One of my favorite summer exercises is to take a glance at the on-ice save percentages (even-strength) compiled at various hockey databases. Despite organizations getting smarter with time -- the New Jersey Devils are apparently hiring the legendary Sunny Mehta of the Oilogosphere to work their Hockey Analytics department -- there are still plenty of suckers hanging around the water cooler.Though on-ice save percentage tells us plenty about what happened in the past, it tells us virtually nothing about future performance. Consequently, it's more or le s usele s when trying to craft an opinion about where a player's game is headed.Take, for just one example, this plot of all regular skaters Marwin Gonzalez Jersey over the last two seasons. The x-axis indicates their on-ice relative save-percentage -- players with high (positive) numbers saw a favorable number of shots stopped relative to his team's ability to do the same with him off of the ice; players with low (negative) numbers, the opposite. The y-axis is what the player's on-ice relative save percentage looked in the following year.This isn't groundbreaking stuff. There's nothing there. Random chance or variance can describe in hockey, and this is just one of many exhibits in the file.What I do find interesting with these numbers is the discu sions that generally follow around the players at the poles. Over the course of a full season, you'll see team executives, coaches, and the local media blast guys on the really MLB Houston Astros unfavorable side. The miscues are remembered. But what really gets placed in the memory bank are the seemingly endle s number of goals against when the player's on the ice. We know, based on the graph above (and millions of others prior to that), that players can be killed in the save percentage front in one season, and normalize just fine in the next. There's a ton of variables in there, but the negative random chance bug disappearing and natural regre sion can bring those guys right back Yuli Gurriel Jersey around.Since the numbers are readily available, it's easy to go back and pick out the guys at the poles last season. Here would be your ten best and ten worst, and what their numbers looked like the following season:Everyone moved towards zero. The le son, as always: the best defensive players (what on-ice save percentage really doesn't capture) are excellent at limiting shots seen by their goaltender. The bounces will fall. But the best teams mitigate random chance by playing at the other team's end with frequency over a long period of time.On-ice save percentage can't tell us whether a player is particularly strong at driving play or deterring shots -- Corsi% and RelativeCorsi% is where you'll want to go to measure that. But, if you're trying to find guys whose value has been dirtied because of stuff largely out of his control (or looking to preemptively Anthony Gose Jersey move a guy who was getting his goaltender's best this season), it's a good starting point.Here would be this year's group:So, a lot of these guys haven't seen their value dinged in any way, particularly if they scored a ton. Guys like Matt Duchene and Bobby Ryan are all-stars and are going to be treated as much.It's the guys on the left side -- particularly that first trio -- I find amazing. If you had to come up with a list of the three most overly criticized players this year, how many of you would go: (1) Nail Yakupov; (2) Dustin Byfuglien; (3) Evander Kane? It's almost incredible it worked out that way, especially because it adds to the overarching point . Every single player in that group has seen his reputation taken a Kevin Lowe/Brian Burke barn-level beating this year, and to the extent that we're talking about their contributions Carlos Correa Jersey to brutal on-ice save percentages, it doesn't make any sense.Two great hockey players, and one who is very likely going to be great in a couple of years. You'd like to think Winnipeg and Edmonton would resist the force of selling on guys who haven't found the mysterious shot quality against formula. But, if I'm an NHL executive, these are precisely the kind of guys I'm targeting to improve my club next year.