This article explains important analytics to know when learning to bet on hockey.
The use of advanced statistics and analytics has grown exponentially over the past decade in hockey. While still in it's infancy stage compared to other sports, many public analysts have created invaluable tools for sports bettors seeking a winning edge. Natural Stat Trick, Evolving Hockey, Hockey Viz and The Athletic are some of the major websites that consistently provide advanced modelling that can be used by bettors.
In this series, we will explore the rising significance of advanced statistics in hockey and how bettors can incorporate analytics into their betting strategies to make more informed and profitable wagers. The first step is understand which statistics are important, what they are and how they can be used in analysis.
Advanced statistics, commonly referred to as "the numbers" or "analytics," go beyond traditional box score stats like goals, assists, shots on goal, plus-minus, and penalty minutes. Instead, they delve deeper into various aspects of the game to provide more accurate and insightful performance measurements.
The purpose of analytics is evaluate the greater picture of either a team or individual. Analytics is an unbiased approach to understanding how a team or player is performing in any given situation. Furthermore, it is a quick snapshot of how a player's performance is impacting team performance. All of the information is helpful when determining how a team will perform in the long-term, in certain matchups and if their current form is sustainable. Those factors are important when considering wagers, be it on individual games or futures bets.
Some of the essential advanced statistics used in hockey include:
Corsi For Percentage (CF%): This metric measures shot attempts (shots on goal, missed shots, and blocked shots) taken by a team compared to their opponents. A higher CF% indicates better puck possession and offensive pressure. CF% has shown to be a predictor of wins, as controlling the shot share is indicative of being the better team. A percentage above 52% is considered to be very good, as anything above 55% over the course of a season puts a team in the top-5. Eight of the top ten teams in CF% last season made the playoffs.
Expected Goals (xG): This metric uses historical data to estimate the probability of a shot resulting in a goal, providing insights into a player's or team's quality of scoring chances. When evaluating team or player performance, this is the go-to statistic. While CF% tracks shot share, expected goals is more indicative of repeatability and therefore, more reliable. Every xG model is different due to mathematical approaches, however they should be within the same range, regardless of source.
A shot from the half wall will count for the same as a shot from the inner slot where CF% and shots are concerned, but the xG value is significantly different. It is more likely a player will score consistently from the inner slot than the half wall. The shot from the inner slot has a much higher xG value and is credited as such. It is important to note that public models do not account for pre-shot movement, which is a key measure of shot quality. For example, a shot off a rebound or cross-ice pace will have a much higher value than a shot from the same spot that didn't involve a rebound or pass.
xGF% is predictive of future performance because it is a truer depiction of performance. If a player or team's CF% is 49%, but their xGF% is 53%, this is an indication that they are getting repeatable, quality scoring chances and holding their opponent to low quality attempts. The Vegas Golden Knights did exactly that this past season and won the Stanley Cup. While a 4% difference does not seem like much, it is a Corsi differential of -297 and an xGF differential of +20. In this case, Vegas would be expected to score 20 more goals than they concede at 5v5 over the course of the season, despite giving up more shot attempts per game.
Expected goals can be displayed as counts or rates, that is as a percentage or as a continuous number. The stat can be used game by game, over multiple games and the course of a season. It is not the be all end all of hockey analytics, but it certainly provides the quickest snapshot into performance.
High-Danger Chances For Percentage (HDCF%): HDCF% evaluates the percentage of high-quality scoring chances a team generates compared to their opponents. The high danger area is also known as the home plate (shown below). Shots from the below the hashmarks in the home plate area are considered "inner slot" shots. It is the most high danger area. However, all shots within the home plate area are considered high-danger, especially if they follow a rebound or royal road/cross-ice pass.
HDCF% is important because it is folded into expected goals. The higher quality the chance, the more likely it is to be a goal, the higher the xG value. Nine of the top-10 teams in HDCF% last season made the playoffs. All four Conference Finalists owned a HDCF% of 52.5% or more. Conversely, five of the bottom-six teams in league standings owned the worst HDCF%, all below 44%. This will be important to note when analyzing how this can help make betting decisions.
PDO (Shooting Percentage + Save Percentage): PDO is used to evaluate a team's luck factor, as it indicates if they are outperforming or underperforming relative to their shot quality and goaltending. Hockey is a random sport when compared to the likes of basketball and American football. There is a lot less left to random chance in those sports, making them easier to predict. You have to be good to be lucky and luck to be good in hockey. Vegas was very lucky in the playoffs, owning a 106.5 PDO, compared to New Jersey who had a 96.2 PDO.
Within a playoff series, this is key. A hot goaltender, Sergei Bobrovsky, can steal a series as he did against Toronto and Carolina. In both series, Florida lost the xG and HDCF% battles, but were buoyed by hot goaltending. Bobrovsky's play raised Florida's PDO to best both teams by nearly 10 percentage points each. That is historical. However, many predicted a fall off was coming once the clock struck midnight on Bobrovsky's play. Bobrovksy's luck ran out in the Final and so did Florida's Stanley Cup hopes.
PDO is important because betting lines don't always account for it. If a team is buoyed by goaltending or shooting percentages, they are unlikely to last the entire year unless their play dictates it, as Boston's did last season. This is a great spot to look for a team to win if their PDO is down, but xG and HDCF% are positive and vice versa.