Picture a hockey player at their absolute best. That sweet stretch somewhere between an unpolished apprenticeship and the first signs of betrayal from one’s physical gifts. A fully bloomed talent contributing as greatly as they ever will to an NHL team. Prime. Peak. Pinnacle. Regardless the descriptor, this is a revealing point in a career. Yet, for each individual player, it has never been formally identified or quantified. Meet High Noon.
We’ve gone to great lengths to develop PPS to be the primary driver in setting Hall of Fame standards by era and position. We’ve created the single digit that tries to best encapsulate a hockey career. But there is additional meaning in identifying a player’s year-to-year place in the league’s hierarchy. When a player approaches retirement, it’s common to reflect on their greatest height in the NHL. In some cases, this is obvious – Terry Sawchuk, Bobby Orr, or Mario Lemieux stood far above their peers at their best. Their greatest height was being the greatest player at their position. That’s as high as it gets. For those outside the league’s inner circle, perhaps this is a place in the top 10, 50, or 100 at their position at the peak of their powers. Regardless the level attained, we want to identify their league-wide rank throughout their career.
In his 1995 book, Whatever Happened to the Hall of Fame?, Bill James detailed the “Keltner List,” a series of 15 questions deployed on baseball Hall of Fame candidate Ken Keltner. While not meant to definitively identify a Hall of Famer, the list served to ask yes or no questions to frame the decision for players under consideration. The first three questions on the Keltner List jumped out at me, eventually leading to the development of the High Noon metric for NHL players. Those questions were as follows:
- Was he ever regarded as the best player in baseball?
- Was he the best player on his team?
- Was he the best player in baseball at his position?
If a player was never among the top players in the league or at their position, let alone their individual team, it becomes more difficult to push their Hall of Fame candidacy. These questions – and those that followed on the list – were meant by James to be subjective and provoke thought. Yet, the questions seemed quantifiable. They could be answered. High Noon was born.
Originally labeled “Apex,” the measure needed more flair. When it comes to flair, look no further than my favourite band, Hamilton, Ontario’s Arkells. While their third studio album, High Noon, was named both for its political undertones and a 1950’s Western movie of the same name, the title fit the new stat seamlessly. Merriam-Webster defines high noon as “the most advanced, flourishing, or creative stage or period.” This was exactly the data point I was looking to identify within a hockey career.
Fancy naming aside, how do we calculate High Noon and how does it work?
Now, the day, week, or month at one’s best cannot be undisputedly pinpointed. So, our approach is best taken using a rolling average, that is, a smoothed average covering a particular window of time. Rolling averages eliminate short-term fluctuations and noise. A red-hot month of play or even a full season on a scorching power play should not place a player among the NHL’s best. We want the sample size to be sufficient to encompass a reasonable body of work.
Our approach then is to take a three-year average of a player’s Adjusted Point Share pace (AdPS+). After each season, a player is assigned a value equal to their AdPS+ covering the preceding three NHL seasons. This value can be compared to all others at their position (i.e. forward, defenceman, or goaltender) over the same three-season stretch. There is some similarity in approach to the world rankings used in professional golf (two-year rolling average) and tennis (one-year rolling average). However, the rankings in those sports award points based on tournament results and roll weekly. High Noon rolls annually, given a week or month has no significance in hockey, a team sport.
Consider High Noon to be the world rankings of hockey, systematically deployed to individual players like golf and tennis.
Let’s work through an example to bring the concept to life. We’ll demonstrate on respected blueliner, Sergei Gonchar.
SERGEI GONCHAR, HIGH NOON SCORE BY SEASON
|High Noon Score||X||X||8.6||8.2||8.2||9.6||11.3||11.2||10.8||10.4||9.1||8.3||8.7||9.7||9.8||7.6||6.3||6.4||5.3||5.3|
AdPS = Adjusted Point Shares; AdGP = Adjusted Games Played; High Noon Score = [Total AdPS ÷ Total AdGP x 82] over the preceding 3 seasons
Gonchar’s three-year pace, identified above by season, is what we will call his High Noon score. You’ll notice the “X” in his first two years since a player must appear in three NHL seasons to qualify. His top High Noon score, and therefore, his best three-season stretch of play in the NHL is 11.3, encompassing his age-24 to 26 seasons. While High Noon score values his performance, we are also interested in a player’s High Noon rank – their place in the league in that particular season.
SERGEI GONCHAR, HIGH NOON RANKING BY SEASON
|High Noon Rank||19||17||17||9||4||5||5||2||9||12||9||3||4||14||33||34||60||72|
High Noon Rank = High Noon Score ranked among NHL by position
Remarkably, Gonchar ranks as a top-12 NHL defenceman for a full decade (covering 12 seasons, given it takes three seasons to first be eligible). It’s not until his age-37 season that his ranking drops out of the top 20. His best standing among all blueliners is at #2, encapsulating his only two post-season all-star selections (second team). Sergei Gonchar’s High Noon is therefore his age-27 to age-29 seasons where our methodology labels him the second-best defencemen in the NHL.
A few takeaway for Gonchar’s case:
- By individual performance, Gonchar’s score was better in each of the three seasons before reaching his High Noon. However, some combination of Chris Pronger, Al MacInnis, Nicklas Lidstrom, and Rob Blake had higher scores at that time, until Gonchar crept past them.
Interestingly, Gonchar’s only two second-team all-star selections come immediately after ages 27 and 28. At age 24, Gonchar scored at a 37-goal adjusted pace, yet only played in 53 games.
