Empty Net & Overtime Scoring: Their Effects on Scoring Distribution and the Resulting Inflation of Point Totals for the League's Top Players

[TheStatican]

Empty net scoring has seen a huge uptick in recent seasons, largely thanks to analytics revealing that playing with an empty net is actually of a net benefit for a team aiming to make a comeback. Goals conceded during these situations are now considered to be largely inconsequential; after all, what does it matter if you lose by two or three goals instead of just one or two when it counts the same in the standings? However, these goals and points are still very consequential to individual players the biggest beneficiaries of which are the top skilled players on teams. What many don't realize is that scoring rates with an empty net are close to three times as high as powerplay scoring rates;

Secondly, overtime scoring also needs to be considered. The introduction of 4-on-4 overtime in 1999-2000, followed by 3-on-3 overtime in 2015-16, has led to a large uptick in OT goals. The increase isn't as dramatic as the ENG totals, but once again this is another setup that favors the most talented players.

Here are the numbers which verifies that scoring in these situations has dramatically increased;

Spoiler: Total Goals EN & OT

Season​	Total Games​	Total Goals​	PP Goals​	EN Against​	EN 6v5 for​	OT Goals​
1963-64​	210​	1166​	241​	13​	​	0​
1964-65​	210​	1208​	294​	13​	​	0​
1965-66​	210​	1277​	298​	11​	​	0​
1966-67​	210​	1253​	254​	10​	​	0​
1967-68​	444​	2476​	490​	37​	​	0​
1968-69​	456​	2718​	551​	31​	​	0​
1969-70​	456​	2649​	660​	29​	​	0​
1970-71​	546​	3409​	752​	45​	​	0​
1971-72​	546​	3348​	731​	57​	​	0​
1972-73​	624​	4088​	781​	55​	​	0​
1973-74​	624​	3989​	786​	51​	​	0​
1974-75​	720​	4932​	1157​	75​	​	0​
1975-76​	720​	4913​	1188​	68​	​	0​
1976-77​	720​	4783​	946​	62​	​	0​
1977-78​	720​	4747​	970​	67​	​	0​
1978-79​	680​	4757​	1045​	70​	​	0​
1979-80​	840​	5902​	1284​	111​	​	0​
1980-81​	840​	6457​	1608​	125​	​	0​
1981-82​	840​	6741​	1540​	111​	​	0​
1982-83​	840​	6493​	1493​	113​	​	0​
1983-84​	840​	6627​	1552​	125​	​	54*​
1984-85​	840​	6530​	1497​	102​	​	48​
1985-86​	840​	6667​	1716​	116​	​	56​
1986-87​	840​	6165​	1517​	121​	​	55​
1987-88​	840​	6237​	1861​	91​	​	49​
1988-89​	840​	6286​	1778​	132​	​	52​
1989-90​	840​	6189​	1599​	145​	​	55​
1990-91​	840​	5805​	1494​	109​	​	54​
1991-92​	880​	6123​	1700​	133​	​	52​
1992-93​	1008​	7311​	2081​	138​	​	65​
1993-94​	1092​	7081​	1975​	148​	​	74​
1994-95​	624​	3727​	964​	96​	​	26​
1995-96​	1066​	6701​	1927​	165​	​	64​
1996-97​	1066​	6216​	1422​	179​	​	70​
1997-98​	1066​	5624​	1491​	175​	​	54​
1998-99​	1107​	5830​	1533​	155​	​	60​
1999-00​	1148​	6306​	1496​	190​	117​	115**​
2000-01​	1230​	6781​	1877​	208​	100​	122​
2001-02​	1230​	6442​	1601​	201​	102​	121​
2002-03​	1230​	6530​	1787​	185​	104​	156​
2003-04​	1230​	6318​	1717​	189​	127​	145​
2005-06​	1230​	7442​	2545​	187​	149​	136​
2006-07​	1230​	7082​	2099​	205​	141​	117​
2007-08​	1230​	6691​	1871​	213​	112​	116​
2008-09​	1230​	7006​	1938​	226​	123​	123​
2009-10​	1230​	6803​	1664​	203​	102​	117​
2010-11​	1230​	6721​	1571​	227​	115​	148​
2011-12​	1230​	6545​	1408​	236​	98​	119​
2012-13​	720​	3822​	872​	138​	59​	65​
2013-14​	1230​	6573​	1441​	221​	117​	129​
2014-15​	1230​	6549​	1403​	284​	124​	136​
2015-16​	1230​	6565​	1429​	351​	151​	168***​
2016-17​	1230​	6704​	1404​	289​	118​	190​
2017-18​	1271​	7449​	1561​	349​	170​	193​
2018-19​	1271​	7577​	1466​	392​	153​	184​
2019-20​	1082​	6448​	1289​	327​	157​	164​
2020-21​	868​	5031​	992​	267​	134​	130​
2021-22​	1312​	8150​	1562​	474​	184​	186​
2022-23​	1312​	8248​	1662​	434​	177​	207​
2023-24​	1282​	7896​	1630​	439​	193​	187​

*Start of regular season overtime play(5 on 5)
** Start of 4 on 4 overtime
*** Start of 3 on 3 overtime
There may be some small errors in the data as I had to manually add up the EN totals from morehockeystats since there seems to be no website which provides the total yearly league-wide totals.

