Power Ranking Methodology
Power Rankings are a common sight on hockey websites these days. They aim to provide fans with a concise view of overall team strength relative to one another as the season progresses. Some rely solely on statistical models involving individual player performance, others only use actual team results in the win-loss column, and many are pure opinion pieces with clear bias sprinkled throughout. There are also those that use combinations of these methods, aiming for a balance between the logical and factual. However, no single Power Ranking has been able to provide a legitimate, foolproof guide to overall team strength, that is, until now.
Puck Propaganda aims to succeed where others have failed. Our goal is to provide hockey fans with the definitive NHL Power Rankings, one that you can trust implicitly. In order to assemble this infallible guide, we’ve outlined the following 3 requirements…
- Divergence. The impact of the expected success must equal the impact of the actual success. This means that 50% of a team or player score is based on the expected season performance and 50% is based on the actual season to date performance.
- Organization. The impact of the overall team must equal the impact of the individual team members combined. This means that 50% of this score is based on the team and 50% is based on the players.
- History. The impact of the recent success must equal the impact of the past success. This means that 50% of a team or player score is based on recent success (the length of which can vary) and 50% is based on past success (the length of which can also vary).
These requirements are then met by using a wide assortment of data and analysis before being combined to produce the overall score for team strength.
D + O + H = DOH!
First, we’ll look at D; expected vs. actual. Expected success for a team or player is defined by Puck Propaganda as the reliable assumptions that can be made regarding the output given the known qualities. This encompasses a wide range of areas, but basically boils down to the following…
- What is expected of a team based on its inherent ability? We’ve identified a set of qualities that an all-around average team should have. To evaluate the expected performance of a team we compare these inherent abilities to those of the all-around average team and then score the team based on the degree to which it’s better or worse. As for the quality of the team’s abilities, those are determined by the players.
- What is expected of a player based on his inherent ability in relation to his teammates? First, we determine the expected performance of a player based on his personal abilities if he was placed on the all-around average team at a position suited to his skill level. Then we offset this expected performance by the quality of the team and/or linemates/defensive partner he spends the majority of his even-strength ice time with.
We’ll use Sidney Crosby as an example of how this plays out into calculating the expected. On an all-around average team, Crosby would be the clear #1 center due to his elite talent level. Therefore we would calculate his expected performance when paired with average 1st line wingers on this team . Then we would offset this by the quality of his actual team (Penguins) and even-strength linemates (Dupuis/Kennedy).
As for actual success, that’s simply defined as the actual, recorded output of the team or player up to the current point in the season. Both the expected and actual are then used in computing the Divergence, or D.
Next we’ll look at O; team vs. players. Puck Propaganda defines the team as the overall collection of talent being used by the organization on game day. This includes the active roster and coaching staff, but does not include injured players, non-roster prospects, or upper levels of management. The team is graded based on team statistics that we have specifically recognized as being important to team play. They are even-strength goal differential, shot differential, takeaway/giveaway differential, power play percentage, penalty kill percentage, hits, and faceoff percentage. A weighted combination of these statistics determines the team’s score.
Puck Propaganda defines the players as those making up the game day roster. These players are then graded individually before being combined to create the players score. Each player is graded based on player statistics that we have specifically recognized as being important to an individual’s performance. For forwards/defensemen they are goals, assists, hits, shots, blocked shots (defensemen only), faceoff percentage (centers only) and corsi. For goaltenders they are goals against average, save percentage, and shutouts. A weighted combination of these statistics determines the individual player’s score. Finally, a weighted combination of these individual scores creates the players overall score. The weight assigned to each player is based on his role with the team. For example, 1st line forwards have more weight than 2nd line forwards, 1st pair defensemen have more weight than 2nd pair defensemen, and starting goaltenders have more weight than backup goaltenders.
Both the team and players scores are then combined to give us the Organization score, or O.
Finally we look at H; recent vs. past. Recent success for the team could be based on an active hot/cold streak or play since a lineup change. Recent success for a player could also be based an active hot/cold streak, play since moving up/down the lineup, or play since returning from an injury. Past success for the team could be based on performance in prior season(s) with a similar lineup or current season success prior to the defined recent success. Past success for a player could be based on any historical performance of the individual under varying circumstances.
In summation, we consider both the long-term (past) and short-term (recent) success in the Power Rankings as both are essential to the makeup of a team’s strength. This gives us the History, or H.
In conclusion, Puck Propaganda‘s NHL Power Rankings are ordered based on the teams’ DOH! scores, from best to worst. This means that for the point in time that the Power Rankings capture they determine that any higher ranked team should defeat any lower ranked team in a best-of-7 series. The only reason Puck Propaganda says should is to account for in-game abnormalities that could affect the outcome, such as if a player leaves mid-game via an injury or game-misconduct. Otherwise these rankings are as good as gold. Trust in them like you would trust yourself.