Hacker News new | past | comments | ask | show | jobs | submit login

You're supposed to publish results that can be reproduced as you describe.

And it wasn't an outlier.

The simplest comparison I can make to illustrate the point without boring you is measuring the power of a car engine.

Imagine that you get a very high torque number, but horsepower doesn't match. Issue is: torque and horsepower are linearly dependent

torque = HP * RMP / 5252

Measurements that linerarly correlated with the voltage published were all consistent with a number 10/15% lower.




> You're supposed to publish results that can be reproduced as you describe.

You are supposed to try. Whether you succeed is necessarily subject to any measurement issues.

> And it wasn't an outlier.

You said it was.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: