Clean Historical Sports Datasets
Budget: $750 – $1,500 USD
Our sports analytics platform is ready for the data-engineering phase, and the immediate goal is to assemble a rock-solid history of NBA, NFL, and College Football games. I need every season pulled, cleaned, and delivered in a format that drops straight into Google Sheets.
Scope
• Source the raw game logs yourself—public archives, league sites, or paid databases are fine as long as the final files reach me as Excel workbooks (my preferred acquisition format).
• Normalize each file: identical column order, names, and data types across all seasons.
• Standardize team names, dates, locations, and scoring conventions so future formulas behave the same way everywhere.
• Return the finished datasets as Sheets-ready CSVs or a single Google Sheet with separate tabs per season.
Acceptance criteria
• Columns match 1-to-1 across every season (no extras, no missing).
• No duplicate rows, misspelled teams, or mixed date formats.
• Spot-checks on random seasons show identical stats to the authoritative source.
• The files open in Google Sheets with no further tweaks.
This phase is data only—no predictive modeling or formula building yet—so your expertise should be in large sports data projects, not ML. If you have past examples of cleaned historical sports datasets, please attach a quick screenshot or shareable link so I can gauge fit.
Fixed-price engagement; once these three sports are complete, we’ll discuss expanding to other leagues.
Scope
• Source the raw game logs yourself—public archives, league sites, or paid databases are fine as long as the final files reach me as Excel workbooks (my preferred acquisition format).
• Normalize each file: identical column order, names, and data types across all seasons.
• Standardize team names, dates, locations, and scoring conventions so future formulas behave the same way everywhere.
• Return the finished datasets as Sheets-ready CSVs or a single Google Sheet with separate tabs per season.
Acceptance criteria
• Columns match 1-to-1 across every season (no extras, no missing).
• No duplicate rows, misspelled teams, or mixed date formats.
• Spot-checks on random seasons show identical stats to the authoritative source.
• The files open in Google Sheets with no further tweaks.
This phase is data only—no predictive modeling or formula building yet—so your expertise should be in large sports data projects, not ML. If you have past examples of cleaned historical sports datasets, please attach a quick screenshot or shareable link so I can gauge fit.
Fixed-price engagement; once these three sports are complete, we’ll discuss expanding to other leagues.
Related categories:
Data Processing
Data Entry
Excel
Web Scraping
Data Extraction
Data Analysis
Data Integration
Data Management