Hackathons: Kaggle (page 10)
Hackathons · Kaggle: 157 entries in total; this page lists entries 136–150.
Keep 'em Separated
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Spatial analysis of receiver and defender separation.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Down Set Hut
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Pre-snap tracking analysis and modeling workflow.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
It's a Bluff: Predicting Receiver Decoy Motions
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Prediction of receiver decoy motion from pre-snap tracking features.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Trench Chess
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Modeling tactical interactions around the line of scrimmage.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Exploring Sports Sponsorship – Ads Exposure Analysis
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Long-context video analysis workflow for measuring sponsor exposure.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
KeepTrack: Use Gemini to Keep Track of Anything
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Long-context multimodal workflow for extracting and tracking events from video.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Building Process Documentation from Video with AI
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Converts long-form process video into structured operational documentation.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Build AI Agents With Google's LLM Gemma
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Agent construction workflow using Gemma for data tasks.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
PyGEM: A Chatbot for Python Questions using Gemma
Evidence grade: for discovery only; must not by itself support a formal conclusion.
A Gemma-based assistant specialized for Python questions.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Gemma meets LangChain - summarize kaggle writeups
Evidence grade: for discovery only; must not by itself support a formal conclusion.
A LangChain and Gemma pipeline for summarizing Kaggle write-ups.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Unleashing Gemma's Power by Prompt Engineering
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Prompt-engineering workflow and evaluation for Gemma.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
KaggleBot - Gemma-7b-it RAG w few-shot prompting
Evidence grade: for discovery only; must not by itself support a formal conclusion.
A retrieval-augmented Kaggle assistant combining Gemma with few-shot prompting.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Pull the Plug
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Player-tracking analysis of tackling decisions and outcomes.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Uncovering Missed Tackle Opportunities
Evidence grade: for discovery only; must not by itself support a formal conclusion.
Modeling missed-tackle opportunities from tracking data.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24
Momentum-based Fractional Tackles
Evidence grade: for discovery only; must not by itself support a formal conclusion.
A momentum-aware method for assigning fractional tackle credit.
Evidence use: For discovering how winning projects were implemented, their data resources, evaluation methods and application constraints; specific conclusions must be checked against the original project.
Limitations: Hackathon awards are not equivalent to peer review or independent reproduction. Project descriptions may change with the Kaggle page; check code, data and dependency versions before use.
Link last checked: 2026-09-24