TLN Roundtable: Should players be up to speed on analytics?

While a number of teams have evidently dipped into their analytics budgets this summer, it’s almost an understatement to say we were floored by what the Leafs did – bringing in Dubas, buying out the popular site ExtraSkater, and eventually creating what looks to be the heaviest analytics department in the league from an outsider’s perspective. Though I don’t consider Dubas an “analytics guy” the way he’s perceived by some, he certainly is the type of manager that will use all the tools at his disposal, and he’s brought in some top-notch resources.

A question that arises from all of this, of course, is how does it all fit together? 

How far should these numbers reach? Is it simply a management tool for finding value where other teams may not, or does the coaching staff need be heavily involved as well? And should this type of stuff ever get to the player level?

Instead of trying to answer all of these questions myself, I asked some of my TLN friends to weigh in.

JEFF (ALSO KNOWN AS JEFFLER or JEFF VEILLEDDE)

Management should obviously be the most aware of these statistics. At the end of the day, their job is to try to acquire the best possible talent for the lowest possible cost, and these numbers typically lead people in the right direction. They should absolutely be used, along with any other variables, to the best of their abilities. As for the coaching staff, I think there’s a value in knowing that your system isn’t working properly. Randy Carlyle shouldn’t walk into the dressing room and declare that his team needs to get all of the Corsis on a given night, but there’s a benefit to being able to quantify when your results are being driven by luck or are truly valuing performance.

The players are where this begins to become a grey area. I don’t think that there’s a value in them knowing their stats, advanced or regular, in a specific form. After all, you don’t play to pad individual stats, you play to win hockey games. Individual statistics are useful in evaluation because they show what you did, but you’re not going to shoot instead of pass on a 2-on-1 because you want to improve your shots per 60. It’s up to the staff to evaluate these numbers, identify strengths and weaknesses, and connect them with video and (gasp) eyeballs to see how the player can refine his game to help them the team. A player doesn’t need to know his CF%, but if the staff finds out that it improves and helps the team if he switches a move up when he gets to the hashmarks, for example, that’s something that should be relayed. Not “your numbers look low, get better numbers”.

JON

What’s great about guys like Kyle Dubas, Tyler Dellow, and Cam is the fact that none of them limit themselves strictly to the numbers. They are all capable of moving beyond reporting stats and drawing only statistical conclusions, they take what they’ve learned from numbers and implement it into a hockey context.

Cam’s writing on this site was a lot more focused on player usage than it was about player recruitment and I think it would add tremendous value to a receptive coaching staff and that’s exactly what we’ll see in Edmonton. Dellow is essentially going to be Eakins guy by most accounts. Having a guy like Dubas influence personnel decisions is great, but analytics need to be a part of the coaching process too. If the Leafs figure out how to properly use the players they have it’s reasonable to consider them a playoff team. 

As for the extent that players need knowledge of statistics, I think it helps from an evaluation standpoint. If a player gets bumped down in the lineup it’s nice to be able to point to something like “we’re seeing too many shots against us when you’re with Player X and Player Y, we need to move things around” or the more positive “the number of shots we see generated when your on the ice should get you some powerplay time.”

BOBBY

I think that any analytics are to be primarily used by management to identify weaknesses and strengths of their roster, and look at other teams and make more educated evaluations before making a trade or signing a free agent. The numbers are also incredibly important because they allow teams to see what the coach can do to improve – whether that is changing a player’s zone starts, changing who enters the zone with possession or dumps, how to clear the D zone, etc. 
Then, it is important that management educates the coaching staff on what the numbers mean and why they matter. There is absolutely no reason that someone can say these numbers are useless once they understand what they mean. No, they aren’t a cure-all, but they can certainly help with roster optimization. Hell, they can even work in conjunction with “old school” hockey attitudes! What craziness! Once coaches understand them, they will be able to use them effectively. It will also help them view the game in a slightly different way, and compare what they think they are seeing with the numbers after. Good coaches adapt, don’t force their players to adapt to them.
I don’t think it is necessary that the players understand and know the numbers. It is the coach’s responsibility to use them to put his players in a situation to succeed. Players can be taught the high-level stuff like “entering the zone with possession is better than dumping it in”, but don’t need to know that they have a 48% corsi with player A and 45% with player B.

JUSTIN

At this point in the evolution of advanced analytics in hockey, it goes without saying that front office and coaching staffs benefit from having access to new and better information. Simply put, advanced stats allow us to better evaluate players in certain contexts, and that will only help teams make better decisions. For general managers, it’s about finding hidden gems in free agency, making stronger trades, and maximizing the money spent in a hard cap world. For coaches, it’s about optimizing line combinations and deployment. We’re past the point where this can still be argued.
Players, on the other hand, are an entirely other case. It’s unrealistic to tell Nazem Kadri that he needs to improve his Corsi Rel QoC numbers, and expect him to execute. High level athletes aren’t stupid, but they have enough to think about when they step onto the ice.
That being said, there’s something to be said about educating players on team-level analytics. Players need to believe in the system their playing, and need to believe that the coaching staff aren’t asking them to play inefficient hockey. Without getting into specific player statistics, why can’t a coaching staff educate their entire team, or even smaller groups of players, how to play more efficiently?
I don’t personally see an issue pulling the third line aside, and explaining to them that “the numbers” tell us that controlled zone entry will result in more and better chances then dumping and chasing. They don’t need to know exactly how everything works or how the numbers came to be, but they’ll probably be happy to know that the strategies and schemes they are asked to execute are in their best interest.
Used as a high-level education tool, I don’t see how exposing players to these statistics can do harm.
  • The coaching staff needs to be able to understand how the numbers relate to their strategy, and be able to give players specific instruction (e.g. Carry in don’t dump) rather than point to numbers. I mean, that would be like saying: ok Phil, we need more goals from you, son. No kidding, coach. It’s the translation factor that is key.

  • CMpuck

    If this proves to be useless in helping the Leafs can’t wait to see the never ending apologetics for advanced stats on this site. I get it though, your boy Cam did great for himself and I wish him luck but let’s be honest, this site will be pro advanced stats and inflates their influence no matter the reality on the ice.

  • silentbob

    I think it is important for players to understand the advanced stats. They aren’t going anyway any time soon and are only going to become a bigger part of the game.

    With Dubas and the new analytic’s department in MLSE, advanced stats are only going to play a bigger role in how the management and coaches evaluate players. It won’t hurt a player to know what Corsi Rel QoC number’s mean when management and coaches are using them to evaluate their play.

    Also if a player does spend a little bit of time to learn about advanced stats and what they mean, a coach can then use Corsi or Fenwick or other stats as a means of communicating and understanding individual and team goals. I agree its unrealistic to tell a player to improve his Corsi, but telling a player to do X and Y because it will improve his and the teams Corsi, that should be no different then telling a player what to do to create more scoring chances.

    Its no different then expecting a MLB player to understand what the OBP stat is so a hitting coach can communicate the reason behind a different approach at the plate.

      • silentbob

        Of course not, just like a player shouldn’t take low % shots on goal or cherry pick instead of back check to increase their goal stat and make themselves more marketable.

        That doesn’t mean its not a good idea for a player to understand the advanced stats so they can understand what the management and their coaches are looking at and talking about. They all need to be able to communicate right? Plus if a player does have a grasp of those stats, they can be used in conversation to help understand concepts, goals and strategies.