Spoiler: Per-game rates and % of all goals

Season​	G/Gm​	PPG/Gm​	EMG/Gm​	OTG/Gm​	EN+OT​	PPG %​	ENG % of all goals​	OTG % of all goals​
1963-64​	5.55​	1.15​	0.06​	0​	0.06​	20.7%​	1.1%​	0%​
1964-65​	5.75​	1.40​	0.06​	0​	0.06​	24.3%​	1.1%​	0%​
1965-66​	6.08​	1.42​	0.05​	0​	0.05​	23.3%​	0.9%​	0%​
1966-67​	5.97​	1.21​	0.05​	0​	0.05​	20.3%​	0.8%​	0%​
1967-68​	5.58​	1.10​	0.08​	0​	0.08​	19.8%​	1.5%​	0%​
1968-69​	5.96​	1.21​	0.07​	0​	0.07​	20.3%​	1.1%​	0%​
1969-70​	5.81​	1.45​	0.06​	0​	0.06​	24.9%​	1.1%​	0%​
1970-71​	6.24​	1.38​	0.08​	0​	0.08​	22.1%​	1.3%​	0%​
1971-72​	6.13​	1.34​	0.10​	0​	0.10​	21.8%​	1.7%​	0%​
1972-73​	6.55​	1.25​	0.09​	0​	0.09​	19.1%​	1.3%​	0%​
1973-74​	6.39​	1.26​	0.08​	0​	0.08​	19.7%​	1.3%​	0%​
1974-75​	6.85​	1.61​	0.10​	0​	0.10​	23.5%​	1.5%​	0%​
1975-76​	6.82​	1.65​	0.09​	0​	0.09​	24.2%​	1.4%​	0%​
1976-77​	6.64​	1.31​	0.09​	0​	0.09​	19.8%​	1.3%​	0%​
1977-78​	6.59​	1.35​	0.09​	0​	0.09​	20.4%​	1.4%​	0%​
1978-79​	7.00​	1.54​	0.10​	0​	0.10​	22.0%​	1.5%​	0%​
1979-80​	7.03​	1.53​	0.13​	0​	0.13​	21.8%​	1.9%​	0%​
1980-81​	7.69​	1.91​	0.15​	0​	0.15​	24.9%​	1.9%​	0.​
1981-82​	8.03​	1.83​	0.13​	0​	0.13​	22.8%​	1.6%​	0%​
1982-83​	7.73​	1.78​	0.13​	0​	0.13​	23.0%​	1.7%​	0%​
1983-84​	7.89​	1.85​	0.15​	0.06​	0.21​	23.4%​	1.9%​	0.8%​
1984-85​	7.77​	1.78​	0.12​	0.06​	0.18​	22.9%​	1.6%​	0.7%​
1985-86​	7.94​	2.04​	0.14​	0.07​	0.20​	25.7%​	1.7%​	0.8%​
1986-87​	7.34​	1.81​	0.14​	0.07​	0.21​	24.6%​	2.0%​	0.9%​
1987-88​	7.43​	2.22​	0.11​	0.06​	0.17​	29.8%​	1.5%​	0.8%​
1988-89​	7.48​	2.12​	0.16​	0.06​	0.22​	28.3%​	2.1%​	0.8%​
1989-90​	7.37​	1.90​	0.17​	0.07​	0.24​	25.8%​	2.3%​	0.9%​
1990-91​	6.91​	1.78​	0.13​	0.06​	0.19​	25.7%​	1.9%​	0.9%​
1991-92​	6.96​	1.93​	0.15​	0.06​	0.21​	27.8%​	2.2%​	0.8%​
1992-93​	7.25​	2.06​	0.14​	0.06​	0.20​	28.5%​	1.9%​	0.9%​
1993-94​	6.48​	1.81​	0.14​	0.07​	0.20​	27.9%​	2.1%​	1.0%​
1994-95​	5.97​	1.54​	0.15​	0.04​	0.20​	25.9%​	2.6%​	0.7%​
1995-96​	6.29​	1.81​	0.15​	0.06​	0.21​	28.8%​	2.5%​	1.0%​
1996-97​	5.83​	1.33​	0.17​	0.07​	0.23​	22.9%​	2.9%​	1.1%​
1997-98​	5.28​	1.40​	0.16​	0.05​	0.21​	26.5%​	3.1%​	1.0%​
1998-99​	5.27​	1.38​	0.14​	0.05​	0.19​	26.3%​	2.7%​	1.0%​
1999-00​	5.49​	1.30​	0.17​	0.10​	0.27​	23.7%​	3.0%​	1.8%​
2000-01​	5.51​	1.53​	0.17​	0.10​	0.27​	27.7%​	3.1%​	1.8%​
2001-02​	5.24​	1.30​	0.16​	0.10​	0.26​	24.9%​	3.1%​	1.9%​
2002-03​	5.31​	1.45​	0.15​	0.13​	0.28​	27.4%​	2.8%​	2.4%​
2003-04​	5.14​	1.40​	0.15​	0.12​	0.27​	27.2%​	3.0%​	2.3%​
2005-06​	6.05​	2.07​	0.15​	0.11​	0.26​	34.2%​	2.5%​	1.8%​
2006-07​	5.76​	1.71​	0.17​	0.10​	0.26​	29.6%​	2.9%​	1.7%​
2007-08​	5.44​	1.52​	0.17​	0.09​	0.27​	28.0%​	3.2%​	1.7%​
2008-09​	5.70​	1.58​	0.18​	0.10​	0.28​	27.7%​	3.2%​	1.8%​
2009-10​	5.53​	1.35​	0.17​	0.10​	0.26​	24.5%​	3.0%​	1.7%​
2010-11​	5.46​	1.28​	0.18​	0.12​	0.30​	23.4%​	3.4%​	2.2%​
2011-12​	5.32​	1.14​	0.19​	0.10​	0.29​	21.5%​	3.6%​	1.8%​
2012-13​	5.31​	1.21​	0.19​	0.09​	0.28​	22.8%​	3.6%​	1.7%​
2013-14​	5.34​	1.17​	0.18​	0.10​	0.28​	21.9%​	3.4%​	2.0%​
2014-15​	5.32​	1.14​	0.23​	0.11​	0.34​	21.4%​	4.3%​	2.1%​
2015-16​	5.34​	1.16​	0.29​	0.14​	0.42​	21.8%​	5.3%​	2.6%​
2016-17​	5.45​	1.14​	0.23​	0.15​	0.39​	20.9%​	4.3%​	2.8%​
2017-18​	5.86​	1.23​	0.27​	0.15​	0.43​	21.0%​	4.7%​	2.6%​
2018-19​	5.96​	1.15​	0.31​	0.14​	0.45​	19.3%​	5.2%​	2.4%​
2019-20​	5.96​	1.19​	0.30​	0.15​	0.45​	20.0%​	5.1%​	2.5%​
2020-21​	5.80​	1.14​	0.31​	0.15​	0.46​	19.7%​	5.3%​	2.6%​
2021-22​	6.21​	1.19​	0.36​	0.14​	0.50​	19.2%​	5.8%​	2.3%​
2022-23​	6.29​	1.27​	0.33​	0.16​	0.49​	20.2%​	5.3%​	2.5%​
2023-24​	6.16​	1.27​	0.34​	0.15​	0.49​	20.6%​	5.5%​	2.3%​

