Since the 2007-08 season, when it became easy to scrape the data accurately, Corsi has been the figurehead of the analytics movement in hockey. A simple measure of how many shots each team attempts, Corsi accomplishes a few things that conventional performance measures of the time didnt: Its controlled for situation, which reduces noise in the data, a big problem with plus/minus; and it contains a much larger sample size. The combination of a bigger sample size and controlling for situation Eric Weddle Jersey , looking only at what happens when both teams have five skaters on the ice, gave Corsi and its sister statistic Fenwick a huge advantage in accurately measuring team and player performance. MORE: Even in that raw state, Corsi predicted future goals better than goals themselves did. When it was discovered that teams played differently based on the score, it led to the creation of score-adjusted Corsi. Adjusting for score allowedCorsi even more predictive power than it had before, giving good teams credit for spending more time defending leads, and the opposite for bad teams. However, part of what made Corsi such a great team statistic is that for a long time, there were big gaps between teams. The NHLs long stated goal under Gary Bettman has been parity, and it has accomplished that in false ways with things like the shootout and consistent disappearance of power plays, but teams have become bunched closer together as Darrell Henderson Jersey well. One thing I find interesting islooking at how many teams in the NHL are with plus or minus one and two percentage points of 50in score-adjusted Corsi each season. When Corsi first came to prominence as a Jamon Brown Jersey statistic, le s than half the league was within two percent, though more and more teams fell into that middle of the pack as the seasons went on until 2011-12 to 2014-15, where on average under half the league was in that range. Since then, every season the spread has become smaller. Until so far this season, as a whopping 19 of 31 teams are within two percent of even in score-adjusted Corsi. MIXED BAG: When you cut it down to within one percent of even, things get a bit more random, but last season saw nearly 50 percentof the league 14 of 30 teams within one percent of 50 in score-adjusted Corsi. At that point, using Corsi to measure the strength of a team becomes difficult. Similarly, the gap between the top teams in the league and those in the middle is shrinking as well. Over the last four seasons, the gap between the best teams in the league and those in the middle has been barely 4.5 percenton average,about half of what Brandin Cooks Jersey it was a decade ago. None of this is to say that Corsi is usele s now. The math behind it remains solid, but as teams have hired analytics specialists to help make them stronger at controlling the puck, the margins have shrunk. You can still get a good idea forhow well a team is Rodger Saffold Jersey playing by their score-adjusted Corsi, and I dont believe its value for individual players has changed at all. But teams are simply closer together now than they ever have been. Corsi was never the only statistic analysts used to divine team strength, but it was one of the easiest to explain and understand and one of the most important. As analysis evolves and we get finer details about the game ma saged into digestible metrics, its value and prominence will likely continue to decrease, though its not dead yet, even if the NHLs parity is working hard to make it so.