Win Probability Added (WPA) - playoff point scoring

[overpass]

I created a WPA model with a Poisson scoring distribution for NHL playoff scoring. Briefly, the model calculates the team's probability of winning before the goal, after the goal, and assigns a value to the goal based on the difference.

Win probability is calculated using the assumption that a team is equally likely to score at any time throughout the game (i.e. a Poisson distribution), with a scoring frequency equal to the average of team GF and opponent GA.

No difference is made between goals and assists, and a player gets full credit for a goal on which he receives a point (just as he does with the points statistic).

Here are the top WPA point scoring playoff seasons since 1980.

1. Evgeni Malkin 2009 - 7.88 WPA
2. Joe Sakic 1996 - 7.08 WPA
3. Doug Gilmour 1993 - 6.77 WPA
4. Nikita Kucherov 2020 - 6.53 WPA
5. Brayden Point 2020 - 6.39 WPA
6. Connor McDavid 2024 - 6.38 WPA
7. Mario Lemieux 1991 - 6.38 WPA
8. Mark Recchi 1991 - 6.27 WPA
9. Wayne Gretzky 1993 - 6.24 WPA
10. Peter Forsberg 2002 - 6.23 WPA

I think it's interesting to see a playoff scoring leaderboard that isn't dominated by Gretzky and Lemieux, although they are present.

Here are the top WPA per game playoff seasons since 1980 (minimum 15 GP)
1. Leon Draisaitl 2022 - 0.372 WPA/GP
2. Connor McDavid 2022 - 0.354 WPA/GP
3. Mario Lemieux 1992 - 0.342 WPA/GP
4. Wayne Gretzky 1988 - 0.333 WPA/GP
5. Evgeni Malkin 2009 - 0.328 WPA/GP
6. Doug Gilmour 1993 - 0.323 WPA/GP
7. Joe Sakic 1996 - 0.322 WPA/GP
8. Wayne Gretzky 1985 - 0.312 WPA/GP
9. Peter Forsberg 2002 - 0.312 WPA/GP
10. Matthew Tkachuk 2023 - 0.298 WPA/GP

And here are the top WPA per point playoff seasons since 1980 of the seasons I've calculated (which is far from all of them).
This shows whose points were the most valuable, on average. It favours players in lower scoring eras, as well as players who frequently scored key points and rarely scored meaningless points.

1. Joe Nieuwendyk 1999, 0.258 WPA/P
2. Jarome Iginla 2004, 0.252 WPA/P
3. Matthew Tkachuk 2023, 0.248 WPA/P
4. Patrick Kane 2013, 0.231 WPA/P
5. Peter Forsberg 2002, 0.231 WPA/P
6. Evgeni Malkin 2009, 0.219 WPA/P
7. Joe Sakic 1996, 0.208 WPA/P
8. Daniel Briere 2010, 0.206 WPA/P
9. David Krejci 2013, 0.206 WPA/P
10. Sidney Crosby 2008, 0.203 WPA/P

And the lowest WPA per point (higher scoring eras, more meaningless points). Playing for dominant teams also lends itself to lower WPA.

1. Mike Bossy 1981, 0.115 WPA/P
2. Wayne Gretzky 1985, 0.119 WPA/P
3. Steve Yzerman 1998, 0.127 WPA/P
4. Wayne Gretzky 1984, 0.128 WPA/P
5. Rick Middleton 1983, 0.130 WPA/P

 

[Bear of Bad News]

I like this a lot - it's also consistent with the knowledge that in high-scoring eras, it takes more goals to "buy" a victory.

 

[overpass]

I like this a lot - it's also consistent with the knowledge that in high-scoring eras, it takes more goals to "buy" a victory.

Yeah, the scoring era adjustment is built in.

The scoring level of goals/game can be compared with the inverse of WPA/game. Both could be considered the number of goals to buy a victory.

 

[Hockey Outsider]

Thanks for sharing. This is exactly what I was planning to get to (at some point). Do you have a summary of the highest career scorers under this method? (Obviously that favours players who played in more playoff games, but curious to see how Gretzky looks compared to Crosby/Sakic/Yzerman).

 

[ScottyMascotty]

 

[overpass]

Thanks for sharing. This is exactly what I was planning to get to (at some point). Do you have a summary of the highest career scorers under this method? (Obviously that favours players who played in more playoff games, but curious to see how Gretzky looks compared to Crosby/Sakic/Yzerman).

I do not. I'm going to play around with automating the calculation more and I'll see if I can get complete numbers.

 

[MadLuke]

Not sure i get this comments:
Given that it's very unlikely that any 3rd period goal changes the probability of the game even a fraction of a percent...
Here's the final* Win Probability Added for skaters in the 2023 Playoffs.

Also it seem's like 7 Panthers added 20.13 games (so I imagine some panthers were in the negative win added to balance, for how much they did not score ?)

 

[Bear of Bad News]

Not sure i get this comments:
Given that it's very unlikely that any 3rd period goal changes the probability of the game even a fraction of a percent...
Here's the final* Win Probability Added for skaters in the 2023 Playoffs.

The score was 6-1, Vegas, going into the third period, so the premise was that nothing that happened in the third period was likely to change Vegas's near-100% win probability at that point so he may as well post the tables.

 

[MadLuke]

A that the kind of context reading something 2 years after easily lost, make sense...

 

[Hockey Outsider]

Also it seem's like 7 Panthers added 20.13 games (so I imagine some panthers were in the negative win added to balance, for how much they did not score ?)

This is an interesting issue. The win probability added doesn't necessarily add up to 100% in each game.

Let's say a team takes a 1-0 lead five minutes into the first period. Their goalie gets a shutout. The player who scored that goal might get 0.10 (ie the win probability went from 50% to say 60%). Who does the remaining 40% get allocated to? There are no other "events" aside from the passage of time.

On the other hand, there could be a game that ends up being 6-5 in overtime. The teams keep alternating 1-goal leads. The goal that makes it 1-0 five minutes into the first would still be worth 0.10. Then when the opponent ties it five minutes later, that goal would be worth slightly more (maybe 0.12). As the game progresses into the second and third, these go-ahead and tying goals might start being worth 0.2 or even 0.3. The OT winner, by definition, is worth 0.5. So the winning team could easily end up with a 150% win probability (50% for the OT goal and then perhaps 20% for each of its five goals).

The only way the system "works" (ie each game adds up to 100%) is if the skaters get credit for the goals scored, and the opposing goalie gets penalized for the decrease in win probability. That doesn't seem fair (or informative) because goalies would only have negative values.

 

[Hockey Outsider]

Here's a theoretical discussion about how the framework might work. I have some ideas for how to incorporate goaltending into it, but I'll save that for another time.

Let's take game 7 of the 2001 Stanley Cup finals as an example:

• the game starts 50-50. After the first several scoreless shifts, the game's still tied, and the probability remains 50-50. Each shift by each player is worth zero.
• Alex Tanguay scores at 7:58 of the first period. The probability is now (hypothetically) 60-40 for the Avalanche. The five skaters on the ice would get the +0.1 shared between them (ie 50% to 60% = 10%, or 0.1 as a decimal). (We can debate how that gets allocated - presumably Tanguay would get the biggest portion. Hinote, with the only assist, would get a bit less. But arguably the other three skaters - Sakic, Bourque, and Foote - deserve something. Maybe it's +0.04 for Tanguay, +0.03 for Hinote, and +0.01 for the other three). The five Devils on the ice (Stevens, Holik, etc) would get equal blame for the goal against (ie they get -0.02 each). The system "balances" so both sides still add up to zero.
• The rest of the period passes. By the end of the period, the win probability might now be 70-30 in favour of Colorado. With each scoreless shift, each Avalanche gets a very small plus, and each Devil gets a very small minus. At the end of the first period, Colorado's players are collectively +0.20 (of which +0.10 is from the goal, and +0.10 is from the passage of time). Tanguay might be up to +0.041, Bourque at +0.012, Blake (who wasn't on the for the goal, but who played lots of minutes) at +0.002, etc. New Jersey's players are collectively -0.20 (of which -0.10 is from the goal, and -0.10 is from the passage of time).
• Colorado scored two goals at approx the five and six minute marks of the second period. By that point, up 3-0 close to halfway through the game, they probably had a 90% chance of winning. Tanguay, with two goals, a secondary assist, and several zero-event shifts, might now be up to +0.20. Sakic, with a goal and an assist, might be +0.10. The scoreless Ray Bourque (who was on the ice for at least two of these goals) might be something like +0.05. Add everything up and the team would be +0.40 overall (reflecting an increase in win probability from 50% to 90%). Similarly, Holik and Stevens would be something like -0.08 each.
• the Devils score midway through the second. Sykora scores a big goal (cutting the probability from say 90-10 to 75-25) - but he was also on the ice for a goal against. My guess is he probably ends the game on the positive side, but not by much.
• with each passing shift in the third period, the probability of Colorado winning increases. It might have been 80-20 at the start of the third (with a two goal lead), but by the end it's 100-0. So there's an extra +0.2 to allocate to Colorado players, and -0.2 to allocate to New Jersey players (reflecting that a scoreless shift in the third period when the Avalanche are up 3-1, is a good thing for the winning team and a bad thing for the losing team).
• ultimately, all Colorado skaters end up at +0.5 and all New Jersey skaters end up at -0.5.

Obviously this is a theoretical discussion, but I think the framework makes sense conceptually.

 

[Bear of Bad News]

I've done something similar to what you describe with just the goalies (using a Poisson approximation). It ascribes all credit for "nothing happening" to the leading goaltender (for instance).

 

[overpass]

I've run the numbers for a number of great playoff scorers.

Here are the career leaders in playoff point WPA. The number in parentheses is how much of the WPA are from clutch timing. This portion is the secret sauce of WPA, the clutchness, the part you won't get from adjusting playoff points by scoring level or opponent.

1. Wayne Gretzky - 55.2 (+6.1)
2. Mark Messier - 46.9 (+7.9)
3. Jaromir Jagr - 38.7 (+6.4)
4. Joe Sakic - 36.8 (+3.4)
5. Sidney Crosby - 36.3 (+1.2)
6. Brett Hull - 35.8 (+4.3)
7. Evgeni Malkin - 31.6 (+2.1)
8. Jari Kurri - 33.1 (+3.3)
9. Doug Gilmour - 32.7 (+4.3)
10. Glenn Anderson - 32.4 (+4.9)

The top 10 are the same as the top 10 in playoff points, except Evgeni Malkin replaces Paul Coffey.

I found that most players ended up with positive clutch timing. Which tells me that playoff scoring does not actually follow a Poisson distribution. Star players from the last 40 years in the playoffs are more likely to score in close games than when the lead gets big.

Gretzky, Messier, and Jagr really stand out for their clutch scoring in the playoffs. Especially Messier.

Some leading defencemen
Ray Bourque - 30.7 (+3.2)
Nicklas Lidstrom - 30.6 (-1.8)
Paul Coffey - 28.3 (+2.2)
Denis Potvin - 25.7 (+2.3)
Cale Makar - 14.8 (+1.8)
Bobby Orr - 14.4 (+0.2)

Lidstrom scored 3 more playoff points than Bourque, in a lower scoring era, so he's easily ahead of Bourque on scoring-level-adjusted playoff points. But Bourque edges him in WPA based on better timing of his points.

Coffey, the #1 among defencemen in playoff points, drops to #3 in WPA due to scoring most of those points in high-scoring environments.

Cale Makar has more playoff WPA than Bobby Orr in fewer games!

Some recent playoff stars
Alexander Ovechkin - 29.5 (+5.0)
Patrick Kane - 28.7 (+3.2)
Nikita Kucherov - 28.6 (-0.2)
Connor McDavid - 18.6 (-0.4)
Leon Draisaitl - 17.6 (+0.8)

Ovechkin and Kane were clutch monsters and played in a lower scoring era, so they overtook Kucherov in WPA despite his higher playoff point total.

McDavid and Draisaitl are off to all-time starts to their playoff careers.

90s stars
Peter Forsberg - 31.2 (+0.1)
Steve Yzerman - 29.5 (-0.8)
Sergei Fedorov - 28.8 (-1.0)
Mario Lemieux - 25.8 (+1.9)

70s-80s stars
Bryan Trottier - 27.3 (+2.3)
Mike Bossy - 24.2 (+2.5)
Phil Esposito - 20.1 (-1.0)
Guy Lafleur - 19.9 (+0.4)

Original Six stars
Jean Beliveau - 27.6 (-3.6)
Stan Mikita - 26.0 (+0.8)
Gordie Howe - 25.9 (-4.4)
Bobby Hull - 22.8 (+0.5)
Maurice Richard - 21.6 (-0.6)
Ted Kennedy - 11.2 (+0.5)

The O6 stars in general had lower WPAs due to timing of their goals. Especially Howe and Beliveau, and especially in the dynasty portions of their careers.

I would take away two points from this. First, WPA might unfairly penalize a player on a dynasty team with more scoring than they need to win. Second, it's likely that Original Six playoff scoring tended towards scoring in blowouts more than one would expect. So an improved WPA model, one that uses empirical data rather than a Poisson distribution to calculate win probabilities, would need a different model for each era. I already found that using a larger home ice advantage for earlier eras was necessary.

As it is, the model may underrate goals early in the game from the Original Six era, because it was less common for the trailing team to come back at that time.

Career WPA/game leaders (of the players listed above)
1. Wayne Gretzky - 0.266
2. Connor McDavid - 0.251
3. Leon Draisaitl - 0.249
4. Mario Lemieux - 0.242
5. Joe Sakic - 0.214
6. Peter Forsberg - 0.208
7. Cale Makar - 0.206
8. Sidney Crosby - 0.202
9. Patrick Kane - 0.201
10. Mark Messier - 0.199

And some of the best stretches of roughly 60-70 playoff GP from the players above, sorted by WPA/GP.

1. Wayne Gretzky, 1981-1985 - 0.283 (+0.012)
2. Connor McDavid, 2020-2024 - 0.281 (-0.001)
3. Mario Lemieux, 1989-1993 - 0.276 (+0.023)
4. Jaromir Jagr, 1995-2000 - 0.275 (+0.050)
5. Wayne Gretzky, 1986-1990 - 0.263 (+0.027)
6. Leon Draisaitl, 2020-2024 - 0.259 (+0.020)
7. Peter Forsberg, 1999-2002 - 0.257 (+0.025)
8. Wayne Gretzky, 1991-1997 - 0.257 (+0.049)
9. Joe Sakic, 1996-1999 - 0.249 (+0.028)
10. Sidney Crosby, 2008-2012 - 0.248 (+0.014)

Doug Gilmour, 1993-1997 - 0.244 (+0.039)
Alexander Ovechkin, 2008-2013 - 0.242 (+0.056)
Evgeni Malkin, 2008-2012 - 0.241 (+0.030)
Brett Hull, 1988-1993 - 0.235 (+0.052)
Gordie Howe, 1955-1964 - 0.231 (-0.023)
Guy Lafleur, 1975-1979 - 0.231 (+0.005)
Mark Messier, 1988-1991 - 0.230 (+0.034)
Mike Bossy, 1980-1983 - 0.225 (+0.016)
Nikita Kucherov, 2020-2022 - 0.222 (0.000)
Patrick Kane, 2012-2016 - 0.216 (+0.035)
Bobby Orr, 1970-1975 - 0.215 (+0.003)
Cale Makar, 2020-2024 - 0.214 (+0.020)
Maurice Richard, 1944-1951 - 0.213 (+0.027)
Jari Kurri, 1987-1990 - 0.209 (+0.043)
Bobby Hull, 1961-1971 - 0.207 (0.000)
Phil Esposito, 1969-1975 - 0.203 (-0.028)
Bryan Trottier, 1980-1983 - 0.202 (+0.009)
Glenn Anderson, 1985-1988 - 0.195 (+0.032)
Denis Potvin, 1976-1980 - 0.189 (+0.039)
Stan Mikita, 1961-1967 - 0.187 (+0.013)
Paul Coffey, 1989-1995 - 0.185 (+0.022)
Jean Beliveau, 1965-1971 - 0.184 (-0.003)
Jean Beliveau, 1954-1960 - 0.181 (-0.050)
Sergei Fedorov, 1992-1997 - 0.176 (-0.004)
Ted Kennedy, 1947-1951 - 0.174 (+0.008)
Steve Yzerman, 1984-1994 - 0.171 (+0.020)
Ray Bourque, 1988-1991 - 0.169 (+0.030)
Paul Coffey, 1983-1986 - 0.165 (+0.010)
Steve Yzerman, 1998-2002 - 0.154 (-0.032)
Nicklas Lidstrom, 1998-2002 - 0.150 (-0.008)

How about Wayne Gretzky? His entire playoff career, excluding his rookie year of 1980, breaks evenly into three sections of 65-70 playoff games. And each of those 3 sections finished in the top 10 above. As his points per game dropped later in his career, his clutchness increased to compensate.

McDavid's 2020-2024 could be argued to be the best playoff point scoring stretch of all time, but WPA drops it just below peak Gretzky due to clutch timing. Lemieux and Jagr round out the top 4. Bet you didn't expect to see Jagr there, but take a look at his playoff scoring logs from the mid to late 90s. Very few points in blowouts, and a ton of tying and go-ahead points.

 

[BraveCanadian]

Bet you didn't expect to see Jagr there, but take a look at his playoff scoring logs from the mid to late 90s. Very few points in blowouts, and a ton of tying and go-ahead points.

After going back to re-examine his playoff career with contemporary newspaper research a few years back, I did!

He is constantly put down for not winning a cup as "the guy" but he did about as much as anyone could on those post-dynasty Pens.

You're likely on the right track here because a lot of these numbers and names attached match the ones people have long said were great playoff performers.

 

[overpass]

Here's a look under the hood for a few playoff seasons to get an idea of how these numbers are calculated and what they mean.

Gretzky and Messier, 1984

First, let's take Wayne Gretzky and Mark Messier's 1984 playoff, in which the Oilers won the Cup for the first time. Gretzky outpointed Messier 35-26, but Messier was awarded the Conn Smythe Trophy. And in fact, WPA rates Messier's 26 points scored as more valuable than Gretzky's 35 points, 4.59 to 4.36. When you add in Messier's matchup role against other top centres, and his physical presence, it's not so hard to see why the voters considered Messier's playoff more valuable.

I'll sort these points into 5 buckets based on WPA value.

0.000 - 0.049 - relatively low value, usually scored late in blowouts, or very late in 1-2 goal games.
0.050 - 0.099 - moderately low value
0.100 - 0.199 - average value, usually scored in the first half of games. Goals in high scoring eras tend toward the lower part of this range, goals in low scoring eras toward the higher part of this range.
0.200 - 0.299 - high value
0.300 - 0.500 - very high value, usually tying goals or go-ahead goals in the second half of games.

Wayne Gretzky 1984
8 points between 0.000 and 0.049 (totalling 0.08 WPA)
Example: G1 vs Winnipeg, assisted on Ken Linseman goal 7:24 into period 2 to make the score 6-1. Increased win probability from 0.953 to 0.983.

3 points between 0.050 and 0.099 (totalling 0.22 WPA)
Example: G3 vs Minnesota, scored a SH goal 10:42 into period 3 to make the score 8-5. Increased win probability from 0.908 to 0.981.

19 points between 0.100 and 0.199 (totalling 2.59 WPA)
Example: G4 vs Islanders, scored a goal 1:53 into period 1 to make the score 1-0. Increased win probability from 0.587 to 0.714.

3 points between 0.200 and 0.299 (totalling 0.65 WPA)
Example: G7 vs Calgary, assisted on Glenn Anderson goal 13:50 into period 2 to tie the game 4-4. Increased win probability from 0.401 to 0.605.

2 points between 0.300 and 0.500 (totalling 0.82 WPA)
Example: G2 vs Calgary, scored a goal 19:15 into period 3 to tie the game at 5-5. Increased win probability from 0.030 to 0.507.

Mark Messier 1984
2 points between 0.000 and 0.049 (totalling 0.06 WPA)
Example: G4 vs Minnesota, assisted on a Jari Kurri EN goal 19:11 into the 3rd period to make the score 3-1. Increased win probability from 0.972 to 0.999.

0 points between 0.050 and 0.099

19 points between 0.100 and 0.199 (totalling 2.78 WPA)
Example: G3 vs Islanders, scored a goal 8:38 into the 2nd period to tie the game 2-2. Increased win probability from 0.377 to 0.563.

3 points between 0.200 and 0.299 (totalling 0.77 WPA)
Example: G6 vs Calgary, scored an unassisted PP goal 6:53 into the third period to tie the game 4-4. Increased win probability from 0.253 to 0.527.

2 points between 0.300 and 0.500 (totalling 0.98 WPA)
Example: G2 vs Winnipeg, assisted on Randy Gregg OT goal. Increased win probability from 0.500 to 1.000.

When you bin the points for Gretzky and Messier in this way, it's pretty clear why Messier came out ahead. Gretzky had 11 low value points, and Messier had only 2. They had the same number of average and high value points. And within those average to high value bins, Messier's points tended to be higher WPA than Gretzky's, which gave him the overall lead.

Maurice Richard and Gordie Howe

I'll also compare Maurice Richard's 1951 playoff, an exceptionally valuable one by WPA, with Gordie Howe's 1955 playoff. Howe set a playoff scoring record with 20 points in 1955, but WPA rates those 20 points as much less valuable than Richard's 13 points in 1951. Richard's 13 points were worth 3.86 WPA and Howe's 20 points were worth 2.58 WPA.

Maurice Richard 1951
0 points between 0.000 and 0.049

0 points between 0.050 and 0.099

5 points between 0.100 and 0.199 (totalling 0.84 WPA)
Example: G4 vs Toronto, scored a goal 14:41 into the first period to tie the game 1-1. Increased win probability from 0.257 to 0.447.

2 points between 0.200 and 0.299 (totalling 0.46 WPA)
Example: G5 vs Toronto, scored a goal 8:56 into the second period to make the score 1-0. Increased win probability from 0.397 to 0.641.

6 points between 0.300 and 0.500 (totalling 2.57 WPA)
Example: G4 vs Toronto, assisted on an Elmer Lach goal 13:49 into the third period to tie the game 2-2. Increased win probability from 0.093 to 0.485.

Gordie Howe 1955
6 points between 0.000 and 0.049 (totalling 0.10 WPA)
Example: G1 vs Montreal, assisted on a Ted Lindsay EN goal 19:42 into the third period to make the score 4-2. Increased win probability from 0.994 to 1.000

1 point between 0.050 and 0.099 (totalling 0.07 WPA)
Example: G2 vs Montreal, scored a goal 17:11 into the first period to make the score 4-0. Increased win probability from 0.882 to 0.956.

10 points between 0.100 and 0.199 (totalling 1.55 WPA)
Example: Game 2 vs Montreal, assisted on a goal by Ted Lindsay 9:57 into the first period to make the score 2-0. Increased win probability from 0.720 to 0.886.

2 points between 0.200 and 0.299 (totalling 0.46 WPA)
Example: G1 vs Montreal, assisted on a PP goal by Alex Delvecchio at 14:00 of the second period to tie the score 1-1. Increased win probability from 0.273 to 0.527.

1 point between 0.300 and 0.500 (totalling 0.39 WPA)
Example: G1 vs Montreal, assisted on a goal by Vic Stasiuk 13:05 into the third period to tie the score 2-2. Increased win probability from 0.121 to 0.511.

Per WPA, the six points below from Maurice Richard's 1951 playoffs had basically the same value as all Gordie Howe's 20 points in the 1955 playoffs.
1951, Game 1 vs Detroit - Richard assisted a Bert Olmstead goal, 9:08 of 3rd, tied game 2-2 (0.319)
1951, Game 1 vs Detroit - Richard scored an OT goal (0.500)
1951, Game 2 vs Detroit - Richard scored an OT goal (0.500)
1951, Game 6 vs Detroit - Richard scored a goal, 9:39 of 3rd, 2-1 lead (0.358)
1951, Game 2 vs Toronto - Richard scored an OT goal (0.500)
1951, Game 5 vs Toronto - Richard assisted an Elmer Lach goal, 13:49 of third, tied game 2-2 (0.392)

 

[overpass]

And here's a Gretzky-Gretzky comparison, to show why his 20 points in the 1997 playoffs were actually more valuable than his 38 points in the 1983 playoffs.

Wayne Gretzky 1983
10 points between 0.000 and 0.049 (0.09 WPA)
3 points between 0.050 and 0.099 (0.23 WPA)
23 points between 0.100 and 0.199 (3.14 WPA)
1 point between 0.200 and 0.299 (0.22 WPA)
1 point between 0.300 and 0.500 (0.34 WPA)
38 points, 4.03 WPA (0.106 WPA per point, on average)

Wayne Gretzky 1997
1 point between 0.000 and 0.049 (0.04 WPA)
1 point between 0.050 and 0.099 (0.09 WPA)
10 points between 0.100 and 0.199 (1.64 WPA)
5 points between 0.200 and 0.299 (1.09 WPA)
3 points between 0.300 and 0.500 (1.34 WPA)
20 points, 4.20 WPA (0.210 WPA per point, on average)

 

[overpass]

Why does WPA rate Nikita Kucherov's 167 points just below Patrick Kane's 138 and Alex Ovechkin's 141?

Kucherov scored 15-20 more low-value points that added very little to his WPA.

Kucherov also led in average value points, but Kane and Ovechkin were well ahead in high value points. Partly because they played in lower scoring environments than Kucherov, on average. And partly because more of their points were scored at high value times.

Total
Kucherov 167 (28.58 WPA) (average point=0.171 WPA)
Ovechkin 141 (29.5 WPA)
Kane 138 (28.63 WPA)

Points where WPA<0.100
Kucherov 38 (1.1 WPA)
Kane 22 (1.0 WPA)
Ovechkin 15 (0.6 WPA)

Points where WPA=0.100-0.199
Kucherov 84 (13.1 WPA)
Ovechkin 64 (10.3 WPA)
Kane 59 (9.3 WPA)

Points where WPA=0.200-0.299
Ovechkin 39 (9.2 WPA)
Kane 32 (7.5 WPA)
Kucherov 26 (6.2 WPA)

Points where WPA>0.300
Kane 25 (10.9 WPA)
Ovechkin 23 (9.4 WPA)
Kucherov 19 (8.2 WPA)

Average WPA value per point
Ovechkin 0.209
Kane 0.207
Kucherov 0.171

 

[overpass]

Here's a look at the top 4 stretches of best playoff point scoring ever, per WPA. You can see Jagr overcomes a lower point total to join this group, based on having about twice as many high-value points as the rest, and far fewer low-value points.

If you wanted to put a higher value on goals vs assists, Lemieux and Jagr would probably be 1-2.

Wayne Gretzky, 1981-1985 (153 points in 67 games)
0.00-0.49 WPA - 17 G, 29 A, 46 P (1.4 WPA)
0.100-0.199 WPA - 32 G, 61 A, 93 P (13.1 WPA)
0.200-0.299 WPA - 3 G, 5 A, 8 P (1.9 WPA)
0.300-0.399 WPA - 2 G, 4 A, 6 P (2.5 WPA)
Total - 54 G, 99 A, 153 P, 18.9 WPA (0.282/game)

Connor McDavid, 2020-2024 (108 points in 61 games)
0.00-0.49 WPA - 4 G, 17 A, 21 P (0.8 WPA)
0.100-0.199 WPA - 21 G, 45 A, 66 P (10.0 WPA)
0.200-0.299 WPA - 5 G, 9 A, 14 P (3.2 WPA)
0.300-0.399 WPA - 2 G, 5 A, 7 P (3.2 WPA)
Total - 32 G, 76 A, 108 P, 17.2 WPA (0.281/game)

Mario Lemieux, 1989-1993 (115 points in 60 games)
0.00-0.49 WPA - 8 G, 15 A, 23 P (0.7 WPA)
0.100-0.199 WPA - 38 G, 37 A, 75 P (11.2 WPA)
0.200-0.299 WPA - 4 G, 7 A, 11 P (2.5 WPA)
0.300-0.399 WPA - 2 G, 4 A, 6 P (2.1 WPA)
Total - 52 G, 63 A, 115 P, 16.6 WPA (0.276/game)

Jaromir Jagr, 1995-2000 (83 points in 61 games)
0.00-0.49 WPA - 4 G, 4 A, 8 P (0.7 WPA)
0.100-0.199 WPA - 23 G, 23 A, 46 P (7.1 WPA)
0.200-0.299 WPA - 7 G, 8 A, 15 P (3.6 WPA)
0.300-0.399 WPA - 8 G, 6 A, 14 P (5.6 WPA)
Total - 42 G, 41 A, 83 P, 16.8 WPA (0.275/game)

 

[Hockey Outsider]

@overpass - great research, as always. Like you noted, many of the stars from the Original Six era have surprisingly mediocre results. Was there anybody who looked unexpectedly good?

 

[seventieslord]

@overpass - great research, as always. Like you noted, many of the stars from the Original Six era have surprisingly mediocre results. Was there anybody who looked unexpectedly good?

If they weren't the ones scoring important goals, someone else had to be, right?

 

[Hockey Outsider]

If they weren't the ones scoring important goals, someone else had to be, right?

The other point to consider - maybe there were just fewer lead changes in the Original Six era? Like I talked about earlier (post 10), each game doesn't add up to 100%. Higher scoring eras conceivably have more WPA's awarded per game, by virtue of there being more lead changes (even if the impact of each goal, on average, is less in a higher-scoring environment).

