The Adjusted Hockey project has primarily been about adding context, a critical ingredient most hockey statistics and conversations lack. Removing the noise of reputation, fame, and timing – both good and bad – we want to focus solely on the player’s on-ice contributions and achievements. The end game has always been to create a single, all-encompassing figure for each player, a platform to fairly compare generations.
How will the unique case of the revolutionary Bruins' star Bobby Orr grade out given his short NHL career? “Bobby Orr 001” by rchdj10 is licensed under CC BY-ND 2.0
The system’s output will serve three critical purposes:
- Objectively rank the careers of every player in NHL history;
- Evaluate which players are historically underrated or overrated;
- Create a Hockey Hall of Fame induction standard by era and position to guide future selections.
The introductions to context, adjusted statistics, eras, value, and adjusted point shares were building blocks to arrive at this point. The creation of the Pidutti Point Share system will allow us to meet the above three main objectives, but also opens up a lot of doors for future analysis and articles. With a career evaluation model in place, we can isolate individuals, positions and eras for comparison, dive into the best of the NHL’s all-time best, explore Hall of Fame selections, create revised top player lists and all-time team selections, and, most importantly, have a comprehensive go-to methodology for any player we want to highlight. Such a canvas provides limitless opportunities for both fun and interesting hockey analysis.
The Inspiration
The initial inspiration for the PPS system comes from three places:
- Bill James’ The New Bill James Historical Baseball Abstract (2001), specifically his player ratings. Using aspects of his revolutionary Win Shares system, James rated the top baseball players on record. It shouted at me that hockey needed something similar to launch historical player analysis.
- Jay Jaffe’s JAWS methodology (Jaffe Wins Above Replacement Score). The system is part of the fabric of Hall of Fame evaluation in baseball today, which culminated in his book, The Cooperstown Casebook (2017). It sparked me to attempt to create one in hockey where thoughtful public Hall of Fame conversations are glaringly absent.
- Lastly, not from a book, but a website – Hockey Reference’s adjusted statistics and point shares. The concepts demonstrated that others have tried to contextualize hockey history through statistics on a credible platform. In these creations, I saw building blocks for a methodology to compare every NHL player in history.
The initial inspiration for my player evaluation system comes from three places. The first is Bill James’ aforementioned The New Bill James Historical Baseball Abstract, specifically his player ratings. Using aspects of his revolutionary Win Shares system, James expertly rates the top all-time MLB players on record. When I purchased the book in 2011 – a decade after it hit shelves – it was an eye-opening experience. The sensibility, depth, and results of his player rating system shouted at me that hockey needed something similar to launch objective historical player analysis.
The second inspiration is Jay Jaffe’s brilliantly named JAWS methodology (Jaffe Wins Above Replacement Score). While Jaffe’s approach had been developed in 2004, it slowly and rightfully became mainstream and part of the fabric of Hall of Fame evaluation in baseball today. This culminated in 2017’s The Cooperstown Casebook: Who’s in the Baseball Hall of Fame, Who Should Be In, and Who Should Pack Their Plaques, a book I devoured within weeks of its publication. It sparked me to attempt to create a system to hockey where proper public Hall of Fame conversations are glaringly absent.
Lastly, the third inspiration comes not from a book, but rather from a website, in the form of Hockey Reference’s adjusted statistic and point share methodologies. The concepts demonstrated that others have tried to contextualize hockey history through statistics, to the point of inclusion on a credible platform popular in hockey circles. In Hockey Reference’s creations, I saw the potential to bring new, objective thinking about statistical history to hockey’s mainstream. It made me believe there were pieces in place to attempt my lifelong pie in the sky – create a methodology to objectively compare every NHL player in history.
