It fails to capture the variance in the # of PP called in different seasons which is not reflected in changes to the league GPG.
All adjustments fail at some point (GPG, VxX, whatever), but using raw data across eras is almost always worse than using any adjustment.
IMO, a performance vs. peers is the starting point for a statistical analysis, not a adjusted points vs. adjusted points. This assumes that a peer group of the best 2o to 30 offensive forwards is stable year after year save for outliers like Wayne and Mario, and consideration for the O6 era.
IMO, when two players from two different eras have proven to be similarly dominant vs. their peers, they are on the same tier in terms of offensive production.
This method is great when the players' leagues are close to the same size, but starts to fall apart when that number gets too far apart. Finishing 10th is a 31 team league is a lot more impressive than finishing 10th in a 6 team league.
The top 30 goal scorers in 1982 posted these goal totals:
1. Gretzky - 92
2. Bossy - 64
3. Maruk - 60
4. Ciccarelli - 55
5. Vaive - 54
6. Stoughton - 52
7. Middleton - 51
8. Messier - 50
9. Dionne - 50
10. Trottier - 50
11. Stastny - 46
12. Barber - 45
13. Hawerchuk - 45
14. Secord - 44
15. Propp - 44
16. Pederson - 44
17. Lukowich - 43
18. Smith - 43
19. Goulet - 42
20. McDonald - 40
21. Duguay - 40
22. Napier - 40
23. Sutter - 39
24. Tardiff - 39
25. Wilson - 39
26. Taylor - 39
27. Broten - 38
28. Petterson - 38
29. Walter - 38
30. Gillies - 38
These 30 players combined for 1402 goals out of the 6741 that were scored League-wide in 1982 (or approximately 21% of the total).
The top 30 goal scorers in 1999 posted these goal totals:
1. Selanne - 47
2. Jagr - 44
3. Yashin - 44
4. Amonte - 44
5. LeClair - 43
6. Sakic - 41
7. Lindros - 40
8. Fleury - 40
9. Satan - 40
10. Robitaille - 39
11. Kariya - 39
12. Graves - 38
13. Berezin - 37
14. Demitra - 37
15. Tkachuk - 36
16. Naslund - 36
17. Straka - 35
18. Modano - 34
19. Hull - 32
20. Clark - 32
21. Bondra - 31
22. Turgeon - 31
23. McEachern - 31
24. Shanahan - 31
25. Sundin - 31
26. Forsberg - 30
27. Primeau - 30
28. Guerin - 30
29. Kozlov - 29
30. Khristich - 29
These 30 players combined for 1081 goals out of the 5830 that were scored League-wide in 1999 (or approximately 19% of the total).
The decrease in the percentage of goals scored by the top 30 goal scorers between 1982 and 1999 (a decrease of 23%) is out of proportion to the decrease in the percentage of goals scored per game between 1982 and 1999 (a decrease of approximately 34%). This indicates that a larger percentage of goals were being scored by the top 30 scorers in 1999 than in 1982.
The reasons for this could be related to increased PPOs and/or ice time for elite scorers. It's also possible that higher-end scorers simply were simply better able to overcome the improvements in goaltending and defensive play than less talented players.
If you use the analogy of equating goals to currency, then adjusted goal totals accurately represent the values of goals on an adjusted scale. To this extent, these numbers have a purpose. Individual goals are worth more when goals in general are less common.
However, whether point and goal adjustments based on League-wide scoring rates are "fair" to the elite scorers of a given era is a different question. Even if we account for advantages and disadvantages in PPOs, games played and ice time for elite player, it's possible that adjusted point and/or goal totals of elite players from higher scoring eras will be more negatively impacted than perhaps they should be by scoring adjustments based on League-wide scoring rates than the scoring totals of elite players from lower scoring eras.
In my opinion, when differences in games played, PPOs and ice time for elite players are accounted for, adjusted goal and point totals are still a more meaningful basis for comparing the scoring totals of players from different eras since the 1967 expansion than using raw point totals, which obviously heavily favor players from higher scoring eras.
I'm sure you'll agree Maruk's goal total of 60 from 1982 cannot be compared directly to Jagr's goal total (44) from 1999 in any meaningful way. Their adjusted goal totals from those seasons (51 for Jagr in 81 games and 44 for Maruk) are a better representation of the impact of each player's goal scoring in 1982 and 1999 and a better indicator of each player's goal scoring aptitude as well in this case.
In 1917, when their were 4 teams... the top-30 scorers scored 99% of the goals.
In 2117, when there will be 60 teams... the top-30 scorers will score 5% of the goals.
1982, 21 teams (21 top players = 15.6% of the goals)
1999, 21 teams (27 top players = 17.0% of the goals)
Say Bure gets hurt that year and Jagr finishes number 1 with the exact same goal total. Does that make Jagr's season more impressive? It shouldn't and that's why I don't think comparing scoring relative to their peers is really all that indicative of being a better player.
And this is when we need to have the option of looking at per game adjustments too in addition to adjusted totals.
Do you have a recent example of a player whose numbers should be adjusted?
Chris Mason 06-07 (spent the year beating up on Chi, St.L & CLB)
Every player who played in the SouthLeast division when we called it the SouthLeast division.