# Log loss and test data

**URL:** <https://community.drivendata.org/t/log-loss-and-test-data/2923>\
**Category:** Warm Up: Predict Blood Donations\
**Created:** [November 30, 2018, 3:16pm UTC](https://community.drivendata.org/t/log-loss-and-test-data/2923 "2018-11-30T15:16:53Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![SaraLS](https://avatars.discourse-cdn.com/v4/letter/s/a3d4f5/32.png) [@SaraLS](https://community.drivendata.org/u/SaraLS)\
**Post date:** [November 30, 2018, 3:16pm UTC](https://community.drivendata.org/t/log-loss-and-test-data/2923/1 "2018-11-30T15:16:53Z")

</div>

I am curious how the public scores are calculated. I am trying to correlate the score I get with the log loss I am calculating. Sometimes I get a better score with a higher log loss and worse score with a lower log loss. My best score right now is with a model that generated a log loss of #Log\_loss = 7.915164020966691 and my second best score had a log loss of #Log\_loss = 6.987650697117431. These are on the modeled data, of course, not the new test batch, which generates the score. So I am not sure how to relate the two.