ELO Ratings Questions

ELO Ratings Questions

Author: Noah Gift September 18, 2025 Duration: 3:39

Key Argument

  • Thesis: Using ELO for AI agent evaluation = measuring noise
  • Problem: Wrong evaluators, wrong metrics, wrong assumptions
  • Solution: Quantitative assessment frameworks

The Comparison (00:00-02:00)

Chess ELO

  • FIDE arbiters: 120hr training
  • Binary outcome: win/loss
  • Test-retest: r=0.95
  • Cohen's κ=0.92

AI Agent ELO

  • Random users: Google engineer? CS student? 10-year-old?
  • Undefined dimensions: accuracy? style? speed?
  • Test-retest: r=0.31 (coin flip)
  • Cohen's κ=0.42

Cognitive Bias Cascade (02:00-03:30)

  • Anchoring: 34% rating variance in first 3 seconds
  • Confirmation: 78% selective attention to preferred features
  • Dunning-Kruger: d=1.24 effect size
  • Result: Circular preferences (A>B>C>A)

The Quantitative Alternative (03:30-05:00)

Objective Metrics

  • McCabe complexity ≤20
  • Test coverage ≥80%
  • Big O notation comparison
  • Self-admitted technical debt
  • Reliability: r=0.91 vs r=0.42
  • Effect size: d=2.18

Dream Scenario vs Reality (05:00-06:00)

Dream

  • World's best engineers
  • Annotated metrics
  • Standardized criteria

Reality

  • Random internet users
  • No expertise verification
  • Subjective preferences

Key Statistics

MetricChessAI Agents
Inter-rater reliabilityκ=0.92κ=0.42
Test-retestr=0.95r=0.31
Temporal drift±10 pts±150 pts
Hurst exponent0.890.31

Takeaways

  1. Stop: Using preference votes as quality metrics
  2. Start: Automated complexity analysis
  3. ROI: 4.7 months to break even

Citations Mentioned

  • Kapoor et al. (2025): "AI agents that matter" - κ=0.42 finding
  • Santos et al. (2022): Technical Debt Grading validation
  • Regan & Haworth (2011): Chess arbiter reliability κ=0.92
  • Chapman & Johnson (2002): 34% anchoring effect

Quotable Moments

"You can't rate chess with basketball fans"

"0.31 reliability? That's a coin flip with extra steps"

"Every preference vote is a data crime"

"The psychometrics are screaming"


Resources

  • Technical Debt Grading (TDG) Framework
  • PMAT (Pragmatic AI Labs MCP Agent Toolkit)
  • McCabe Complexity Calculator
  • Cohen's Kappa Calculator

🔥 Hot Course Offers:

🚀 Level Up Your Career:

Learn end-to-end ML engineering from industry veterans at PAIML.COM


Noah Gift guides you through a year-long journey with 52 Weeks of Cloud, a weekly exploration designed for anyone building, managing, or simply curious about modern cloud infrastructure. Each episode digs into a specific technical topic, moving beyond surface-level explanations to offer practical insights you can apply. You’ll hear detailed discussions on the platforms that power the industry-like AWS, Azure, and Google Cloud-and how to navigate multi-cloud strategies effectively. The conversation regularly delves into the orchestration of these systems with Kubernetes and the specialized world of machine learning operations, or MLOps, including the integration and implications of large language models. This isn't just theory; it's a focused look at the tools and methodologies shaping how software is deployed and scaled today. By committing to this podcast, you're essentially getting a structured, expert-led curriculum that breaks down complex subjects into manageable weekly segments, all aimed at building a comprehensive and practical understanding of the cloud ecosystem.
Author: Language: English Episodes: 225

52 Weeks of Cloud
Podcast Episodes
Academic Style Lecture on Concepts Surrounding RAG in Generative AI [not-audio_url] [/not-audio_url]

Duration: 45:17
Episode Notes: Search, Not Superintelligence: RAG's Role in Grounding Generative AISummaryI demystify RAG technology and challenge the AI hype cycle. I argue current AI is merely advanced search, not true intelligence, a…
Pragmatic AI Labs Interactive Labs Next Generation [not-audio_url] [/not-audio_url]

Duration: 2:57
Pragmatica Labs Podcast: Interactive Labs UpdateEpisode NotesAnnouncement: Updated Interactive LabsNew version of interactive labs now available on the Pragmatica Labs platformFocus on improved Rust teaching capabilities…
Meta and OpenAI LibGen Book Piracy Controversy [not-audio_url] [/not-audio_url]

Duration: 9:51
Meta and OpenAI Book Piracy Controversy: Podcast SummaryThe Unauthorized Data AcquisitionMeta (Facebook's parent company) and OpenAI downloaded millions of pirated books from Library Genesis (LibGen) to train artificial…
Rust Projects with Multiple Entry Points Like CLI and Web [not-audio_url] [/not-audio_url]

Duration: 5:32
Rust Multiple Entry Points: Architectural PatternsKey PointsCore Concept: Multiple entry points in Rust enable single codebase deployment across CLI, microservices, WebAssembly and GUI contextsImplementation Path: Initia…
Python Is Vibe Coding 1.0 [not-audio_url] [/not-audio_url]

Duration: 13:59
Podcast Notes: Vibe Coding & The Maintenance Problem in Software EngineeringEpisode SummaryIn this episode, I explore the concept of "vibe coding" - using large language models for rapid software development - and compar…
DeepSeek R2 An Atom Bomb For USA BigTech [not-audio_url] [/not-audio_url]

Duration: 12:16
Podcast Notes: DeepSeek R2 - The Tech Stock "Atom Bomb"OverviewDeepSeek R2 could heavily impact tech stocks when released (April or May 2025)Could threaten OpenAI, Anthropic, and major tech companiesUS tech market alread…
Why OpenAI and Anthropic Are So Scared and Calling for Regulation [not-audio_url] [/not-audio_url]

Duration: 12:26
Regulatory Capture in Artificial Intelligence Markets: Oligopolistic Preservation StrategiesThesis StatementAnalysis of emergent regulatory capture mechanisms employed by dominant AI firms (OpenAI, Anthropic) to establis…
Rust Paradox - Programming is Automated, but Rust is Too Hard? [not-audio_url] [/not-audio_url]

Duration: 12:39
The Rust Paradox: Systems Programming in the Epoch of Generative AII. Paradoxical Thesis ExaminationContradictory Technological NarrativesEpistemological inconsistency: programming simultaneously characterized as "automa…
Genai companies will be automated by Open Source before developers [not-audio_url] [/not-audio_url]

Duration: 19:11
Podcast Notes: Debunking Claims About AI's Future in CodingEpisode OverviewAnalysis of Anthropic CEO Dario Amodei's claim: "We're 3-6 months from AI writing 90% of code, and 12 months from AI writing essentially all code…
Debunking Fraudulant Claim Reading Same as Training LLMs [not-audio_url] [/not-audio_url]

Duration: 11:43
Pattern Matching vs. Content Comprehension: The Mathematical Case Against "Reading = Training"Mathematical Foundations of the DistinctionDimensional processing divergenceHuman reading: Sequential, unidirectional informat…