A summarization of the above data;
Empty Net Goals

Seasons​	Situation​	EN Goals​	Games​	Total Goals​	ENG/Gm​	ENG % of all Goals​
1963 to 1967​	O6 sample​	47​	840​	4904​	0.056​	1.0%​
1967 to 1979​	Post Expansion​	647​	7256​	46809​	0.09​	1.4%​
1979 to 1994​	WHA Expansion​	1820​	13060​	96614​	0.14​	1.9%​
1994 to 2014​	Post lockout​	3883​	21557​	119160​	0.18​	3.3%​
2014 to 2019​	Post 2014​	1273​	4961​	27267​	0.26​	4.7%​
2019 till now​	Post 2019​	2322​	7099​	43283​	0.33​	5.4%​

Before the initial expansion, empty net goals accounted for a small percentage of overall scoring, roughly a dozen goals annually, or about 1% of total scoring. Following the O6 expansion, there was a slight but noticeable increase, bringing it up to about 1.5%. Subsequently, after the NHL's second major expansion in 1979, there was another uptick, raising the total to about 2% of all scoring. These figures remained relatively consistent until the first lockout in 1994, after which for some reason the percentage of empty net goals surged into the 2.5-3% range. This trend continued to slowly escalate until 2014-15, when the numbers significantly jumped again to well over 4%, since then then numbers have continued to climb into the 5-6% range, reaching an all-time high of 5.8% in 2021-2022.

Overtime Goals

Years​	Situation​	OT Goals​	Games​	Total Goals​	OT Goals/Gm​	OTG % of all Goals​
1942 to 1983​	No overtime​	-​	-​	-​	-​	-​
1983 to 1999​	5 on 5 OT​	888​	14629​	99119​	0.06​	0.9%​
1999 to 2015​	4 on 4 OT​	1865​	17858​	97611​	0.10​	1.9%​
2015 till now​	3 on 3 OT​	1605​	10830​	64001​	0.15​	2.5%​

During the period from 1943 to 1983, the NHL did not incorporate regular season overtime. However, starting in 1983, the NHL reinstated regular season overtime. The OT period along resulted in approximately a 1% increase in scoring compared to the pre-overtime era. Adjusted scoring figures likely already account for this increase, at least to some extent... It's possible they do not fully capture its impact since overtime is sudden death and teams typically deploy their top lines to start the period. That means if they successfully score and end the game this would slightly skew the ice time distribution and point allocation to their favor. Further analysis is required to precisely determine the extent to which 5-on-5 overtime scoring benefited top players during this period, though I surmise it's overall impact would have been quite minimal.

In 1999, the NHL transitioned to 4-on-4 overtime, effectively doubling the overtime scoring benefit. Then, in 2015, the NHL adopted 3-on-3 overtime, further increasing the advantage. Currently, approximately 2.5% of all scoring occurs in overtime, with this scoring benefit favoring the top players, who often receive a larger share of available ice time during OT periods.

Empty net and overtime goals now collectively account for over 600 goals per year, constituting 8% of all scoring. The following chart clearly profiles their dramatic rise over time;

Per game goals in the highest scoring situations;

It's worth noting that while 4-on-4 overtime benefits higher skilled players, it wouldn't have provided the same degree of benefit as 3-on-3 overtime currently does, hence why I've differentiated it in the chart above.

6v5 Goals
Additionally, goals scored while playing with an empty net (6-on-5) are another overlooked aspect affecting point distribution. Scoring rates per 60 minutes in this situation are comparable to power play scoring rates(somewhat lower in recent seasons due to higher PP scoring rates). These goals further skew totals toward higher-skilled players, who are often deployed in comeback situations. Notably, Auston Matthews leads the league in both goals and points in this senario. Unfortunately, without available data before 1999-2000, it's challenging to ascertain the frequency and increased prominence of these goals. Nonetheless, the per-game rates for them have increased by about 50% since 1999-2000 based on available data;

Spoiler: 6v5 totals

Season​	6v5 G’s​	6v5/Gm​	6v5 %​
1999-00​	117​	0.10​	1.9%​
2000-01​	100​	0.08​	1.5%​
2001-02​	102​	0.08​	1.6%​
2002-03​	104​	0.08​	1.6%​
2003-04​	127​	0.10​	2.0%​
2005-06​	149​	0.12​	2.0%​
2006-07​	141​	0.11​	2.0%​
2007-08​	112​	0.09​	1.7%​
2008-09​	123​	0.10​	1.8%​
2009-10​	102​	0.08​	1.5%​
2010-11​	115​	0.09​	1.7%​
2011-12​	98​	0.08​	1.5%​
2012-13​	59​	0.08​	1.5%​
2013-14​	117​	0.10​	1.8%​
2014-15​	124​	0.10​	1.9%​
2015-16​	151​	0.12​	2.3%​
2016-17​	118​	0.10​	1.8%​
2017-18​	170​	0.13​	2.3%​
2018-19​	153​	0.12​	2.0%​
2019-20​	157​	0.15​	2.4%​
2020-21​	134​	0.15​	2.7%​
2021-22​	184​	0.14​	2.3%​
2022-23​	177​	0.13​	2.1%​
2023-24​	193​	0.15​	2.4%​

Finally an example of the cumulative totals of the top 10 point scorers from certain selection seasons and the portion of their totals that were Empty Net and Overtime Points;

Season​	Pts​	EN(against) Pts​	OT Pts​	Total​	% of points​
1981-82​	1325​	25​	n/a​	25​	1.9%​
1992-93​	1353​	25​	(12)​	25​	1.8%​
1995-96​	1210​	30​	(10)​	30​	2.5%​
2005-06​	1050​	15​	25​	40​	3.8%​
2023-24​	1136​	62​	36​	98​	8.6%​

There was no OT in 1981-82. OT in 92-93 & 05-96 was 5 on 5 which would have provided only a very limited scoring benefit, if any to these players, hence why the totals are listing in brackets and not added up in the totals column. 05-06 had 4 on 4 OT which has notably higher scoring rates than normal 5 on 5 play and of course this season features 3 on 3 OT which provides the highest scoring benefit - equivalent to being on a powerplay.

Now for the juicy bits, examples of player scoring rates in these situations. First, here's a list of the best player seasons of the 4 on 4 OT era. Unfortunately naturalstattrick doesn't have the TOI and Pt's for 4 on 4 play, but they do have the empty net for(6v5) and against numbers. The total overtime point totals these players had was 18, and most of these would have been 4 on 4.

Player​

Season​

Pts​

​

4v4​

6v5​

EN​

​

4v4 TOI​

6v5 TOI​

EN TOI​

​

ES/60​

PP/60​

4v4/60​

6v5/60​

EN/60​

Crosby​

10-11​

66​

​

​

2​

2​

​

​

14:52​

10:10​

​

4.2​

5.1​

​

8.1​

11.8​

Crosby​

13-14​

104​

​

​

3​

5​

​

​

21:40​

17:37​

​

2.9​

6.6​

​

8.3​

17.0​

Malkin​

08-09​

113​

​

​

2​

6​

​

​

21:28​

14:42​

​

3.2​

5.4​

​

5.6​

24.5​

Malkin​

11-12​

109​

​

​

4​

3​

​

​

24:39​

06:42​

​

3.6​

6.2​

​

9.7​

26.9​

Ovechkin​

07-08​

112​

​

​

4​

4​

​

​

27:24​

16:19​

​

3.2​

4.8​

​

8.8​

14.7​

Ovechkin​

08-09​

110​

​

​

2​

8​

​

​

15:57​

18:43​

​

2.9​

6.5​

​

7.5​

25.6​

Ovechkin​

09-10​

109​

​

​

3​

5​

​

​

18:25​

12:06​

​

3.6​

5.9​

​

9.8​

24.8​

Sedin​

09-10​

112​

​

​

0​

4​

​

​

22:45​

11:48​

​

3.1​

8.5​

​

0.0​

20.3​

Totals​

​

835​

​

18*​

20​

37​

​

​

167:10​

108:07​

​

3.3​

6.1​

​

7.2​

20.5​

% PerGm​

​

​

​

2.15​

2.4%​

4.4%​

​

​

00:16​

00:10​

​

​

​

​

​

​

These players averaged only 10 seconds a game paying against an empty net, which is less than 1% of their total ice time, and yet it yielded them 4.4% of their point totals. Their cumulative scoring rate playing against an empty net was 20.5/per 60 more than 6 times higher than their even strength scoring rate and 3 times higher than their powerplay rate. Their scoring rate playing with an empty net (6v5) was also higher than their powerplay scoring rate. Thanks in part to averaging a little more time in these comeback situations these points yielded them 2.4% of their totals. Overtime points yielded them 2.15% of their totals, these would have been mostly 4 on 4 with some 5 on 4 time.

