UGG Baratas online. Australia UGGs Outlet España.

Your #1 Source for Nashville Predators News, Updates, and Opinions – A Nashville Predators Blog

Advanced Statistics Breakdown: Predators vs Coyotes – October 31st, 2013


Nov 2, 2013


Need to understand any of the terms in the article below? Check out our Introductory Guide to Advanced Hockey Analysis

There’s really no way to take the sting out of this loss. Looking at the numbers, it would appear that the Predators did virtually everything right. They received scoring from depth players, they scored the first goal (6-0-1 when accomplishing that, going into last night), and they hammered the Coyotes with shot attempts all night. The raw statistics for this game are terrific for the Predators, one of the best games of the season. Unfortunately, as we always say–sometimes there’s no accounting for “acts of Hockey God,” be it defensive lapses, poor goaltending, or good old-fashioned dumb luck–all of which this game had in spades. If there’s any redemption, it’s that the Predators did a lot of good things in terms of puck possession, after a bad streak in the previous few games.

In the 36 “close” 5-on-5 minutes, the Predators dominated Corsi by a margin of 61 to 39 percent. The teams combined for 72 Corsi events, 44 of which belonged to the Predators. Nashville accounted for 22 shots, 11 misses, and 11 blocks.  The Coyotes countered with nine shots, eight misses, and 10 blocks for a total of 28 Corsi events. This leaves an even more lopsided Fenwick total of 66/34% in favor of the Predators. On the strength of this possession performance, the Predators climbed a few spots back up to 17th–just below these same Phoenix Coyotes.

A stat that we don’t typically like to look at on a game-to-game basis now bears mention. PDO, which you’ll recall is used to calculate a team’s “good/bad bounce” factor, or to simplify it’s “luck”, is calculated by adding the 5-on-5 close on-ice shooting percentage to the on-ice save percentage in the same period. With 13 games played, we can get a good idea of whether a team is getting a lot of good breaks, and are due to have the bottom fall out–or there’s a run of abnormally bad fortune at play.

For example, the Colorado Avalanche have been the surprise feel-good story of the NHL. They are winning a lot of games against all odds, just a few months removed from winning the draft lottery. When I say “against all odds,” that point bears emphasis, because they’re operating under some percentages that are almost certainly not sustainable. The Avs hold the third best shooting percentage in the league with about 11%. Over the past several years, the league average tends to fall somewhere around 8%, which is to say that the Avalanche are seeing a lot of their shots go in, and that trend isn’t likely to continue. Traditionalists may point to the virtue of shot quality here, but there have been entire articles dedicated to explaining away this notion (Travis Yost of HockeyBuzz sums it up well).

Now, we would expect a let-down based on Colorado’s bloated shooting percentage alone, but their save percentage, which makes up the other half of the PDO equation, is similarly inflated. The Avalanche currently maintain an on-ice save-percentage of 96.3%, which is the second best in the entire league, behind only the Montreal Canadiens. Adding these values together, we get Colorado’s PDO–a league-leading 107.1%. As we discussed in the guide to advanced statistics, historical data has demonstrated that teams will always regress to a mean of around 100 percent. This means that anything over 100 tells us that everything is going right for a team, whereas a sub-100 percentage suggests the opposite. While the Avs’ PDO doesn’t state absolutely that they’re going to crash and burn, it does tell us that they’ve been enjoying a ride with their feet off the pedals, but eventually the slope is going to turn to an incline.

If the Colorado Avalanche are exemplary of getting all the breaks, the Predators are the “tails-up” side of that coin. Their shooting percentage is a paltry 5.2%, the fourth-worst number in the league. Meanwhile, Nashville’s save percentage is in the high middle of the pack at 93%, just above the league average of about 92%. The sum of these percentages is 98.3 percent, which is to say that their luck could and should improve. With a relatively strong save percentage, to still fall short of that 100% mark, the shooting percentage is understood to be quite dreadful. This might be cause for a bit of optimism–statistically, there is no way the Predators can continue to hold such a low shots/goals ratio. I had believed that last night could be a step toward normalization, but then the Predators accompanied their four goals with a higher-than-average amount of shots. We spend a lot of time wringing our hands about the lack of scoring, but a shooting % that poor tells us, “keep at it, the pucks will start going in.” It’s somewhat heartening that the Predators have managed to stay in the race, in spite of a bottom ten PDO.


Kevin Klein: A strong night for Klein, who lead the team in both Corsi and Fenwick percentage. As usual, Klein’s usage was fairly balanced, so this accomplishment isn’t sullied by soft minutes.

Mike Fisher: Fisher’s 53% Corsi-For is great out of context, but extremely impressive when you factor in that he had the second highest concentration of defensive zone starts on the team and the second lowest offensive zone start percentage–just 13.3 percent. In aggregate, you could reason that much of time when Fisher was on the ice, the puck started in the Predators’ end and wound up in the Coyotes’ zone.

To conclude, I wanted to address a reader comment.  Essentially, his reaction to my statement that Legwand had an uncharacteristically poor possession game against Winnipeg, on the whole, was:

Legwand…had an awful night (2 assist) and Hendricks had a good night (2 minute penalty)? That’s another feather in the cap for advanced stats.

I think he raises a good point, one that I need to clarify. From the start, I’ve posited that the value in advanced statistics isn’t really contained in a small sample, nor does it stand on its own. All of this information is designed to complement the more traditional numbers inherent to the game. Goals and assists don’t tell the whole story, nor does Corsi rating or zone deployment. When you combine them with the most important gauge–actually watching and absorbing the game itself, you can get a good idea of where the team stands and what direction it is likely heading.

Legwand’s numbers are typically around the “even” Corsi mark, but he had an unusually difficult game in terms of being on the ice for far more shot attempts against than for. That’s not intended to invalidate Legwand’s assists, en route to the victory, nor do those assists rub out any deficiencies he might have shown in the all of the other collective minutes he played. The example I used was,

If, say, Colin Wilson loses puck battles all night, is constantly chasing opposing players without the puck, and is generally on the ice for far more shots against than shots for…but in one 30 second span, gets a pass from Weber and scores the game winning goal, can you say that overall, he played a strong game? It’s not about discrediting players in a small sample, it’s about establishing predictability and overall likelihood of long-term success–and I did state that typically Legwand falls in the “even” range. I was just pointing out that, on the whole, he was weaker than normal in driving play.

To summarize–this blog is never intended to condemn or take anything away from a player’s accomplishments in the course of a game, especially when those actions result in a victory. Rather, the idea is to paint the big picture, and use that image to establish a trend to determine a player’s overall contribution to his team’s chances of future success.

Statistics Courtesy of The Extra Skater