Methodology — how AlphaDash measures what it publishes

Every number on this site follows the same discipline: it is recorded the moment it happens, it is never rewritten afterward, and the parts we cannot defend with evidence are labeled or removed. This page explains exactly how.

Write-once, at signal time

Our core telemetry rule: information describing a decision at time T is written once, at T, and never updated. When the radar first flags a stock, we freeze the alert time, the price at that moment and the score at that moment. Live displays may keep updating through the day — the frozen record is what all published evidence is built on. This prevents the most common trick in this industry: retroactively grading yesterday's calls with today's information.

Runner Score — what it is and what the numbers mean

The Runner Score ranks live movers by their pre-explosion profile: float size, relative volume, float rotation and catalyst presence. Two honest details matter when reading its results:

1. Peaks, not profits. "Max gain after alert" is the stock's intraday high measured strictly after the first alert — never including the move that happened before we flagged it. Capturing all of it would mean selling the exact top; treat it as the size of the opportunity, not a realized return.

2. A clean evidence era. Rows recorded before 2026-08-17 used an older measurement that could include pre-alert movement — they are permanently excluded from the public record. And because a live score can rise with price action, score-banded evidence is built only on the frozen first-alert score, restarting from zero when that fix shipped. We never pool the old era with the new one.

The track record

Every alert is logged the moment it fires with its exact time and price, and it stays on the board — winners and losers. Nothing is removed, nothing is regraded. The public ledger is generated from the same database rows the system wrote at alert time.

The Day-0 watchlist

Our nightly ranking of next-session candidates is validated out-of-sample as information — the ranking identifies stocks that subsequently show unusually large intraday expansion at well above the universe base rate. It is not a validated entry strategy, and we do not publish probabilities from it: the ranking is what survived evaluation, so the ranking is what we show.

How research decides what we may claim

Behind the terminal runs a systematic research process: hypotheses are preregistered with frozen parameters before results are seen; validation requires surviving out-of-sample data, realistic costs and placebo checks; and negative results are recorded permanently. Most ideas die there — options strategies, analog-retrieval systems, alert families that showed no edge after honest measurement. When something is killed, it stays killed, and the product copy changes to match. That is why you will not find the words "high-probability" or profit promises on this site: our own research does not support them as general claims.

What we do not claim

We do not claim validated profitable entry signals. We do not convert intraday peaks into "returns our traders made". We do not show model probabilities that have not passed formal calibration review. The Options Radar is a context monitor — strikes, implied volatility, expected move — not a validated signal edge. AlphaDash is an information tool; trading involves substantial risk, and nothing here is financial advice.

Past performance does not guarantee future results. Not financial advice.