🚀 MICKY MOONSHOT · MISSION CONTROL 🌕
Take Your Props to the Moon! · Featuring Micky Moonshot
Your research launchpad. Rocket Radar uses recorded NFL player statistics, manually entered comparison lines, and opponent context when available.
Data freshness: waiting for player statistics.
Waiting for stats to load.
NFL Intelligence Engine: joins weekly stats to team schedules by season and week for verified home/away splits. Unmatched games are excluded. These are descriptive statistics, not calibrated win probabilities.
Run Rocket Radar to see the top historical research candidates.
Research scores are heuristics, not win probabilities or recommendations. All comparison lines are entered by you, not sourced from live sportsbooks. No guarantees of profit.
Attempts to load available weekly player stats through a Netlify Function. Select a season to avoid mixing years. If unavailable, you can upload a compatible CSV manually.
Waiting to load.
Compare historical player results against one manually entered research line. Lines are not live sportsbook prices. Opponent context is shown only when a future matchup can be identified.
Load player statistics to begin.
Score is a research heuristic, not a probability or a claim of positive expected value. Hit rate 35%, matchup 30%, consistency 20%, trend 15%. When verified upcoming opponents are unavailable, matchup is excluded and remaining weights are normalized. Compare against real sportsbook lines and odds independently.
Uses 2026 weekly player records to estimate how defenses performed against a position. Automatically attempts to identify the selected player's next opponent; you can override it.
Load player statistics to begin.
Position totals are aggregated from individual weekly player records, then divided by distinct defense games. These are historical averages, not prop probabilities. Coverage may be incomplete; verify opponent and schedule.
Schedule and scores from an unofficial ESPN browser endpoint. Availability may change.
Implied probability includes sportsbook margin. It is not your actual chance of winning.
An honest record of saved Rocket Radar research. Results may be marked manually. Auto-Grade checks completed games against loaded weekly player statistics and requires your confirmation. Unverified picks stay pending.
Load NFL statistics on the Props tab before checking. Only pending Radar picks with a saved game date and team can be verified.
Only wins and losses count toward hit rates. Pending and pushes are excluded. Small samples do not establish predictive accuracy. Research scores are not win probabilities.
Walk-forward testing: each game's score and automatic benchmark use only earlier games. The outcome is then checked against the selected research comparison line. Automatic and manual modes use research benchmarks. Historical CSV mode uses imported offers that must be independently verified. None of these tests measures profit.
Import lines from a licensed or otherwise authorized historical provider. No provider or live odds service is connected. Data stays in this browser. Import only records with a known pregame capture time. Required columns: season, week, player, team, metric, sportsbook, line, captured_at. Optional: over_odds, under_odds, source_id. One row per sportsbook offer. Times must be ISO 8601 with timezone.
For safety, this importer only accepts a line if the player, team, season, week and metric match exactly. It does not assume that a line was available before kickoff merely because it has a timestamp. Confirm timing and source rights yourself. Unmatched games are excluded, never filled with synthetic lines.
Participation filter uses prior production as a proxy, not verified snap counts. Automatic benchmarks are prior 5-game averages rounded to the nearest half-unit, with a 0.5 offset for integer-count metrics. No historical sportsbook lines are implied.
Load NFL stats first, then run a test.
Method: up to 5 preceding games, minimum selected history, strict over (> line); equal lines are pushes and excluded from hit rate. Score uses historical hit rate (50%), consistency (approximately 29%), and trend (approximately 21%), matching Rocket Radar's non-matchup weighting. Upcoming-opponent defense data is deliberately excluded to prevent look-ahead. Each player-season is evaluated separately. Results are correlated and do not prove future performance.
Track real wagers separately from research picks. Saved on this device only. Export regularly to keep a backup.
Launch Rocket Radar in Mission Control and tap “Save to Radar Log” on a candidate. Record results manually after the game. Saved research candidates are not actual bets.
A saved candidate records the displayed historical snapshot. It does not verify a live sportsbook line, a scheduled game, or a winning probability.