Back to All

For any experiment that already has a winning variation assigned to it, is it still possible to restart it but with a different distribution?

How does setting a winning variation and segmentation work exactly?