Caveat minimum 50% of the schedule over the 3-year period (i.e., min. 123 AdGP, or average of 41 AdGP per season). Same for goalies. While argument can be made to lower it (i.e., 60/82 rate seen elsewhere, or 90 AdGP), did not want to potentially include backup goalies or goalies with one year as a starter in a 3-year cycle as vying for status as best in NHL. By this measure, a goalie would likely need at least 2 healthy seasons in starter role in the cycle. playoffs
If a players plays <123 AdGP, they simply do not qualify – do not have a score for the 3-year cycle. Could have one in the year before or after – use an example.
Completely missed seasons do not hurt someone – Kucherov injury, Lemieux semi-retirement, Thornton lockout. Jagr 3-year sabbatical to KHL, Lafleur retirement. Unfair to do so, war, major injury. Forsberg possible example. Pronger. 3 year hiatus max, presuming returning as similar player before departure. A lot of such players don’t qualify anyway (missed games).
Exclude extreme examples – Hull left a 33-year old 50-goal superstar, returned 8 years later after WHA. Howe returned at 51-year old, not same 5th best player that left nearly decade earlier. Keon. Johnny Mowers, returning to play 7 games. War commonly 2 seasons (prime years, 20s) or 2+ or 3 in some cases.
3-year cycle means first and second seasons won’t register – need 3 years of results to carry an Apex Value (Apex still make sense?). Gretzky or Crosby… clearly top player. Lasts longer – career fading but good in previous 2 years would still have a declining but solid rating (example)
Career vs peak mentioned throughout
We recognize there is more to any debate. Sustained excellence, best at position can heavily influence people, even if a career on balance was blah. Similarly big totals compiled can overlook a player never was elite.
Peak value attempts to cover this – assign a score to best 7-year stretch, which weighs heavily in properly evaluating shorter career players
Odd guy like MacInnis or Orr almost no games in a year still best in 3-year cycle. Makes sense as play never fell off.
Apex top 1,2,3 NOT in Hall of Fame? Top # times NOT?
Part of HOF evaluation:
F = Top 15
D = Top 10
G = Top 5
It is of course more difficult to be a top 5 G presently. Or even a top 5 F given the pool of players, but HOF really hasn’t responded to growth of player pool so logical way.
If a player is never in the top 15 F, should they be in HOF? Examples
When evaluating a new player, what was their max Apex? Marleau?
Which eligible forward was top 15 at their position most times and is NOT in the HOF? Turgeon (10 times, reaching as high as 7th). D – Hollett, Tremblay) top 10 for 11 seasons. Gonchar modern candidate (9 times).
Which HOF forwards never reached the top 15 even once? One big surprise – Sundin (others Gainey 111, Carbonneau 82, Pulford 16, Duff 16, Laprade 16, Sundin 16). Andreychuk , Gartner 15; Gilmour, Francis 14. D Lowe (41), Hap Day (13), Leo Boivin (12) not to reach top 9.
Which players reached #1 spot but not elected? Heatley (1), Gaborik, Aurie, Naslund, Simmer (2). Does this mean HOF? No, of course not. D – Hollett (1), Tremblay, Gonchar, Talbot (2)
Using Apex – existing HOF (all-time greats, typical HOF, on the fence HOF)
All-time greats – D Shore (top 3 every eligible season). Compare that to others. Consistency.
EDDIE SHORE, APEX RANKING BY SEASON
Existing HOF – F Robitaille (eligible seasons 17, top 1 0, top 5 0, top 15 6x), best 7th
Shows power of the scoring environment (Robitaille is 16th best F in 1993-94, coming off 151 Goals in 3 years, never got a Hart vote). Compare to Elias later
LUC ROBITAILLE, APEX RANKING BY SEASON
On the fence HOF – F Barber (eligible seasons 10, top 10 1x, top 15 3x), best 7th
BILL BARBER, APEX RANKING BY SEASON
Similar to Robitaille through season 12. Would it matter if Robitaille played 7 more years, compiling stats as the 50th to 150th best F in NHL? Debtable. But Robitaille finds a second wind, being a top 30 F from seasons 13 through 16, including reaching as high as #14 again, before full decline sets in. Barber, a top 15 forward for 3 years only – good enough? A top 30 forward essentially his entire career? Is THAT good enough?
3 more types of players – HOF candidates. The eligible player, the retired or nearly retired player, the active player
CURTIS JOSEPH, APEX RANKING BY SEASON
Best goalie in the NHL covering bulk of his time in St. Louis. Top 10 goalie 8x. Compare to some other G from era with Stanley Cups that often sneak into HOF discussions – Vernon (peaked at #7, only time in the top 10), Osgood (peaked at #12).
PATRIK ELIAS, APEX RANKING BY SEASON
Elias is 16th best F in 2000-01, has just 92 G in preceding 3 years. They both have same value and standing in the NHL, but one of them banks 59 more G toward a HOF case. Similar career to Elias, albeit a little healthier, but both 74/75 Point Fs that history will never see that way. Elias loses entire year to lockout at age 28.
PATRICK MARLEAU, APEX RANKING BY SEASON
Top 15 player 1 time. Top 30 player 3 x. Took 9 years to be a top 50 player, stayed there for a solid 8 years, spent last 7 years chasing records averaging being 145th best F in NHL. Impressive to be a middle six forward for that long late in a career, not an elite player. Entire case is based on career totals – that’s okay. But pretending Marleau was a franchise player for any stretch of time would not be true.
ROMAN JOSI, APEX RANKING BY SEASON
Top 10 D 6 years in a row, and given rolling age likely has a few more seasons. With the rush to put more forwards in constantly, these are type of D that Era 4 will produce nearly a dozen of that should almost certainly deserve inclusion.