Now lets look at the modern era, these are all the top performing seasons since 3 on 3 overtime was incorporated, numbers for which we do have on naturalstattrick:

Player​

Season​

Pts​

​

3v3​

6v5​

EN​

​

3v3 TOI​

6v5 TOI​

EN TOI​

​

ES/60​

PP/60​

3v3/60​

6v5/60​

EN/60​

Kane​

15-16​

106​

​

4​

2​

5​

​

21:55​

41:49​

12:57​

​

2.9​

8.7​

11.0​

2.9​

23.2​

Kane​

18-19​

110​

​

2​

5​

7​

​

28:40​

35:31​

14:37​

​

3.2​

6.0​

4.2​

8.4​

28.7​

McDavid​

18-19​

116​

​

6​

6​

4​

​

25:04​

36:04​

13:12​

​

3.3​

7.4​

14.4​

10.0​

18.2​

McDavid​

20-21​

105​

​

3​

3​

7​

​

06:43​

21:55​

21:22​

​

4.1​

9.5​

26.8​

8.2​

19.7​

McDavid​

21-22​

123​

​

5​

2​

11​

​

31:40​

32:55​

25:24​

​

3.2​

8.9​

9.5​

3.6​

26.0​

McDavid​

22-23​

153​

​

3​

5​

7​

​

23:07​

29:31​

28:34​

​

3.2​

13.2​

7.8​

10.2​

14.7​

McDavid​

23-24​

130​

​

2​

4​

5​

​

19:13​

35:52​

20:34​

​

3.9​

10.3​

6.2​

6.7​

14.6​

Draisaitl​

19-20​

110​

​

3​

2​

6​

​

28:55​

26:49​

25:45​

​

3.1​

9.7​

6.2​

4.5​

14.0​

Draisaitl​

22-23​

128​

​

3​

3​

7​

​

23:28​

25:28​

28:16​

​

2.8​

11.7​

7.7​

7.1​

14.9​

Matthews​

21-22​

106​

​

3​

2​

5​

​

23:01​

30:06​

16:15​

​

3.6​

7.7​

7.8​

4.0​

18.5​

Matthews​

23-24​

104​

​

4​

8​

2​

​

36:06​

46:10​

19:22​

​

3.5​

6.6​

6.6​

10.4​

6.2​

Kucherov​

18-19​

128​

​

2​

1​

6​

​

22:13​

30:13​

19:48​

​

3.6​

9.4​

5.4​

2.0​

18.2​

Kucherov​

22-23​

113​

​

0​

7​

6​

​

16:16​

37:46​

15:12​

​

2.9​

9.1​

0.0​

11.1​

23.7​

Kucherov​

23-24​

139​

​

2​

6​

14​

​

15:40​

53:39​

24:52​

​

3.8​

9.8​

7.7​

6.7​

33.8​

MacKinnon​

22-23​

111​

​

3​

1​

4​

​

25:13​

33:06​

17:36​

​

3.6​

7.0​

7.1​

1.8​

13.6​

MacKinnon​

23-24​

137​

​

3​

1​

8​

​

23:37​

33:37​

20:28​

​

3.7​

7.8​

7.6​

1.8​

23.5​

Huberdeau​

21-22​

115​

​

8​

5​

8​

​

26:33​

39:49​

16:16​

​

3.9​

7.7​

18.1​

7.5​

29.5​

Gaudreau​

21-22​

115​

​

2​

2​

11​

​

18:37​

18:37​

11:31​

​

4.2​

6.2​

6.4​

6.4​

57.3​

Pastrňák​

22-23​

113​

​

1​

3​

6​

​

16:31​

20:52​

33:34​

​

3.5​

7.1​

3.6​

8.6​

10.7​

Panarin​

23-24​

116​

​

4​

6​

7​

​

19:16​

50:10​

15:55​

​

3.3​

9.1​

12.5​

7.2​

26.5​

Totals​

​

2162​

​

57​

67​

124​

​

401:13​

602:39​

374:56​

​

3.5​

8.6​

8.5​

6.7​

19.8​

% PerGm​

​

​

​

2.6%​

3.1%​

5.7%​

​

00:17​

00:26​

00:17​

​

ES​

PP​

3v3​

6v5​

EN​

Several observations stand out prominently. Firstly, it's evident that the increasing number of goals scored in these situations is indeed due to an increase in their occurrences. "EN against" TOI increased from 10 seconds to 17 seconds and "EN for" increased from 16 seconds to 26 per game. The scoring rates when facing an empty net, both for and against, were marginally lower compared to the previous sample group. It's unclear whether this indicates a genuine decline or if the initial sample size is insufficient for a reliable comparison. Nevertheless, scoring rates in situations involving an empty net remain significantly higher than in any other game-time scenario. In fact, these instances accounted for 5.7% of total goals scored, slightly exceeding the league's overall average for the 2023-2024 season, which stood at 5.5%.

Playing with an empty net contributed 3.1% of total points, significantly surpassing the league average of 2.2%. This suggests that top players are extensively utilized during comeback attempts. Additionally, 3-on-3 overtime accounted for another 2.6% of total goals. Remarkably, the scoring rates in 3-on-3 overtime are comparable to those on powerplays, underscoring the benefit of playing during this time. This is particularly noteworthy given that powerplay scoring rates have increased notably since the 2008-2014 period. All combined these situations are responsible for 11.4% of the top 10 players point totals. In comparison, power play goals currently account approximately 20% of all goals, while shorthanded goals comprise a little under 3%.