While it’s nice to be inspired, nothing happened to move that inspiration forward. I was a full-time, often-exhausted accountant, not a sports writer. As far back as elementary school, I had made attempts to compare baseball players on paper, often the 1990s Toronto Blue Jays. In 2013, I had looked at how Sidney Crosby and Alex Ovechkin stacked up to past legends using adjusted statistics as my first real foray into the topic. But the effort was no more than a crude four-page document that went unfinished and unseen. Formerly working 60 hours a week, frequently travelling Northern Ontario, I lacked the requisite time and energy needed to tackle such a project. It remained a pipe dream until a change in career aspirations and a global pandemic simultaneously brought more free time than I had ever known. Starting by blindly downloading player data for each of the NHL’s first 102 seasons during the first of many COVID lockdowns, a spark was lit, and a project was born.
Initially, I thought it best to replicate JAWS in the NHL. The beauty of Jaffe’s system is how it manages to be both expressive and uncluttered. Take a player’s career value (Wins Above Replacement, or “WAR,” in baseball) and the value of his seven best seasons and average them. That’s the score. This approach inherently applies double the credit to those top seven, non-consecutive seasons (i.e., peak years). JAWS is a career/peak model, you might say. By taking the mean of each Hall of Famer’s score by position, it creates a positional standard. To be a worthy candidate to uphold the standard, an eligible shortstop outside the Hall of Fame, for example, should exceed (or at least approach) the average score of all Hall of Fame shortstops. Each enshrined Hall of Famer can also be assessed on where their score fits against the JAWS positional standard to identify the no-doubters and the suspect selections.
But issues instantly arose when applying a similar approach on the NHL, substituting baseball’s WAR with my adjusted version of point shares. Many great hockey players often have compromised careers – a workplace hazard of a violent, physical sport. Orr, Bossy, Lemieux, Lindros, Bure, Forsberg, just to name a few. Players with longer careers were disproportionately higher on the leaderboards, even those that hung around chasing Stanley Cups or milestones as depth players (eg. Dave Andreychuk or Patrick Marleau). Even within individual seasons, many players had incomplete years at their peak due to injury. So, while JAWS is a superb fit for baseball, any iteration of its ideology did not produce suitable results when unleashed on the NHL.
From there, it was clear the approach needed a way to boost players who had shorter careers. Orr is widely considered the best defenceman of all-time, despite playing only 36 games after his age-26 season. Lemieux is regarded among the top handful of forwards in history, yet he played 65 games or more just three times after his age-23 season. To evaluate players appropriately, total output needed to be a lesser part of the puzzle. After dozens of revisions spanning months, incorporating iterations of values for career and peak, a proper combination failed to emerge.
To clarify, the goal of objective analysis is not to replicate popular opinion. After all, what would be the point? But having spent nearly a year reviewing the personal database I constructed and examining the Hockey Reference pages of hundreds of players many times over, I had gained a wealth of knowledge contextualizing careers. Basic stats, adjusted stats, point shares, paces, playoff performance, award votes. Over and over. It was clear the system’s recipe was missing a key ingredient.
A major breakthrough appeared when factoring in a third major element to the system – pace. When creating adjusted pace statistics for seasons and careers, the concepts proved extremely insightful. So, when I added a form of pace value (i.e., value per game played), it launched the quality of the ratings forward. Instantly, players with solid, long careers slide down the ratings, and players with elite, shorter careers slid up the ratings. From the research and experimentation throughout the Adjusted Hockey project, it became clear these three measures should form the bedrock to most appropriately encapsulate an NHL career. The system was now rewarding truly great players appropriately, whether or not they were still a middle-six forward at age 36. From exhaustively tweaking the combinations, a formula founded on a split of career, peak, and pace proved a well-timed hat trick at a key point in the game. The system finally had an engine.
