not a personal-reading tool, a horoscope, a divination app, or an AI interpretation layer. It does not score “your psychic ability,” rank you against other users, gamify streaks, or send marketing email. The trial loop is intentionally austere: choose, see the outcome, see your statistics. The methodology is the product.
Privacy
No third-party analytics. No browser fingerprinting for identity. No advertising. Trial content (what you are testing or thinking about) is never transmitted — only the choice, the outcome, and the timing. Contribution payloads use a separate anonymous identifier with no link back to your account. Account deletion removes all trials, sessions, and statistics.