In summarization, empty net situations and changes in overtime rules have clearly influenced the distribution of points in recent seasons, underscoring the evolving nature of NHL scoring dynamics. It is evident that these overlooked aspects disproportionately favor the league's higher-skilled players and that future scoring adjustment metrics should incorporate these insights in order to maintain accuracy.

Click to read more...

 

[BraveCanadian]

Really glad to see the distribution of scoring getting some work done.

It is such a huge factor in comparing seasons by players over time.

 

[The Panther]

This isn't relevant to your larger point, but since we're on the history board, a couple things to note:

-- It's "WHA", not WHL
-- 1979 was not the NHL's second major expansion. The 1979 expansion grew the League from 17 to 21 teams (+23.5%), but before that, between 1967-68 and 1978-79, the League had grown from 12 to 18 teams (+50%).

Empty-net scoring is kind of 'meh' for me. It's obvious that pro-hockey strategy has shifted to (a) pulling goalies earlier, and (b) encouraging defenders to shoot for the empty net more often (i.e., to risk icing). So, empty net scoring goes up.

As to why there were incremental increases prior to the current era, I don't know.

 

[Hockey Outsider]

This is great analysis. I was planning to do something like this eventually, but haven't had time.

In summarization, empty net situations and changes in overtime rules have clearly influenced the distribution of points in recent seasons, underscoring the evolving nature of NHL scoring dynamics. It is evident that these overlooked aspects disproportionately favor the league's higher-skilled players and that future scoring adjustment metrics should incorporate these insights in order to maintain accuracy.

This is yet another reason why simply adjusting players' stats by the league average goals per game is insufficient.

 

[overpass]

Really good stuff.

I was thinking about these different types of goals recently as a reason why plus-minus as calculated by the NHL for the last 60-plus years has become less and less useful. In 1960, there were very few shorthanded goals, empty net goals, and no 3-on-3 OT, so using total goals minus power play goals worked pretty well for the game as it was played then. But shorthanded goals became much more common after the 1967 expansion, and then there have been these other trends described in the OP.

It makes sense that these different goal types would disproportionately go to the top scorers as well. Great insight.

 

[TheStatican]

Really good stuff.

I was thinking about these different types of goals recently as a reason why plus-minus as calculated by the NHL for the last 60-plus years has become less and less useful. In 1960, there were very few shorthanded goals, empty net goals, and no 3-on-3 OT, so using total goals minus power play goals worked pretty well for the game as it was played then. But shorthanded goals became much more common after the 1967 expansion, and then there have been these other trends described in the OP.

It makes sense that these different goal types would disproportionately go to the top scorers as well. Great insight.

Thank you. Indeed, plus minus was of some value in the past but it's utility is impractical now with all these unique game time situations skewing the data.

-- It's "WHA", not WHL

Fixed.

-- 1979 was not the NHL's second major expansion. The 1979 expansion grew the League from 17 to 21 teams (+23.5%), but before that, between 1967-68 and 1978-79, the League had grown from 12 to 18 teams (+50%).

I never said that it was as significant as the O6 expansion but it remains the second largest expansion in league history, personally I would still categorize that as a major expansion.

 

[TheStatican]

It would be complicated and time consuming to make scoring adjustments for EN & OT points on top of doing so for ES/PP&SH. A more simple means of comparison might be to just remove these points all-together and then do a scoring adjustment. When doing so here are the era single season point leaders;

Original 6 expansion
No OT, very limited EN opportunities

Pts​	Player​	Season​	EN​	OT​
150​	Esposito​	1970-71​	2​	-​
143​	Esposito​	1973-74​	2​	-​
137​	Orr​	1970-71​	2​	-​
136​	Lafleur​	1976-77​	0​	-​
134​	Trottier​	1978-79​	0​	-​
133​	Orr​	1974-75​	2​	-​
132​	Lafleur​	1977-78​	0​	-​
128​	Esposito​	1971-72​	5​	-​
128​	Esposito​	1972-73​	2​	-​
127​	Lafleur​	1978-79​	2​	-​
126​	Esposito​	1974-75​	1​	-​
126​	Bossy​	1978-79​	0​	-​
125​	Esposito​	1968-69​	1​	-​
125​	Lafleur​	1975-76​	0​	-​
125​	Dionne​	1978-79​	5​	-​
121​	Dionne​	1974-75​	0​	-​
120​	Orr​	1973-74​	2​	-​
120​	Trottier​	1977-78​	3​	-​
120​	Dionne​	1976-77​	2​	-​
118​	Orr​	1969-70​	2​	-​
​	Average​	​	1.7​	​

WHA expansion
No OT, limited but increasing EN opportunities

​	Player​	Season​	EN​	OT​
207​	Gretzky​	1981-82​	5​	-​
192​	Gretzky​	1982-83​	4​	-​
157​	Gretzky​	1980-81​	7​	-​
140​	Bossy​	1981-82​	7​	-​
138​	Šťastný​	1981-82​	1​	-​
134​	Gretzky​	1979-80​	3​	-​
134​	Maruk​	1981-82​	2​	-​
132​	Dionne​	1979-80​	5​	-​
132​	Dionne​	1980-81​	3​	-​
131​	Nilsson​	1980-81​	0​	-​
126​	Trottier​	1981-82​	3​	-​
125​	Lafleur​	1979-80​	0​	-​
124​	Šťastný​	1982-83​	0​	-​
121​	Savard​	1982-83​	0​	-​
​	Average​	​	2.9​	​

