# Is TB200 neutralizing the Moore-Penrose Inversion neutralization step in Signals

**URL:** <https://forum.numer.ai/t/is-tb200-neutralizing-the-moore-penrose-inversion-neutralization-step-in-signals/4308>\
**Category:** Signals\
**Created:** [October 10, 2021, 3:32pm UTC](https://forum.numer.ai/t/is-tb200-neutralizing-the-moore-penrose-inversion-neutralization-step-in-signals/4308 "2021-10-10T15:32:49Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![of\_s](https://yyz1.discourse-cdn.com/flex009/user_avatar/forum.numer.ai/of_s/32/2627_2.png) [@of\_s](https://forum.numer.ai/u/of_s)\
**Post date:** [October 10, 2021, 3:32pm UTC](https://forum.numer.ai/t/is-tb200-neutralizing-the-moore-penrose-inversion-neutralization-step-in-signals/4308/1 "2021-10-10T15:32:49Z")

</div>

Using the Signals scoring example provided [here](https://github.com/numerai/example-scripts/blob/ac461829258858b6f342fd2ad55e076016912876/SignalsScoringExample.ipynb), what effect does TB200 have on the second neutralization step?

If the TB200 effect is to reduce the amount by which our predictions are linearly neutralized to their set of factors, then why not just move to a rescaled un-neutralized scoring system?

Providing the un-neutralized scores will also help identify any bias this second neutralization step has currently.

 ![image](https://canada1.discourse-cdn.com/flex009/uploads/numerai/original/2X/f/f5e2379f24c13c6d8d84bbe54b8c525cfd3f17fd.png)