My recommendation is to change your approach altogether.
Instead, count throughput. How many stories does your team get done in a week? Use that metric to forecast.
Your team will save tons of time and a lot of psychological tax. There is a lot of pressure to get the estimate ‘right’ and that has a cost.
Then, in order to feel like the variation is under control, try to break down the work into smaller chunks. Try to get stories down to a gut check of 1-3 days. But don’t fret whether they are accurate or not. Eventually they will be so used to making stories small, you won’t have to discuss how to make them into 1-3 day stories.
This isn’t innovation. It’s removing waste.