Resumption of Overtime
Five minutes of 5 on 5 OT, increasing EN opportunities

​	Player​	Season​	EN​	OT​
206​	Gretzky​	1985-86​	6​	3​
205​	Gretzky​	1984-85​	3​	0​
196​	Gretzky​	1983-84​	7​	2​
190​	Lemieux​	1988-89​	7​	2​
172​	Gretzky​	1986-87​	9​	2​
160​	Lemieux​	1987-88​	4​	4​
158​	Gretzky​	1988-89​	8​	2​
149​	Yzerman​	1988-89​	4​	2​
147​	Gretzky​	1987-88​	2​	0​
141​	Nicholls​	1988-89​	7​	2​
138​	Lemieux​	1985-86​	0​	3​
136​	Coffey​	1985-86​	1​	1​
132​	Kurri​	1984-85​	3​	0​
126​	Kurri​	1985-86​	3​	2​
126​	Hawerchuk​	1984-85​	1​	3​
123​	Dionne​	1984-85​	1​	2​
123​	Savard​	1987-88​	6​	2​
122​	Goulet​	1983-84​	0​	0​
122​	Coffey​	1983-84​	3​	1​
119​	Coffey​	1984-85​	2​	0​
118​	Šťastný​	1985-86​	3​	1​
117​	Hawerchuk​	1987-88​	0​	4​
116​	Bossy​	1985-86​	5​	2​
​	Average​	​	3.7​	1.7​

Early 90's
Same as above but with lower scoring levels

​	Player​	Season​	EN​	OT​
156​	Lemieux​	1995-96​	3​	2​
155​	Gretzky​	1990-91​	6​	2​
154​	Lemieux​	1992-93​	4​	2​
145​	Jágr​	1995-96​	3​	1​
144​	LaFontaine​	1992-93​	3​	1​
138​	Gretzky​	1989-90​	3​	1​
136​	Oates​	1992-93​	3​	3​
135​	Yzerman​	1992-93​	2​	0​
129​	Brett Hull*​	1990-91​	1​	1​
128​	Selänne​	1992-93​	3​	1​
128​	Turgeon​	1992-93​	1​	3​
127​	Lemieux​	1991-92​	3​	1​
125​	Gretzky​	1993-94​	4​	1​
125​	Gilmour​	1992-93​	2​	0​
124​	Messier​	1989-90​	2​	3​
123​	Mogilny​	1992-93​	3​	1​
123​	Robitaille​	1992-93​	1​	1​
121​	Yzerman​	1989-90​	4​	2​
121​	Stevens​	1991-92​	1​	1​
118​	Lemieux​	1989-90​	4​	1​
118​	Sakic​	1995-96​	1​	1​
115​	Francis​	1995-96​	4​	0​
114​	Federov​	1993-94​	2​	4​
113​	Forsberg​	1995-96​	2​	1​
112​	Oates​	1990-91​	1​	2​
109​	Lindros​	1995-96​	6​	0​
​	Average​	​	2.8​	1.4​

Lower scoring DPE & Post lockout
50% increase in EN scoring from 1994 to 1997, start of 4 on 4 OT in 1999

​	Player​	Season​	EN​	OT​
121​	Jágr​	1998-99​	1​	5**​
117​	Thornton​	2005-06*​	1​	7​
118​	Crosby​	2006-07​	1​	1​
116​	Jágr​	2005-06*​	3​	4​
113​	Lemieux​	1996-97​	9​	0​
113​	Sakic​	2000-01​	2​	3​
112​	Thornton​	2006-07​	2​	0​
111​	Jágr​	2000-01​	10​	0​
106​	Sedin​	2009-10​	4​	2​
105​	Malkin​	2011-12​	3​	1​
104​	Malkin​	2008-09​	6​	3​
104​	Ovechkin​	2007-08​	4​	4​
100​	Forsberg​	2002-03​	2​	4​
102​	Ovechkin​	2009-10​	5​	2​
100​	Ovechkin​	2008-09​	8​	2​
97​	Crosby​	2013-14​	5​	2​
62​	Crosby​	2010-11​	2​	2​
​	Average​	​	4.0​	2.5​

* high scoring season
** done in 5 on 5 OT

Current Era
Continued dramatic increase in EN opportunities, start of 3 on 3 overtime in 2015

​	Player​	Season​	EN​	OT​	6v5*​
142 (137)​	McDavid​	2022-23​	7​	4​	5​
125 (119)​	Kucherov​	2023-24​	14​	2​	6​
124 (123)​	MacKinnon​	2023-24​	8​	6​	1​
121 (117)​	McDavid​	2023-24​	5​	4​	4​
119 (118)​	Kucherov​	2018-19​	6​	3​	1​
117 (114)​	Draisaitl​	2022-23​	7​	4​	3​
107 (101)​	Panarin​	2023-24​	7​	4​	6​
107 (105)​	McDavid​	2021-22​	11​	5​	2​
106 (100)​	McDavid​	2018-19​	4​	6​	6​
106 (99)​	Kucherov​	2022-23​	6​	1​	7​
103 (102)​	MacKinnon​	2022-23​	4​	4​	1​
102 (99)​	Pastrňák​	2022-23​	6​	5​	3​
101 (99)​	Draisaitl​	2019-20​	6​	3​	2​
101 (93)​	Matthews​	2023-24​	2​	4​	8​
100 (98)​	Gaudreau​	2021-22​	11​	4​	2​
100 (95)​	Kane​	2018-19​	7​	3​	5​
99 (94)​	Huberdeau​	2021-22​	8​	8​	5​
97 (95)​	Matthews​	2021-22​	5​	4​	2​
95 (92)​	McDavid​	2020-21​	7​	3​	3​
93 (85)​	Kane​	2015-16​	5​	6​	8​
​	Average​	​	6.8​	4.2​	4.0​

