NHL: Here's How I Would Select All-Stars Instead of Depending on Fans

I'm writing this out of protest for the apparent bias that the current NHL procedure has in selection of the players to play in the NHL All-Star Game this season.

It's clear that with four Pittsburgh players (and zero players from some teams) picked that there is a flaw in the system. Here is what I would propose as a semi-permanent selection procedure, until a more fair one comes along.

Forwards:

For the 24 selected forwards, the NHL would pick the top scorer on each of 12 East and 12 West teams.

Defense:

NHL selects eight d-men from each conference who have achieved the highest scores in a weighted formula of ice-time, penalty-kill minutes, blocked shots, etc., making sure to select at least three of them from teams who did not send a forward to the game.

Goalies:

NHL selects the top three goalies in each conference based on a weighted formula of win percentage, games played, save percentage and GAA. Each goalie selected will play one period in the game. 

A similar concept:

Using the above strategy, expand one roster out to select from the 29 non-Cup champs from last season (which will provide some alternates, etc.). Pit them against the team that won the Stanley Cup last season, as they did when the All-Star Game first existed. 

A third idea:

A mesh of the fan-vote system with the first strategy. Fans vote for the team they support, but only the top vote-getter on each of the 30 teams (goalies excluded) can be selected (provided he fits one of the needed roster spots). Three other skaters per team selected by NHL, based on the previously listed criteria.

NHL selects the top three goalies in each conference based on a weighted formula of win percentage, games played, save percentage and GAA. Each goalie selected will play one period in the game.