---
title: "Discussion on \"The draw is 24% of the outcomes and 0.8% of the labels: auditing class collapse in a three-way model\" | Hashnode"
description: "Discussion on \"The draw is 24% of the outcomes and 0.8% of the labels: auditing class collapse in a three-way model\". A three-class classifier that reports one label per row is not a three-class classifier. It is a ranking function collapsed to a decision, and the collapse is where the information dies. On our platfo"
url: "https://hashnode.com/posts/the-draw-is-24-of-the-outcomes-and-0-8-of-the-labels-auditing-class-collapse-in-a-three-way-model/6af3f7e04f16024dc27ebee5"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [The draw is 24% of the outcomes and 0.8% of the labels: auditing class collapse in a three-way model](https://soccerdreamai.hashnode.dev/the-draw-is-24-of-the-outcomes-and-0-8-of-the-labels-auditing-class-collapse-in-a-three-way-model?utm_source=hashnode&utm_medium=feed)

A three-class classifier that reports one label per row is not a three-class classifier. It is a ranking function collapsed to a decision, and the collapse is where the information dies. On our platfo

3d ago · 10 min read · soccerdreamai.hashnode.dev  
Read the article

## Written by

[SoccerDream AI](https://hashnode.com/@soccerdreamai)

@soccerdreamai · Beijing, China

  
[View profile](https://hashnode.com/@soccerdreamai)

Independent prediction models publishing timestamped probability estimates on football fixtures. Graded in the open, retired models included

Independent prediction models publishing timestamped probability estimates on football fixtures. Writing about evaluation integrity, model life cycles, and calibration.

## Responses

Markdown supported · ⌘↵ to post

No responses yet.