The careers of Martin Brodeur and Scott Niedermayer, teammates with Team Canada and New Jersey, are intrinsically linked to success in both the NHL playoffs and international play. “Canada vs. USA” by s.yume is licensed under CC BY 2.0
Playoffs and International Play
Having grown up reading a lot of baseball analysis and with JAWS front of mind, my initial system excluded the NHL post-season entirely. I viewed playoff performances as overstated given the small sample of playoff games. But MLB plays 162 games with a historically limited playoff pool; it was not until 1994 that more than four teams printed playoff tickets. The NHL’s Stanley Cup playoffs have always been much more inclusive, the post-season comprising a larger proportion of the schedule. By 1971, three rounds of playoffs would be in place, arriving at the familiar, four-round, 16-team format in 1980. The NHL’s playoffs simply comprise too many hockey games on a meaningful stage to dismiss from a player’s career.
To illustrate the regular season vs. playoffs split in the NHL, let’s look at the post-season as part of a career body of work. Consider the following three Hall of Famers from Era 4, ranging from minimal to modest to exceptional post-season success.
REGULAR SEASON VS. PLAYOFF GAMES PLAYED, SELECT PLAYERS
| Player | Regular Seasons | Playoff Seasons | Cup Wins/ Cup Finals | Season GP | Playoff GP | Season GP% | Playoff GP% |
| Jarome Iginla | 20 | 8 | 0/1 | 1,554 | 81 | 95% | 5% |
| Luc Robitaille | 19 | 15 | 1/2 | 1,431 | 159 | 90% | 10% |
| Marian Hossa | 19 | 17 | 3/5 | 1,309 | 205 | 86% | 14% |
| Average | 19.3 | 13.3 | 1.3/2.7 | 1,431 | 148 | 91% | 9% |
GP = Games Played; Season GP% and Playoff GP% = % of total career GP in each format
At a typical range of 5% to 15% of an Era 3 or 4 player’s career comprising post-season games, the playoffs need to be part of the equation. This conclusion created a number of complex questions around how to factor in playoff success appropriately; however, all iterations appreciably improved the quality of the rating system given the sport’s storied fascination with the pursuit of the Stanley Cup.
When considering the post-season, hockey’s rich international history emerged as another outlet to potentially factor into the system. Initially, I discarded any inclusion of these streams of play, given the multitude of formats over the years. However, unlike most North American sports, hockey has held regular “best-on-best” tournaments for nearly five decades. Given almost every elite player (in Era 3 or 4) has or will have multiple opportunities to play for their country on a big stage, it was clear a small sliver of the player ratings should capture international play. The extent of this piece within the system, which events warrant inclusion, and how to award points objectively became new questions to answer. But the system continued to round into form, the outputs and order increasingly refined and more effective.
Other Factors
The decisions around the final two ingredients to the methodology came later in the process, each after considerable thought to Bill James’ player ratings. James integrated an “era adjustment” based on the player’s year of birth. His argument was simple. The quality of baseball has improved gradually over time, so it was easier to dominate the further back you go. This is difficult to dispute for hockey. In Era 1 of the NHL (1917-1942), the sport competed with rival professional leagues in North America for talent, the quality of play was diluted from two World Wars, the league’s stability often in a state of flux.
Moving into Era 2 (1942-1967), the Original Six remained almost exclusively an all-Canadian circuit, farm systems were in primitive form, and a formal amateur draft to organize, disperse and develop talent was years away from the product we know today. By Era 3 (1967-1993), the earliest signs of international growth had slipped into the NHL, with occasional elite Swedes, Finns, and other Europeans filtering into the league. While the WHA continued to pose a legitimate threat for talent, its eventual collapse strengthened the league as it continually expanded its number of teams.
In Era 4 (1993-2020), Russian talent could finally play in the NHL in plain sight without requiring calculated defection. The sport has grown in the United States to the point where American-born players make up an increasingly meaningful portion of the league’s membership. Advancements in off-ice training, player development, and technology produced a new breed of athlete, while NHL scouting deftly herded international talent to the world’s best league.
