being on leave from work and having claude code max >>>>
Claude and I build Lineup Labs https://t.co/f848UulUmm Pick any 5 NBA players (from any era) and it will predict their success as a lineup. Choose an opposing 5. Search for the best lineups. Updated for the 26-27 season.
MCP (Model Context Protocol) standardizes how AI models discover and call tools at runtime. A normal API is for code: developers read docs and hardcode fixed endpoint calls. MCP servers list tools with plain-language descriptions and schemas; the model decides which to use mid-task. Most MCP servers simply wrap existing APIs. This solves the integration mess for agents across clients without custom code per app. Official docs: https://t.co/ZZLBkSWxSJ
@TourSwingsTommy Honestly none of these swings were ever the same. He might have the most “good” variability out of anyone - the type that Adam Young talks about
@ibrasquared@JerryEngelmann@JerryEngelmann can you plot it so it controls for the fact that possessions are zero sum. Two people on the same team can’t score in the same possession (I guess that can happen really rarely). Like someone else mentioned plot the residuals of a regression