# Statistical asumptions behind confidence intervals

**URL:** <https://dev-forum.pyth.network/t/statistical-asumptions-behind-confidence-intervals/501>\
**Category:** Uncategorized\
**Created:** [January 5, 2026, 12:26pm UTC](https://dev-forum.pyth.network/t/statistical-asumptions-behind-confidence-intervals/501 "2026-01-05T12:26:32Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![artazor](https://yyz2.discourse-cdn.com/flex034/user_avatar/dev-forum.pyth.network/artazor/32/163_2.png) [@artazor](https://dev-forum.pyth.network/u/artazor)\
**Post date:** [January 5, 2026, 12:26pm UTC](https://dev-forum.pyth.network/t/statistical-asumptions-behind-confidence-intervals/501/1 "2026-01-05T12:26:32Z")

</div>

I have a question regarding the statistical assumptions behind Pyth’s confidence intervals for a Hayashi-Yoshida covariance estimator I’m building.

I am modeling “surprisal bits” using the KL divergence between consecutive (μ, σ) updates to construct information bars (accumulating some given threshold). Is it safe to strictly assume the provided confidence interval maps to a Laplace distribution (at 95% confidence) for this purpose?

Would appreciate any feedback from the math/architecture team

---

<div class="post-metadata">

**Author:** ![KemarTiti](https://yyz2.discourse-cdn.com/flex034/user_avatar/dev-forum.pyth.network/kemartiti/32/6_2.png) [@KemarTiti](https://dev-forum.pyth.network/u/KemarTiti)\
**Post date:** [January 5, 2026, 1:22pm UTC](https://dev-forum.pyth.network/t/statistical-asumptions-behind-confidence-intervals/501/2 "2026-01-05T13:22:43Z")

</div>

No, it is not safe to strictly assume Pyth’s confidence interval maps to a Laplace distribution at 95% confidence.

Pyth’s confidence intervals are computed empirically from publisher submissions using a percentile-based approach, not based on any assumed underlying distribution.