I think systems have a big impact on team possession metrics, which definitely impacts defensive ability, but sometimes this gets overstated as if possession and defense are completely interchangeable.
The way I look at it, the system will dictate how a teams plans to operate with possession, and that will have a big impact on the possession numbers, but great talent will execute at a better rate, which will push the possession number a bit higher as well as cut down on things like how many attempts against are high danger.
If we take a team like Tampa, who if you look at pure possession numbers, they look like a high offense run and gun team that allowed a ton of shots. However, Vashilevsky had the highest expected save percentage of any goalie in the league, which means Tampa allowed the lowest quality of shots of any team in the NHL. Tampa essentially built a style where teams with less offensive talent were allowed to trade chances with them, then used their more talented players to block royal road passes, and keep chances further out. So they essentially banked on their forwards creating higher danger chances, and their D on doing a better job of limiting the other teams high danger chances and created as system where they could run amok on the opposing team offensively. Given the composition of their team, I think this is a great strategy, even if it doesn't lead to dominate possession stats, but this strategy does require high end talent to execute.
Meanwhile, a team like Edmonton tried to dominate games by just limiting shots at all costs, but the lack of talent produced a situation where Talbot actually out produced his expected save percentage but the team still got lit up a lot. This is because players were playing a system that limited shots, but other teams were able to exploit them to create a lot of high percentage chances, while also knowing they could safely take more risks whenever McDavid wasn't on the ice.
Edmonton probably wasn't quite as bad as their record showed last year, but I know a lot of people that expect them to turn it around quickly because they think their possession numbers made it obvious they just got really unlucky.
I think a coach's system can have a massive impact on team defense, and I do think Carlyle might be one of the worst coaches in the league, as his systems tend to both produce lower expected save percentages and more shots against, and because they aren't trading them for more chances at the other end, his teams won't out score their problems, but I do think talent matters a lot in terms of the execution of that strategy.
Carolina is a very nice early example of this. They have been a great defensive and possession team for years, but this year, the combination of adding a 2nd threat from the back end, along with Aho taking a step, Necas coming in, and Svenchnikov, as well as at least average goaltending, is producing much better early results. The injection of higher end offensive talent is what is driving their early success, as they are playing a relatively similar style.