*6v5 points(playing with an empty net) have not been removed from the first PT totals but are shown in brackets. They're another added benefit these players are receiving in a greater amount verse past seasons. But without knowing the numbers for seasons prior to 1999 one can't do a proper comparison, they seem to be be in the range of about 60% of the EN totals. Here's the average points for the above time periods(actual totals in bold, estimates in italics);

​	EN​	OT​	6v5​	Total​
1967 to 1979​	1.7​	0​	1.0​	2.6​
1979 to 1983​	2.9​	0​	1.7​	4.5​
1983 to 1989​	3.7​	1.7​	2.2​	7.6​
1989 to 1996​	2.8​	1.4​	1.6​	5.8​
1996 to 2015​	4.0​	2.5​	2.4​	8.8​
2015-current​	6.8​	4.2​	4.0​	15.0 ​

 

[Leksand]

This is such an interesting and illuminating analysis. I never understood how it can be controversial to somehow adjust scoring statistics for the scoring environment if you want to compare numbers across time - but somehow it is. Adjustments are needed. At the same time, it seems very difficult to do in a straightforward broadly acceptable way. I really haven't seen anything like that (and if there is something I would very interested to read). And even if one thinks it's possible, the lack of data will be very limiting for any longer perspective analysis.

It's really why I more and more appreciate ranking stats - who scored the most goals, had most points, had most assists in a season / several seasons / many seasons over a career. All players play with the same rules during a season, play against the same players, shoot against goalies with the same size pads, face the same tactics, same reffing and so on. In principle I therefore also value Hart, Norris, All Star wins / voting quite a bit, although learning from the many insightful HOH posters over the years, that requires caution.

 

[The Panther]

At the very least, the NHL should stop awarding 'pluses' and 'minuses' for empty-net goals. This would require zero effort on the League's part and make more sense. It's completely illogical that PP goals don't count for plus/minus, but EN goals do.

 

[daver]

This is such an interesting and illuminating analysis. I never understood how it can be controversial to somehow adjust scoring statistics for the scoring environment if you want to compare numbers across time - but somehow it is. Adjustments are needed. At the same time, it seems very difficult to do in a straightforward broadly acceptable way. I really haven't seen anything like that (and if there is something I would very interested to read). And even if one thinks it's possible, the lack of data will be very limiting for any longer perspective analysis.

Because it doesn't account for dynamics like increases and decreases in PP opportunities and the spreading out of scoring throughout the team vs. 1st line/Top 6.

If the focus is a comparison of high end scorers, their performance vs. a reasonable sample size of their peers; players that are also getting similar PP time, OT time and perhaps EN time is the best way to compare with other stars from other seasons/eras.

The OP is interesting but not sure it sheds light on too much when talking about the elite players.

 

[Leksand]

Because it doesn't account for dynamics like increases and decreases in PP opportunities and the spreading out of scoring throughout the team vs. 1st line/Top 6.

If the focus is a comparison of high end scorers, their performance vs. a reasonable sample size of their peers; players that are also getting similar PP time, OT time and perhaps EN time is the best way to compare with other stars from other seasons/eras.

The OP is interesting but not sure it sheds light on too much when talking about the elite players.

I literally don’t understand this reservation against adjusted stats. It seems to be very strongly arguing for adjusted stats.

Concretely: right now there’s a lot of discussion about goal scoring (or assists) in one season, for obvious reasons. Just looking at straight up unadjusted numbers for goals (or assists) during one season, ie Gretzky, Hull and Mario win by a mile against anyone in the last 25 years or so, does that really account for dynamics like increases and decreases in PP opportunities and the spreading out of scoring throughout the team vs. 1st line/Top 6? I just don’t see how it does.

 

[TheStatican]

The OP is interesting but not sure it sheds light on too much when talking about the elite players.

Indeed, the stats I've provided are league-wide with only a relatively small sample size of select individual player-seasons. It gives us an idea but not doesn't clearly establish precisely how much elite players as a whole are benefiting, but the indications seem to be that their advantage is likely even greater than that of the league as a whole.

 

[daver]

I literally don’t understand this reservation against adjusted stats. It seems to be very strongly arguing for adjusted stats.

Concretely: right now there’s a lot of discussion about goal scoring (or assists) in one season, for obvious reasons. Just looking at straight up unadjusted numbers for goals (or assists) during one season, ie Gretzky, Hull and Mario win by a mile against anyone in the last 25 years or so, does that really account for dynamics like increases and decreases in PP opportunities and the spreading out of scoring throughout the team vs. 1st line/Top 6? I just don’t see how it does.

You can compare players by comparing their dominance relative to their peers.

 

[BraveCanadian]

I literally don’t understand this reservation against adjusted stats. It seems to be very strongly arguing for adjusted stats.

Comparing stats adjusted by league average scoring completely ignore the distribution of ES/PP/SH/EN offense and more importantly they completely ignore the distribution of scoring throughout the lineup.

This is easy to see if you look at seasons where scoring was driven more from PP time and the top 6 vs. seasons where there was less PP time and scoring was more even through the lineup. Adjusting by average scoring will largely overrate people in the first category compared the the second.