With that history, as well as social trends in health, fitness and size, it becomes clear that from Howie Morenz to Maurice Richard to Guy Lafleur to Connor McDavid, the quality of play has improved dramatically. With any increase to the quality of play, it becomes more difficult to dominate. Consider your local playground league to house league to AA to AAA as a microcosm of the phenomenon. As the level rises, the best kid on the team is brought closer to the pack in terms of scoring feats or ability to dominate. The NHL is not exempt from this trend, regardless of the notions of nostalgic fans. As the league has grown in structure, development, and international footprint, the competition has indisputably risen. As such, a timeline adjustment is necessary within the player ratings. The approach taken centers on a player’s NHL debut season, as opposed to year of birth.
There is little doubt the NHL of Guy Lafleur's day was above the playing standard of Rocket Richard's time but below the standard of Connor McDavid's. “Guy Lafleur & Borje Salming 001” by rchdj10 is licensed under CC BY-NC-ND 2.0
The final consideration for the methodology involved whether subjectivity should creep into the equation. Of course, we have to accept that despite exhaustive efforts to create fair player ratings, it is unrealistic to conclude that the inputs produce a perfect holistic score. There will always be considerations that numbers may not capture, and there are limits in the data the further you go back in history. While I’m confident this methodology advances career evaluations forward, there is no right way to slice the many elements of a career into a formula.
In his Abstract, James left significant room for points to subjectively incorporate areas he felt had gaps in his system, including inequality in competition, statistically undocumented portions of a player’s career, World Series performance, leadership, clutch play, contributions above and beyond stats, and defensive value unaccounted for in the data. These subjective points allowed James to tinker with his system to align the results for what he felt he could not capture. However, a lot of ink has been spilt on this site on how critical it is to remain objective. I wanted to stand by that. It didn’t seem right to simply say ‘Jean Beliveau had a reputation as a leader,’ so he deserves 3 more points. Believe me, I tried. But my attempts relied on speculation, surely contain biases, and led to conforming to narratives – the very aspects I sought to avoid when conceptualizing a Hall of Fame standard.
One of the reasons JAWS has caught on in baseball is that it can be calculated from available information – it has no room for personalization. Subjectivity is not directly a factor in JAWS, aside from the systematic decisions in arriving at the methodology itself. But now that it is established and legitimized, fans and media can access, track and use the ratings for every player as they see fit. Once subjective points enter the conversation, this is no longer possible. When working on a previous approach with positive or negative subjective points, I found it was simply overriding a well-designed methodology to align to popular opinion. If well-conceived, the system should not need me to regularly intervene with the ratings of each player. As the methodology already accounted for playoff and international results, it had inherently covered some form of the post-season, leadership, and clutch play that James was seeking to encapsulate subjectively.
So, in the end, I did not allow for a subjective element. The only minor piece of popular opinion you can find incorporated is award voting, the final ingredient. As award shares can be systemically calculated using published information, consider it serving as a tiebreaker among closely rated players. While this adjustment can slightly move the needle as a bonus for players the system scores similarly, it is intentionally kept to a minimum. Award shares generally do not notably launch players up or down the pecking order, but they do offer a way to respect Doug Harvey‘s seven Norris Trophies or Bobby Clarke‘s stranglehold on MVP voting in the 1970s.
Now that you’re familiar with the inspiration, logic, and high-level approach to the methodology, it can be referenced more familiarly as the Pidutti Point Share system. Or PPS for short. While the alliteration of my full name and the term ‘point shares’ threads together neatly, PPPSS seemed a bit much. But PPS is short and sweet. Not as cool as JAWS surely, but the kicker is that the abbreviation PPS (post-postscript) means “written after” in Latin. This subtly notes that the system seeks to add more information after something has been written about a player’s career.
With the naming out of the way, we will now dive into the five elements of PPS, bringing the individual components to life. These five factors (seven, when factoring in three types of bonus considerations) are summarized as follows:
- Career value
- Pace value
- Peak value
- Timeline adjustment
- Bonus value
Bonus value consists of: a. Playoffs; b. International; and c. Award shares.
Together, we’ll explore PPS, the new measure for every NHL player’s career.
Era Names, Era development concepts, PPS System from Adjusted Hockey;
All other data from Hockey-Reference.com.
Adjusted Hockey 