CHA
127
MIA
126
PHX
110
POR
114
PHI
109
ORL
97
LAC
121
GSW
126
ORL
CHA
PHX
GSW
Team • Position • MIN
Team • Position • MIN
0 of 30 Teams selected
Loading NBA News...
Wähle Teams aus, um nur deren Spiele anzuzeigen
6thman is a free, non-profit fan project built for NBA analytics enthusiasts who want to understand the game at a deeper level. This isn't about quick picks or shortcuts—it's about learning how NBA predictions really work.
We believe in transparency and education. Instead of hiding behind black-box algorithms, we show you:
No authentication, no paywalls, no hidden agenda—just pure NBA analytics for fans who love the game.
Select two teams, choose your starting five players for each, and let the models analyze the matchup using:
Professional organizations employ thousands of analysts with decades of experience and sophisticated models. Even they only achieve 65-70% accuracy in predicting game outcomes.
NBA games are highly complex. This means most obvious patterns are already well-known. If you think you've found an "easy system," chances are you're missing something.
This app is built to educate, not to guarantee certainty. Understanding why predictions are difficult is more valuable than chasing false confidence.
We combine multiple data sources and modeling approaches:
We calculate player impact ratings based on Neil Paine's legendary RAPTOR formula. RAPTOR (Robust Algorithm using Player Tracking and On/Off Ratings) measures a player's contribution across offense and defense, adjusted for teammates, opponents, and game context.
Points, rebounds, assists, field goal percentage, and more—the fundamental stats that still matter.
ELO ratings, recent win/loss records, and strength of schedule calculations.
We train and compare multiple ML algorithms:
All data used in our calculations comes from legitimate sources:
When a model says "Team A has a 70% chance to win," does that mean they win exactly 70% of the time in similar situations? Calibration tables answer this question.
We group predictions into bins (e.g., 50-55%, 55-60%, etc.) and compare:
A well-calibrated model should have mean probability ≈ actual accuracy for each bin.
Below you can see actual calibration tables from our models. These are the real numbers showing how well our predictions match reality.
In the Historical Analysis section, you can see:
We don't cherry-pick results. We don't hide failures. Every prediction is logged and visible—that's what reproducibility means.
Understanding calibration helps you interpret predictions correctly:
This project is NOT affiliated with, endorsed by, or connected to the NBA, NBA teams, or any official NBA entities. We are an independent fan project.
Team names, logos, and trademarks are used solely for reference and identification purposes. All rights to these marks are owned by their respective organizations.
6thman is a completely free, non-profit project:
This was built purely as an educational tool for fans who love NBA analytics.
This app helps you explore how analytics shape the game of basketball. Data is updated daily and sourced from publicly available statistics.
All data used in this project was acquired through legitimate means:
This software is provided "as is" without any warranties, express or implied.
This project builds on the shoulders of giants in NBA analytics:
The RAPTOR (Robust Algorithm using Player Tracking and On/Off Ratings) formula was created by Neil Paine, a legendary figure in NBA analytics.
We are not affiliated with Neil Paine, but we are huge fans of his work and proud subscribers to his Substack:
The RAPTOR formula is publicly documented and we calculate our own RAPTOR metrics using our data sources.
Player movements, shot quality, defensive metrics, and detailed performance data acquired through private 3rd-party analytics providers at personal cost.
Box score stats, player rosters, team records, and historical data from publicly available sources.
This project was built with:
This project represents countless hours of:
All funded personally, with no sponsorships or monetization. Just pure love for NBA analytics.
This is a fan project and we're always looking to improve. If you have feedback, suggestions, or find bugs, we'd love to hear from you!
Halil Zabun
Email: contact@6thman.app
Halil Zabun
Email: contact@6thman.app
The contents of this website have been created with the utmost care. However, we cannot guarantee the accuracy, completeness or timeliness of the content.
This website contains links to external third-party websites over whose content we have no control. The respective provider is always responsible for the content of linked pages.
Content and works created by the site operator on these pages are subject to German copyright law. NBA, team names and logos are registered trademarks of their respective owners.
This privacy policy informs you about the nature, scope and purpose of the processing of personal data on this website.
Halil Zabun
Email: contact@6thman.app
This website uses Google Analytics 4 (Google LLC). Google Analytics uses cookies and transmits usage data (including IP address, page views, session duration) to Google servers. IP addresses are anonymized before transmission. Legal basis: Art. 6(1)(f) GDPR. Opt-Out: Google Analytics Opt-Out Add-on
This website uses localStorage/sessionStorage exclusively for technically necessary functions (e.g. saving settings). No personal data is stored.
The hosting provider (Firebase/Google Cloud) automatically collects server log files transmitted by your browser (IP address, browser, operating system, time of access).
Google is certified under the EU-US Data Privacy Framework.
You have the right to:
Contact: contact@6thman.app
We reserve the right to update this privacy policy as needed. Last updated: March 2026
When you open the app, you automatically land on today's date with the first matchup of the day. The most frequently used starting lineups for each team in the current season are pre-selected for you.
You can still customize the lineups using the dropdown menus to select different players for each position.
Use the orange "Today's Games" section to:
The team header cards are clickable! Click on any team to view:
In the team game history modal, you can:
Click the "Roster" button on any team card to view:
After analyzing a matchup, you'll see predictions from different ML model categories:
Each category shows:
Below the matchup analysis, you'll see the Top Starting Five Combinations for each team:
Click on "Click to view all combinations" to see detailed lineup statistics:
For each combination, you can see:
In the Historical Analysis modal, you can explore:
Once you've selected both lineups, click "Start Analysis" to see:
Click "Historical Analysis" in the header to: