Adjusting playoff scoring for defense faced

[MadLuke]

As talked about in a thread, it would be maybe interesting to adjust playoff scoring over player career around a weighted defensive score of the opponent they face, 6 games against a 3 goals against, 4 games against a 4 goals a game, faced a average 3.4 goals a games defensive team that year.

player nan	seasson	gameplayed	average defense	points	adjustedpts	ppg adj
Gretzky	19801981	9	3.13	21	24.7	2.74
Gretzky	19841985	18	3.68	47	47.0	2.61
Lemieux	19911992	15	3.26	34	38.4	2.56
Crosby	20212022	6	2.49	10	14.8	2.47
Gretzky	19821983	16	3.56	38	39.3	2.46
Gretzky	19871988	19	3.53	43	44.8	2.36
Crosby	20172018	12	2.89	21	26.8	2.23
Lemieux	19901991	23	3.29	44	49.3	2.14
Gretzky	19881989	11	3.46	22	23.4	2.13
Lemieux	19951996	18	2.74	27	36.3	2.02
Gretzky	19961997	15	2.44	20	30.2	2.01
Crosby	20072008	20	2.52	27	39.5	1.98
Crosby	20092010	13	2.74	19	25.5	1.96
Gretzky	19921993	24	3.20	40	46.1	1.92
Gretzky	19811982	5	4.61	12	9.6	1.92
Bossy	19801981	18	3.92	35	32.9	1.83
Crosby	20112012	6	2.74	8	10.7	1.79
Bossy	19791980	16	2.99	23	28.4	1.77
Gretzky	19851986	10	4.01	19	17.5	1.75
Gretzky	19951996	13	2.61	16	22.6	1.74

It seem to give something of the sorts, it is adjusted using Gretkzy 47 pts record weighted average team regular season defense he faced that year.

Only Bossy, Gretzky, Lemieux, Crosby as a try-out was made, making it for all players would not be particularly hard just a bit long for the code that fetch the api to run.

 

[MadLuke]

Career

Gretzky has 410 adjusted pts in 208 games for 1.97 adjppg.
Lemieux has 205 in 107 for 1.91 adjppg.
Bossy 169 in 129 for 1.31
Crosby has 285 in 180 for 1.58 adjppg

Player	Adjusted Points	Games	Adj PPG
Gretzky	410	208	1.97
Lemieux	205	107	1.91
Bossy	169	129	1.31
Crosby	285	180	1.58

If we limit Gretzky-Lemieux to their "prime" it would not be easy, Gretzky 1997 is fantastic considering he faced in average a defense that allowed 2.44 goal a game during the regular seasons (if my code worked well)

 

[authentic]

Career

Gretzky has 410 adjusted pts in 208 games for 1.97 adjppg.
Lemieux has 205 in 107 for 1.91 adjppg.
Bossy 169 in 129 for 1.31
Crosby has 285 in 180 for 1.58 adjppg

Player	Adjusted Points	Games	Adj PPG
Gretzky	410	208	1.97
Lemieux	205	107	1.91
Bossy	169	129	1.31
Crosby	285	180	1.58

If we limit Gretzky-Lemieux to their "prime" it would not be easy, Gretzky 1997 is fantastic considering he faced in average a defense that allowed 2.44 goal a game during the regular seasons (if my code worked well)

MadLuke this is precisely what I’ve been looking for. Well done my friend.

Forsberg and McDavid are the next ones I’m most curious about.

 

[MadLuke]

Forsberg and McDavid are the next ones I’m most curious about.

Some that I was curious Howe-and the MTL 06 that never played against themselves, but they look really good in the early result (I had 2,600 of 8,000 something players from the API done this morning), the teams they faced probably tended to have good defense already those years they made it.

Jagr I feel like will be one of the best of all-time playoff goalscorer relative to the defense, so to the DPE players like Forsberg could look good.

I will do the same for goals scoring and I think something like the 6 best playoffs and the best X consecutive series (or games or playoff run, something to limit it to players prime in that way).

 

[authentic]

Some that I was curious Howe-and the MTL 06 that never played against themselves, but they look really good in the early result (I had 2,600 of 8,000 something players from the API done this morning), the teams they faced probably tended to have good defense already those years they made it.

Jagr I feel like will be one of the best of all-time playoff goalscorer relative to the defense, so to the DPE players like Forsberg could look good.

I will do the same for goals scoring and I think something like the 6 best playoffs and the best X consecutive series (or games or playoff run, something to limit it to players prime in that way).

Yes it would be interesting indeed to see how Howe and the rest of the O6 stack up. Can’t wait for the results.

 

[jigglysquishy]

How are you scraping the data? I tried to put together data for how Howe/Habs performed against the common opponents in the regular season. Basically, remove the data skew that Beliveau and Richard never played against Harvey/Plante/H.Richard.

Doing it manually was exhausting so I gave up.

 

[MadLuke]

How are you scraping the data?

Mostly I ask Grok how to do it frankly, first I used the nhl api,
to get all the seasons log of all teams for all seasons in python to create an sqlite database with it.

Next I go fetch the game log of all players ("https://api-web.nhle.com/v1/player/{playerId}/game-log/{season}/{gameType}") in another database, playoff are a gametype '3'.

"I" (the AI translate English of what I want into) make sql request to adjust each series scoring for that series opponent goal against average of that year. I did manually validate a couple of results and the code seem right.

The adjustment was made using the average weighted goals against Gretzky faced when he did set the 47 points record.

 

[MadLuke]

Process done, result are surprising in a way and not at all in another way

Most adjusted playoff points (per games value are the adjusted one):

full_name	games	points	adjpts	ppg	goals	adjgoals	gpg	player_rank
Wayne Gretzky	208	382	409.8	1.97	122	130.9	0.63	1
Mark Messier	236	295	317.8	1.35	109	116.8	0.49	2
Sidney Crosby	180	201	284.9	1.58	71	99.4	0.55	3
Joe Sakic	172	188	274.4	1.6	84	122.4	0.71	4
Jaromir Jagr	208	201	273.8	1.32	78	105.3	0.51	5
Evgeni Malkin	177	180	257.7	1.46	67	95.4	0.54	6
Nicklas Lidstrom	263	183	256.7	0.98	54	75.2	0.29	7
Peter Forsberg	151	171	252.6	1.67	64	93.7	0.62	8
Jean Beliveau	162	176	249.3	1.54	79	111.2	0.69	9
Gordie Howe	157	160	246.9	1.57	68	105.4	0.67	10
Brett Hull	202	190	243.5	1.21	103	129.1	0.64	11
Jari Kurri	200	233	238.6	1.19	106	109	0.55	12
Sergei Fedorov	183	176	237.3	1.3	52	71.2	0.39	13
Nikita Kucherov	147	167	236.9	1.61	53	75.6	0.51	14
Steve Yzerman	196	185	236	1.2	70	88.4	0.45	15
Marian Hossa	205	149	224.2	1.09	52	77.8	0.38	16
Glenn Anderson	225	214	222.4	0.99	93	96.4	0.43	17
Paul Coffey	194	196	213.3	1.1	59	63.5	0.33	18
Ray Bourque	214	180	212.2	0.99	41	47.4	0.22	19
Doug Gilmour	182	188	207.1	1.14	60	65.4	0.36	20
Stan Mikita	155	150	205.7	1.33	59	80	0.52	21
Mario Lemieux	107	172	204.6	1.91	76	89.3	0.83	22
Claude Lemieux	234	158	203.8	0.87	80	100.7	0.43	23
Joe Pavelski	201	143	202.8	1.01	74	104.2	0.52	24
Patrick Kane	143	138	202	1.41	53	77.2	0.54	25
Mark Recchi	189	147	201.8	1.07	61	85.7	0.45	26
Mike Modano	176	146	200.8	1.14	58	77.9	0.44	27
Bryan Trottier	221	182	198.7	0.9	71	77.5	0.35	28
Joe Thornton	187	134	198.6	1.06	32	47.9	0.26	29

 

[MadLuke]

most adjusted career ppg (if you are in the top 250 most playoff points, group)

full_name	ppg
Connor McDavid	2.15
Leon Draisaitl	2
Wayne Gretzky	1.97
Mario Lemieux	1.91
Bobby Orr	1.79
Nathan MacKinnon	1.77
Peter Forsberg	1.67
Mikko Rantanen	1.67
Nikita Kucherov	1.61
Joe Sakic	1.6
Sidney Crosby	1.58
Gordie Howe	1.57
Jean Beliveau	1.54
Bobby Hull	1.5
Cale Makar	1.48
Phil Esposito	1.47
Evgeni Malkin	1.46
Ryan Getzlaf	1.42
Patrick Kane	1.41
Gilbert Perreault	1.41
Brayden Point	1.4
Mark Messier	1.35
Jake Guentzel	1.35
Alex Delvecchio	1.34
Mats Sundin	1.34
Stan Mikita	1.33
Gabriel Landeskog	1.33
Jaromir Jagr	1.32
Daniel Brière	1.32

 

[MadLuke]

most adjusted playoff goals:

full_name	games	goals	adjgoals	gpg
Wayne Gretzky	208	122	130.9	0.63
Brett Hull	202	103	129.1	0.64
Joe Sakic	172	84	122.4	0.71
Mark Messier	236	109	116.8	0.49
Jean Beliveau	162	79	111.2	0.69
Maurice Richard	132	82	111	0.84
Jari Kurri	200	106	109	0.55
Patrick Marleau	195	72	106.7	0.55
Gordie Howe	157	68	105.4	0.67
Jaromir Jagr	208	78	105.3	0.51
Joe Pavelski	201	74	104.2	0.52
Claude Lemieux	234	80	100.7	0.43
Alex Ovechkin	151	72	99.7	0.66
Sidney Crosby	180	71	99.4	0.55
Glenn Anderson	225	93	96.4	0.43
Evgeni Malkin	177	67	95.4	0.54
Peter Forsberg	151	64	93.7	0.62
Mario Lemieux	107	76	89.3	0.83
Mike Bossy	129	85	88.9	0.69
Steve Yzerman	196	70	88.4	0.45
Bobby Hull	119	62	86.1	0.72
Mark Recchi	189	61	85.7	0.45
Phil Esposito	130	61	85.6	0.66
Bernie Geoffrion	132	58	84.5	0.64
Esa Tikkanen	186	72	83.3	0.45
Yvan Cournoyer	147	64	83.2	0.57
Joe Nieuwendyk	158	66	83.1	0.53
Henrik Zetterberg	137	57	81.8	0.6
Brendan Shanahan	184	60	80.4	0.44

 

[MadLuke]

Most adjusted goal per games (if you are in the top 250 points group)

full_name	games	goals	adjgoals	gpg
Maurice Richard	132	82	111	0.84
Mario Lemieux	107	76	89.3	0.83
Jake Guentzel	69	38	53.7	0.78
Leon Draisaitl	74	41	55.4	0.75
Nathan MacKinnon	88	48	65	0.74
Bobby Hull	119	62	86.1	0.72
Joe Sakic	172	84	122.4	0.71
Jean Beliveau	162	79	111.2	0.69
Mike Bossy	129	85	88.9	0.69
Jarome Iginla	81	37	56.1	0.69
Cam Neely	93	57	62.9	0.68
Brayden Point	87	42	58.9	0.68
Gordie Howe	157	68	105.4	0.67
Connor McDavid	74	37	49.8	0.67
Alex Ovechkin	151	72	99.7	0.66
Phil Esposito	130	61	85.6	0.66
Brett Hull	202	103	129.1	0.64
Bernie Geoffrion	132	58	84.5	0.64
Wayne Gretzky	208	122	130.9	0.63
Peter Forsberg	151	64	93.7	0.62
Daniel Alfredsson	124	51	75.6	0.61
Daniel Brière	124	53	75.2	0.61
Martin St. Louis	107	42	64.9	0.61
Mats Sundin	91	38	55.5	0.61
Henrik Zetterberg	137	57	81.8	0.6
Logan Couture	116	48	69.1	0.6
Steve Shutt	99	50	59	0.6
Rod Gilbert	79	34	46.4	0.59
David Pastrnak	90	39	51.9	0.58

My feeling that Forsberg got underrated as a goal scorer based at how much he did it in the playoff for the era he was, match this.

Same for Sakic being underated with Jagr, because of who they followed, would they have had 1-2 best career for a long prime after anyone else than Mario and Wayne I feel like they would have been seen differently.

