TLN Roundtable – The Leafs and SAS Analytics

As you may have heard, the Toronto Maple Leafs are looking to leverage advanced analytics this year and beyond to improve their hockey club. The organization has even gone so far as to team up with analytics software company SAS in order to implement new tracking and evaluation processes. 

This week, our Roundtable tackles the question…

Isn’t it funny how the Leafs went from anti-analytics to being the first team to announce a huge partnership with an analytics company, offering them permanent residence inside the ACC? They really want people to know that this is the new era Leafs. It’s nice to see, but still shocking.

The most important fact about teaming up with SAS is that the Leafs get exclusive access to information that could have been a big part of SportVU, which is in its initial implementation stage in the NHL. That’s huge.

Having exclusivity is amazing and is what MLSE needs to do to give the Leafs a competitive advantage – especially when there is no NHL-mandated spending limit on analytics.

But spending a bunch of money on analytics doesn’t make them useful, and it seems unlikely to me that the Leafs spent all this money to track possession stats that are once again publicly available. What I would guess, and hope, is that the Leafs want to discover more in-depth causes of those stats, and even come up with their own.

And that’s the best use of the SAS partnership. They still have Cam Charron, Darryl Metcalf, and Rob Pettapiece to track possession and whatever else they deem important. 

The SAS partnership is not an instant lotto ticket to a cup, and it is going to take time for Dubas and co. to deem what numbers are important and build enough data to actually see trends. The real value is going to be figuring out what to track – noticing things that others haven’t noticed – i.e. figuring out what causes the volatility of goaltending would be massive, although unlikely.

If SAS is supposed to give the Leafs a competitive advantage down the road, they need to get creative. Possession, zone entries, breakouts, etc are all great and important, but I have a feeling fancy stats is about to get really really granular. 

The analytical team of the Toronto Maple Leafs have their work cut out for them in trying to isolate factors that correlate to strong on-ice performances and ultimately, sustainable ways of  winning hockey games.

While the most widely known advanced stats tend to shift towards puck possession via shot attempts, it’s just step one of the process. Understanding the percentage of shot attempts a team takes on the ice may tell you what happens, but now teams should be looking as to why exactly is one team better than another at achieving these stats.

Tracking specific on-ice events such as zone exits and zone entries are a step in the right direction, but it’s only the beginning of building a better hockey team. Truthfully, stats don’t mean a heck of a lot to a team without proper application in both the short-term and long-term. It’s one thing to understand what the stats say, it’s a whole other monster to find ways to improve them.

Discovering the effectiveness of various plays, such as breakout strategies, puck cycling, power play setups, neutral zone traps and defensive coverage systems and identifying quantifiable strengths and flaws in the various aspects of the game are key to building a strong analytics department. 

Looking at these ideas at both the individual and team from a microanalysis level of thinking can identify trends, which can help the analytical team see the success rates for different strategies. 

This doesn’t have to remain exclusive to analyzing just the Leafs, however. Emulation of other team’s strategies and comparisons across franchises can help separate why a team like Los Angeles is more effective at controlling the game compared to a team like Toronto. 

But perhaps the most crucial aspect of applying these analytics is effective communication and cohesion throughout the entire organization, including management, coaches and players. Once conclusions are drawn, it’s key for people such as assistant General Manager Kyle Dubas to find a way to connect their message in a way that’s useful for the coaching staff to implement.

If head coach Randy Carlyle and the coaching staff can recognize a moment in-game that they’ve been shown typically results in a low or high success rate, this is a perfect teachable moment to pass on to the players on the bench or post-game.

If Phil Kessel’s more effective in generating a shot on goal off of a 2-on-1 by passing the puck as to shooting it, he should be made aware of this idea. If he’s noticeably more successful at getting around defenders on the right side compared to the left, he should know his strengths. Every little tip helps.

With the Leafs being the largest, most valuable franchise in hockey, the brightest future should be one that sees every factor considered, and nothing too bold or too controversial to be explored as a potential influence on the outcome of hockey games.

There’s so much to look at that it’s hard to pinpoint one single thing that the Leafs should be focusing on. The biggest thing that can be taken from data analysis is optimization. It’s about getting the most out of the payroll and getting the most out of the lineup. We already know tracking puck possession via shot attempts is what correlates best with winning, so the best thing to analyze would be the micro-events that precede shot attempts to understand what causes them. Zone entries and exits are the obvious big ones, but the company needs to go deeper than that. 

They need to look at who’s touching the puck, who’s making passes, where are they during these events, basically anything little thing that can be tracked should be tracked across the entire league for as many seasons as possible so they understand what’s skill and what’s dumb luck. Tracking micro-data will lead to a better understanding of player talent and a better understanding of what leads to wins. It’s looking at every detail in an intricate painting.

With the micro-data in hand and an understanding of the “little things” that make up the game, coach’s will be much better suited to optimize their lineups to its fullest potential, and management will be better able to evaluate and project talent. The two go hand in hand. With video tracking coming to the league very soon, the best thing the Leafs can do is track what’s already happened in the past to sort out what’s important and what isn’t. It’ll give them a huge head start over clubs who haven’t done the same before the big data boom that’ll come with the aforementioned video tracking. With SAS Analytics on board, the Leafs need to jump ahead of the curve as much as possible to find whatever edge they can.