Few months ago we decided to focus more on pair programming. We were doing it from time to time before, but after few conversations with far more experienced people, I spoke with the team and we agreed that it's worth to try doing more pair programming.
If there were so many people, who found pair programming as a one of the greatest technique they have ever experienced, it should be worth to spend some more time to improve it in our Scrum.
What bad could happen either way?
After those few months, I can say that was really good idea to invest more time and improve this technique in our team. Right now almost half of our user stories are implemented in pairs and everyone loves it. Why?