Remark on statistical model fits to particle ratios in relativistic heavy ion collisions

TL;DR

This paper highlights that using ratios of hadron yields instead of yields themselves in statistical fits for heavy ion collision data introduces biases, affecting the accuracy of chemical freeze-out parameter estimation.

Contribution

It demonstrates that the common practice of fitting ratios rather than yields is statistically flawed, proposing a correction to improve parameter extraction accuracy.

Findings

Fitting ratios instead of yields causes biased parameter estimates

Using yields directly yields more accurate freeze-out parameters

The study clarifies proper statistical methods for analyzing heavy ion collision data

Abstract

In order to determine the chemical freeze-out parameters of the hadron-emitting source in relativistic heavy ion collisions some studies in literature perform fits by using as data input a subsample of ratios calculated out of experimentally measured hadron yields instead of yields themselves. We show that this is a statistically incorrect method fit, implying a bias in the extracted parameters.