I made a machine learning model to predict 66.45% of NBA games

In plain words

DeepShot is an NBA game prediction tool that forecasts game outcomes with 66.45% accuracy using machine learning models trained on historical basketball data. It combines team rolling averages, exponentially weighted moving averages, and Elo ratings to analyze matchups, then displays predictions through a web interface built with NiceGUI. The tool is designed for sports fans, data analysts, and developers interested in sports prediction models and machine learning applications.

written from the facts on this page · September 2026

From the sources

DeepShot is a machine learning model designed to predict NBA game outcomes using advanced team statistics and rolling averages. It combines historical performance trends with contextual game data to deliver highly accurate win predictions (66.45%)

AI-powered NBA game outcome predictor that uses advanced team stats and trend-based features to forecast winners and track model performance - saccofrancesco/deepshot

DeepShot is a machine learning-based NBA game predictor using advanced rolling stats (like EWMA) and real historical performance. It helps forecast matchups with visual insights and a clean interactive GUI. git clone https://github.com/saccofrancesco/deepshot.git cd deepshot pip install -r requirements.txt # Train model by running the notebook # Open `model.ipynb` and run the cell to generate `deepshot.pkl` python main.py # Launches the NiceGUI web app 📬 Emailware: Share Your Thoughts DeepShot is emailware . If it helps you or you find it interesting, I’d love to hear from you! This project is licensed under the MIT License — feel free to use it in your ownfrom github.com

In the maker’s words, at launch

Introducing DeepShot: An NBA Game Prediction Model Hey devs, sports fans, and data nerds! After weeks of work, I'm excited to share DeepShot – an advanced NBA game predictor powered by historical data from Basketball Reference, machine learning, and a clean NiceGUI-powered web interface. What it does: DeepShot uses…

• 
  A basketball hoop to maximize shots that go in [video]2020 · youtube.com · ▲971

  https://github.com/OlexG/finclaw ^ repo if anyone wants to contribute. I realized that sometimes I'm pretty good at making random guesses about the future. (I.E, I think OpenAI will beat Anthropic in the short term, etc...) However, I'm bad at converting these guesses into actual market plays I can profit from. As such, I wanted to make a tool that connects to the right APIs and launches a swarm of agent experts to generate related plays based on your prediction. The very first super shitty V0 MPV just has Polymarket/Kalshi bets, but I will be adding a more complex…

• 
  ShotTracker2014 · shottracker.com · ▲79

  Boosts basketball players' shot by measuring performance

• 
  Applying Machine Learning to March Madness2019 · github.com · ▲97

  Machine learned bracketology. Contribute to adeshpande3/March-Madness-ML development by creating an account on GitHub.

• 
  Deep Play-by-Play – labelling NBA plays using a 3D ConvNet2018 · github.com · ▲8

  Labelling NBA action using deep learning :basketball: - GitHub - neeilan/DeepPlayByPlay: Labelling NBA action using deep learning

• 
  AI Sports Prediction2022 · copascore.ai · ▲62

  Machine Learning sports data forecasting

• 
  GameFocus AIDec 2025 · gamefocusai.com · ▲5

  AI-powered NBA player prop - transparent & educational

• 
  Spira Maxima▲578

  Video model that turns script into viral social media videos

  AI · 19d ago · spira.ai

• 
  Rill Browser▲540

  The browser where Claude Code and Codex work beside you

  AI · 19d ago · rill.love

• 
  Germany's new sovereign AI model Kolibri▲424

  Kolibri is Aleph Alpha

  AI · 21d ago · tej.as

• 
  Giving Opus 5.5 a simulated paint canvas▲392

  Paintings by AI models: programs run through a simulation of oil paint.

  AI · 22d ago · stillwet.art

• 
  Beam.win▲379

  Run AI agents on one Mac and use its windows on another

  AI · 5d ago · beam.win

• 
  FastRouter.ai▲364

  Route requests to the right LLM for cost, latency & quality

  AI · 23d ago · fastrouter.ai

• 
  I built a hardware processor that runs Python▲983

  Hi everyone, I built PyXL — a hardware processor that executes a custom assembly generated from Python programs, without using a traditional interpreter or virtual machine. It compiles Python -> CPython Bytecode -> Instruction set designed for direct hardware execution. I’m sharing an early benchmark: a GPIO test where PyXL achieves a 480ns round-trip toggle — compared to 14-25 micro seconds on a MicroPython Pyboard - even though PyXL runs at a lower clock (100MHz vs. 168MHz). The design is stack-based, fully pipelined, and preserves Python's dynamic typing without static type restrictions.…

  Dev tools · 2025 · runpyxl.com

• 
  Unsure Calculator – back-of-a-napkin probabilistic calculator▲930

  The Unsure Calculator is an online tool that lets you calculate with numbers you’re not sure about.

  Life & fun · 2025 · filiph.github.io

• 
  I built an AI that turns GitHub codebases into easy tutorials▲923

  https://the-pocket.github.io/Tutorial-Codebase-Knowledge/

  AI · 2025 · github.com

• 
  Whatting▲898

  The iPad diary, just for you

  Work · 2025 · whatting.com

• 
  Strawberry▲859

  Browser with AI companions that automate web tasks

  AI · 2025 · strawberrybrowser.com

• 
  Peek▲828

  AI personal finance coach that guides you through decisions

  Work · 2025 · peek.money