 

[authentic]

most adjusted career ppg (if you are in the top 250 most playoff points, group)

full_name	ppg
Connor McDavid	2.15
Leon Draisaitl	2
Wayne Gretzky	1.97
Mario Lemieux	1.91
Bobby Orr	1.79
Nathan MacKinnon	1.77
Peter Forsberg	1.67
Mikko Rantanen	1.67
Nikita Kucherov	1.61
Joe Sakic	1.6
Sidney Crosby	1.58
Gordie Howe	1.57
Jean Beliveau	1.54
Bobby Hull	1.5
Cale Makar	1.48
Phil Esposito	1.47
Evgeni Malkin	1.46
Ryan Getzlaf	1.42
Patrick Kane	1.41
Gilbert Perreault	1.41
Brayden Point	1.4
Mark Messier	1.35
Jake Guentzel	1.35
Alex Delvecchio	1.34
Mats Sundin	1.34
Stan Mikita	1.33
Gabriel Landeskog	1.33
Jaromir Jagr	1.32
Daniel Brière	1.32

That is amazing for McDavid. I wonder how that compares to Gretzky and Lemieux up to the same age or number of games played? Same with Jagr as he played a ton of games way past his prime.

 

[MadLuke]

McDavid scoring 42 points against an average 2.63 goals against per games in the 2024 playoffs is by a large amount the most ever.

McDavid 2024: 59 (25 games)
Lemieux 1991: 49 (23 games)
Kucherov 2020: 47 (25 games)
Gretzky 1985: 47 (in just 18 games...)

 

[VanIslander]

4 of the top 6 op stats pegs Gretzky. #1.

That flies.

 

[authentic]

McDavid scoring 42 points against an average 2.63 goals against per games in the 2024 playoffs is by a large amount the most ever.

McDavid 2024: 59 (25 games)
Lemieux 1991: 49 (23 games)
Kucherov 2020: 47 (25 games)
Gretzky 1985: 47 (in just 18 games...)

Per game still Gretzky by quite a margin

 

[daver]

This doesn't really pass the eye test.

I think the focus on "adjusting" things is taken too far. Anything that gets into the hypothetical is always questionable and adds unnecessary noise in discussions.

Is there some statistical anomaly that needs a deeper dive? Is there a player whose relative regular season dominance is way off of their relative playoff performances that can be explained by matchups?

Noone should be shocked that the best playoff run in points and/or PPGs in the DPE and the DPE 2.0 are significantly lower than timeframes before, in between, and after. But throughout all those timeframes, the names at the top of the best playoff runs are completely in line with the list of the best regular season performers (Mario, Jagr, Sakic, Forsberg, Crosby, Malkin, Kane, Kucherov, MacKinnon, Draisaitl, McDavid).

 

[authentic]

This doesn't really pass the eye test.

I think the focus on "adjusting" things is taken too far. Anything that gets into the hypothetical is always questionable and adds unnecessary noise in discussions.

Is there some statistical anomaly that needs a deeper dive? Is there a player whose relative regular season dominance is way off of their relative playoff performances that can be explained by matchups?

Noone should be shocked that the best playoff run in points and/or PPGs in the DPE and the DPE 2.0 are significantly lower than timeframes before, in between, and after. But throughout all those timeframes, the names at the top of the best playoff runs are completely in line with the list of the best regular season performers (Mario, Jagr, Sakic, Forsberg, Crosby, Malkin, Kane, Kucherov, MacKinnon, Draisaitl, McDavid).

I don’t know, look at McDavids gap over the rest of the top scorers like MacKinnon and Kucherov the past 3 playoffs.

 

[authentic]

4 of the top 6 op stats pegs Gretzky. #1.

That flies.

See if that remains once McDavid’s best runs are added, his 2023 playoffs probably ranks fairly high.

 

[Hockey Outsider]

Great data - thanks for sharing.

The top adjusted PPG table shows how "beneficial" it can be (in terms of elevating per-game results) for a player not to have a declining phase. Aside from Gretzky, eight of the next nine names are either currently active, or retired young. I find the next group (players with long careers and still very high per-game production - Sakic, Crosby, Howe and Beliveau) to be really impressive.

The other observation is Bobby Hull looks really good by this metric. That's consistent with some of the research that I've done. Hull was a great playoff performer, far better than his reputation suggests.