If you don't understand the concept of sample size, then look it up. It'll help you analyze stats much better.
I'll try to help. Really.
Take last season. If you look at the sample of all players playing one game or more in goals produced per 60 minutes at even strength, then you have this:
http://www.behindthenet.ca/nhl_stat...13_s&f2=5v5&f7=1-&c=0+1+3+5+4+6+7+17+18+19+20
You'll see that none of the top 10 played more than 6 games and 5 of them were defensemen. You will probably not categorize any of the top 10 on this list as elite players. Ryan Johansen was 25th in this sample size.
Moving the sample to a minimum of 10 games played, you have this:
http://www.behindthenet.ca/nhl_stat...3_s&f2=5v5&f7=10-&c=0+1+3+5+4+6+7+17+18+19+20
Already the defensemen have been eliminated as one would expect. 5 of the top 10 players are commonly recognized as very good players. Jack Skille topped this list
Johansen was 14th on this list.
Moving to 30 games and this is what you get:
http://www.behindthenet.ca/nhl_stat...3_s&f2=5v5&f7=30-&c=0+1+3+5+4+6+7+17+18+19+20
7 of the top 10 are those one would expect on the list. Johansen cracked the top 10 at the 30 game minimum.
Moving to the 60 game mark and this is what one sees:
http://www.behindthenet.ca/nhl_stat...3_s&f2=5v5&f7=60-&c=0+1+3+5+4+6+7+17+18+19+20
By this point, all of the top 10 are represented by top players. This is to be expected once one removes shorter term variables from the equation. Ryan Johansen is tied for 6th in this sample size.
The larger the sample size, the greater the accuracy of the statistic in reflecting true performance. If the Jackets would call up a player from Springfield who scored 1 ES goal in one game and was sent back down for the rest of the season, he would lead the league in ES goals produced per 60 minutes IF ONE USED A ONE GAME MINIMUM SAMPLE. That surely wouldn't indicate that he was the most dangerous goal scorer at even strength in the entire league
No one could possibly argue that the top 10 of the one or 10 game sample are as capable scorers as a group than those on the 60 game minimum list.
I hope that helps explain why size of sample is important in drawing valid conclusions from any given statistic.