CH said:
That Mario had more dominant seasons, when compared to other dominant seasons in history, despite not outscoring his better peer group by as large a percentage as Gordie had.
Ogopogo said:
How does that even make sense? Winning a scoring title by a greater percentage is clearly more dominant.
I HAVE explained this repeatedly. Lemieux scored MORE normalized points.
Gordie was playing against a weaker group of players.
Lemieux's seasons (such as his 199 point season) was a better offensive year by a significant margin then anything Gordie Howe ever managed to accomplish in his lifetime.
Its the same thing as a movie drawing a larger amount of money one week then any other movie in a different week, despite the number one drawing movie in the second week outdrawing the second movie in its week by a larger percentage.
When you compare Lemieux's best seasons to Howe's best seasons, Lemieux's are better. When you compare Howe's best seasons to those of his peer group, Lindsay, Richard etc, Howe may lead by a larger margin then Lemieux over his peer group Gretzky, Yzerman etc., but that does not mean Howe's season was better than Lemieux's. In fact in this case the opposite is true.
Ogopogo said:
Fact is, there are more players from the 50s and 60s among the top 50 players of all time than from the 80s and 90s.
Perhaps, you just don't know much about hockey prior to the 80s so you just deny that it was any good?
Thats easy, the top 50 players as you define them is WRONG. Players in the earlier days had a weaker peer group then they do today, so when you compare players to their peer groups, weaker players in the weaker peer groups will falsely appear better than stronger players in stronger peer groups.
The problem is that you do not know enough about hockey. You see these players as lines of statistics that you are unable to put into context.
Lets try a baseball example (its always good for teaching pruposes to apply methods to a wide range of topics because it will allow you to see things a new light if you apply them in areas where you lack the same preconceived biases)
In 1902 Socks Seybold lead the major leagues with 16 home runs. In 2nd place, there was a 4 way tie of players who hit 11 home runs. Seybold led the majors by a whopping 31%.
In 1998, Mark McGwire hit 70 home runs. In 2nd place, was Sammy Sosa who hit 66 home runs. McGwire led by 5.7%.
By your method, Seybold was more dominant. This is only because of a false normalization, the peer group of home run hitters in 1902 was clearly far worse than it was in 1998. I dont think anyone would argue that Sammy Sosa didn't have a more dominant season in terms of home runs in 1998 then Seybold did in 1902 EVEN THOUGH, Sosa did not even lead the league and Seybold led by over 30%. Alex Rodriguez hit 42 home runs in 1998. That was good enough for 12th overall. And he had a more dominant home run hitting season than Seybold did.
The hockey example is less obvious then the baseball one, but Lemieux played in an NHL with more good players then Howe did. Lemieux scored more points both before and after normalization in his prime seasons then Howe did and he did so against a better group of players.
In fact, it is far easier to "dominate" by % victory in Seybold's day, because the 30+% difference was only 5 home runs. Most of the largest % victory will come from the earlier days when there peer group is not as strong. If you use this number to determine who should be rated in a top 50 players you will falsely include too many early players who take the place of more deserving more recent players. In fact this is exactly what you have done.