(I wonder if WPA, divided by the average number of WPA per game from that season, might be a rough estimate of the scoring environment. Not "scoring environment" in terms of the number of goals per game - which is already accounted for in WPA - but a reflection of the number of lead changes).

 

[Bear of Bad News]

Good point - and by that nature, in games with more lead changes, the players defending against goals would accrue more negative WPA.

 

[overpass]

@overpass - great research, as always. Like you noted, many of the stars from the Original Six era have surprisingly mediocre results. Was there anybody who looked unexpectedly good?

No, every player from the Original Six for whom I have run the numbers had WPA results either equal or worse to what you would expect from their scoring totals. Maurice Richard had some very clutch results in the late 40s and early 50s, but that was offset by his high scoring totals from the war years and the dynasty being less clutch.

If they weren't the ones scoring important goals, someone else had to be, right?

I'm glad you asked that question, because I don't think it's the case. The number of important goals could legitimately vary by era, separately from the scoring level.

Let's start on a single game level. When looking at points scored alone, the WPA for each game does not necessarily add up to 0.500. Consider the following games, all between the same two teams and in the same scoring environment, where 3 goals scored sum to a different WPA total.

1. One team scores 5 minutes into the game, holds a 1-0 lead for most of the game, and then scores an empty net goal with 2 minutes to go and another with 1 minute to go. WPA for the first goal might be 0.200, the second goal might be 0.050, and the third goal might be 0.005. Which sums to 0.255 for the three goals in the game.

2. One team scores with 3 minutes to go in the third period, followed by another goal with 2 minutes to go and another goal with 1 minute to go. WPA for the first goal might be 0.400 instead of 0.200, so the three goals would sum to 0.455.

3. Same goals scored as in example one, but this time the opponent scores halfway through the first period. Now the second goal scored with 2 minutes to go is a go-ahead goal and worth maybe 0.450 WPA rather than 0.050. So the three goals would sum to 0.655.

@Hockey Outsider discussed above in his post how a full framework would include defensive contributions. You could imagine in the examples above how the team in example 1 would get more defensive credit for holding a 1-0 lead for over 50 minutes than the team in example 2, who only held a lead for three minutes. And the team in example 3 might have negative defensive WPA for allowing one goal in seven minutes with the lead.

Going back to the points-only model I'm using, the model uses Poisson distribution to calculate probabilities, which assumes that goals are equally likely to be scored at any time in the game. If players as a group from a particular era have higher or lower WPA. then we know it's not random, likely because score effects are taking place in one direction or another. If players as a group have lower WPA, then the leading team is more likely to score than the trailing team. If players have higher WPA as a group, then the trailing team is more likely to score.

I ran some numbers for each era to see which eras had lower WPA than expected and which had higher. This is based only on the players who I have calculated, not all goals scored.

Eras that had lower WPA than expected (i.e. it was rare for trailing teams to come back)
1952-1960 - 83%
1961-1970 - 97%

Eras with about as much WPA as expected
1998-2004 - 100%

Eras with more WPA than expected (i.e. it was common for trailing teams to come back)
1946-1951 - 118%
1971-1979 - 108%
1980-1985 - 108%
1986-1992 - 119%
1993-1997 - 110%
2006-2011 - 108%
2012-2017 - 106%
2018-2024 - 106%

I remember the 1998-2004 era as one where it was particularly difficult for a trailing team to score, so it's interesting to see this method rate it as neutral. In fact it was more difficult for the trailing team to score than in most other eras, as most other eras have actually been more favourable to the trailing team scoring. Possibly because referees were more likely to call penalties on the leading team, or trailing teams pressed harder to score. The Original Six era was the exception to this rule. And during the period of the 1950s dynasties, it was more likely than ever that goals were scored by the team leading rather than trailing. Maybe the lack of parity was a factor, or the defensive and goaltending strength of the top teams. The 1960s were also more difficult than most eras for the trailing team.

A full WPA model would give more credit to the goaltenders and skaters of the 1950s and 1960s for their defensive strength when leading. Possibly with the result that the top scorers were relatively less valuable in that environment than they were at other times in league history. And that might actually make sense if you see the Original Six league as a league full of strong players and teams where it was more difficult for one player to make an impact than it was in later eras, when the average player and team wasn't as strong.