Our 59-feature model — trained on 3,791 transfers across three data sources — projects how portal players will actually perform in your team's system. Our projections average within 3 points of actual outcomes — among the most accurate transfer prediction models in college basketball analytics.
Free during beta. No signup required.
A player averaging 18 PPG in a slow half-court system might drop to 12 in your up-tempo spread offense. Raw stats lie without context.
The portal window is April 7-21. Coaches need to evaluate hundreds of players and make offers within days, not weeks.
The best transfers aren't the highest-ranked players. They're the ones whose skills align with how your team actually plays.
We pull season stats from Bart Torvik and cluster 365 teams into offensive archetypes — Perimeter Sniper, Paint Bully, Ball Control, and more.
Our Research Lab's model learns from 3,791 real transfers across 5 seasons. Not theoretical — actual before-and-after performance at the new school.
For every portal entrant, we project how they'd perform in each team archetype. You get a ranked list fitted to your system.
A player enters the portal overnight. By 7am, your staff opens MA4B and sees him already ranked against your system with a fit score, projected stats, and comparison to other available players at his position.
"Is this kid worth a call, or do we pass?" — answered in 30 seconds instead of 3 hours of film.
Your staff is deciding between two portal strategies. Add the candidates to the War Room, enter NIL costs your GM provided, and compare: "Do we spend $1.5M on two stars, or $750K on four high-fit role players?" The model shows which combination adds more total impact per dollar.
"Scenario B adds more projected impact at half the cost" — data for the budget meeting.
Your head coach ran a Perimeter Sniper system for 5 years. A 20 PPG scorer from a Paint Bully system enters the portal. Before you offer, MA4B shows that cross-archetype transfers at that PPG level regress an average of 5.9 points. The fit score is 23. You pass — and avoid a $400K mistake.
"The model saved us from a bad fit" — the transfer you didn't make matters as much as the one you did.
Every tool is free. Use the dashboard, War Room, and research to evaluate portal players and build your roster.
Live portal tracker with projections, team archetypes, and historical transfer outcomes with before/after stats.
Open DashboardIn-depth research articles, model validation, and daily portal news and reactions backed by data.
Read ResearchBuild and compare roster scenarios. Add portal candidates, see projected team impact side by side.
Open War RoomOur free tools give you the platform. Our paid reports give you the deep dive — custom analysis built specifically for your school, your roster, and your portal targets.
One-time per school
One-time per school
Your own analytics department during the portal window. Real-time projections as players enter, on-call scenario analysis for your staff, weekly roster updates, and direct access to our research team throughout the portal cycle.
Contact Us for PricingStandard and custom reports delivered within 48 hours of payment. Add $199 for 24-hour rush delivery.
Email support@ma4b.com to get started.
A player enters the portal. Within seconds, our model projects his stats at any school. Here's a real example from the 2025-26 season.
Aberdeen's profile matched what our model looks for: moderate PPG, high BPM relative to usage, same system archetype. The model predicted 12.8 PPG — his actual was 13.5.
Everything is free during beta — the portal dashboard, school reports, War Room scenario builder, team archetypes, and CSV exports. No signup, no paywall. We're focused on building the best product first.
All data comes from public sources: Bart Torvik (T-Rank), Sports-Reference, and 247Sports. We never resell raw data — everything on MA4B is proprietary analysis and projections built from these sources.
NCAA Men's Basketball only, focused on the transfer portal. We go deep on one sport rather than shallow on many.
Our 59-feature model is trained on 5 seasons of real transfer outcomes — 3,791 total transfers, 2,970 with significant playing time at both schools — using data from three sources: Bart Torvik, KenPom, and EvanMiya. Our model explains 55% of the variance in post-transfer scoring changes (R²=0.55) and projections are typically within 3 PPG of the actual outcome. For comparison, simply assuming a player's PPG stays the same after transferring is worse than guessing randomly. Only 16% of the top 50 pre-transfer scorers end up as top 50 post-transfer scorers. We believe this is among the most accurate transfer projection models publicly available in college basketball analytics.
Absolutely. The dashboard, War Room, and all reports are free during beta. Export to CSV for your staff meetings. We're actively looking for feedback from coaching staffs — email support@ma4b.com to tell us what features would help you most.
Free portal dashboard with real-time rankings, team archetypes, and historical transfer data. No signup needed.
Open the Dashboard