Invite only. Blackbox. No public source.
Pure Elo ratings. Seven sports. Platt-calibrated probabilities.
MechaPip rates seven leagues with Elo, then calibrates those ratings into win probabilities using Platt scaling. Player form, team advanced stats, injuries, and starter data enrich each matchup. Daily board on mechapip.com. The engine is private. Access is invite only. This page is not a clone kit. Forecasts are not financial or gambling advice.
Live board Full model explainer
Seven sports. Seven tunings. One HUD.
Same engine. League-specific rest, travel, starters, and sample size.
NBA
82-game grind. Back-to-backs, altitude, star load. Rest deltas stay in the rating path.
NFL
Short season, fat K. Weather, byes, and QB-out flags punch above their sample.
MLB
162 games, starter swings, park factors. Lowest noise, longest memory.
NHL
Goalie sub-ratings. Travel and B2B matter more than a hot stick.
WNBA
Compact schedule, travel density, rest clustering. Same Elo math, tighter window.
NCAAF
Roster churn every year. Heavy regression, FCS/FBS context, Saturday volume.
NCAAMBB
Hundreds of teams. Transfer portal noise. Conference strength baked into the prior.
How it actually works
Elo base
Rating per team with data enrichment layers (player form, team advanced stats, injuries, starters). Sport-tuned home edge at forecast time. Platt calibration maps raw Elo expected scores to calibrated win probabilities.
Data enrichment
Player form ratings, team advanced composites, injury impact, and starter quality adjust base Elo at forecast time. All enrichment is opt-in per sport via pregame flags.
Three nodes
Stream PC paints the HUD. Windows Spot runs the morning bot. Weekly Spot box retrains. S3 is the bus. No second live bot.
Paper first
Live money stays off until an operator confirm. This site does not sell event contracts.
Elo
Each team has a single rating that updates after every game. Enrichment layers adjust ratings for player form, team advanced stats, injuries, and starter quality. Home edge applies at forecast time. Platt calibration maps the raw expected score to a calibrated win probability.
Data enrichment
Player form ratings, team advanced composites, and injury/starter impact adjust base Elo at forecast time. All enrichment is opt-in per sport. The engine is pure Elo underneath.
Platt calibration
A logistic scaler fitted on walk-forward predictions maps raw Elo expected scores to well-calibrated probabilities. Each sport gets its own Platt scaler from the optimizer.
The house behind the board
MechaPip is the sports desk of Phipps Predictions — a small systematic house with one question in front of every number: is this still allowed to live? Rankings before slogans. Paper before live. Seven sports always on the board.
Elo Star
Head of the Sports Desk · MechaPip
"What wound does the crowd refuse to rank correctly?"
Reika Kurogane
Founder & Chief Investment Officer
"What are we allowed to want?"
Yui Aizawa
Head of Quantitative Strategies
"What wound does this exploit?"
Rin Takamiya
Chief Risk Officer
"How is this allowed to die?"
Niko Hoshino
Lead AI Systems Architect
"Where does it live, and what must it never touch?"
Mei Sakuraba
Global Macro & Sentiment Lead
"What change in the world makes this obsolete?"
Aria "Blitz" Kanzaki
Head of Execution & High-Frequency Trading
"What is the click?"
The house charter lives with the private engine. The board, the forecasts, and every public post stay MechaPip.
Access is invite only
The engine is a closed blackbox. No public source. No public clone. Invited operators get the board and the runbook. Everyone else watches forecasts on mechapip.com.
Business inquiries: mechapip@mechapip.com
Live forecasts stay on www.mechapip.com. This host does not replace the board.