This will be the first statistic I will be introducing on this blog. The way hockey's offensive statistics work is that for every assist and goal one accumulates he receives a point. Basically the formula for points is: G (goals) + A (assists) =P (points). This seems very logical and it’s easy to quantify and that's why it’s considered basically the only statistic used for total offensive production. Points actually determine who wins an award in the NHL automatically (The Art Ross Trophy which is given to the player who has the most points at the end of the season). Points also is a heavy contributor in determining other awards like the Hart Trophy (NHL's Most Valuable Player), Norris Trophy (NHL's best defenseman that usually goes to an offensive defenseman), Lester B. Pearson Award (NHL MVP voted by the players), Calder Trophy (NHL rookie of the year) along with the Con Smythe Trophy (Playoff MVP). It also is used to determine potential NHL hall of famers and is used by general mangers when they discuss contracts and trades.
I disagree with the sentiment that points are a great statistic for measuring offensive value. It is flawed because it a, doesn't take into account quality of teammates and linemates b, powerplay time and c, doesn't take into account the rarity factor of how "easier" it is to accumulate assists rather than goals. The first two factors are ones that I hope to address later in time on this blog but the first one I will be addressing today. The ratio of assists to goals is generally considered being in the 1.5 to 1.75 to 1 range. My theory is that now that assists are more common than goals, goals should be considered more valuable in the points statistics because they are rarer. Either that or abolish "the second assist" which is the pass that leads to the pass that leads to the goal. These second assists lead to the overpopulating of the assist statistic. However, abolishing this isn’t going to happen so I am creating a new statistic which is called Rationalized Points. The best way to do rationalized points would be to keep track of the total amount of goals and assists in any NHL season find out the ratio and go from there. But this is an impossibility because of the lack of hockey statistics available regarding league totals and the fact that it would lead to inconsistent formulas along with ridiculous decimals. So I am adopting an arbitrary number of 1.5 as the ratio. This is on the lowest end of the projected ratio so goals are likely still undervalued. I adopted this number because I don’t want overvalue goals and have people complain on that subject and also 1.5 leads to only decimals of .5 and not more than that which makes it neat and clean. Anyway my formula would have the number of goals be multiplied by 1.5 and then added to the assist total to come up with the Rationalized point statistic.
Basically the current equation is: P= G+A
My equation is: RP= 1.5G+ A
Take into account two players:
Player A: 25 goals 90 assists
Player B: 60 goals and 45 assists.
Based on the points statistic:
Player A: would have 115 points
Player B: 105 points
Therefore player A would be considered the better player.
However with RP:
Player A: 127.5 RP
Player B: 135 RP
Therefore player B is the better offensive player. The difference here is pretty big considering Player A is leading by 10 points but loses his RP by 7.5 points. A 17.5 difference. Obviously the range isn’t generally this big but it can happen and RP still is more accurate.
Now I’m going to apply this formula further to account for games played. Obviously missing games is a bad thing and hurts your chances for awards and such but per game statistics also help project statistics for a player when healthy and show his per game ability level. Yes injury prone players may never achieve this but some players may be having a fantastic year but have a fluke injury half way through and their pure numbers don’t reflect how good a season they are having.
Example:
If I had been for most of my career a 30 goal 30 assist man in 82 games, since I had never been injured. This would give me an average of 75 RP a season. But if one year I played 41 games and had 20 goals and 20 assist to start that season my RP would only be a measly 50 for the year and people could consider it a down year for me. However, my RP per game would be 1.22 compared to my usual 0.91.
Anyway the formula for RP per game is simply.
RP per game= RP/GP
If your working from the beginning and want one equation do the following:
RP per game= (1.5G+ A)/GP
One last tidbit argument for RP
Consider the following in 2008/2009 in the NHL there was one 50 goal scorer (Ovechkin). There were twenty 50 assist players in the NHL that season including six who had 60-69 assists and 2 who were over 70 assists.
There were eight 40 goal scorers that season, and fifty three players who had 40+ assists.
This trickles down to 30 goal/30 assist guys, 20 goal/20 assist guys etc. Of course the ratios do decrease but it is always a significant more players who reach the markers with their assists than players who reach the markers with goals.
Thought on RP and RP per game ladies and gentleman? This is our first new NHL statistic and I will be using it throughout my comparisons of players. In my next post I will likely do a couple rerankings of past seasons showing who finished where in the Art Ross race and who would finish where in the RP race.
PS: I made up this statistic myself. Any statistical formula I find and I like that I post here I will source to its original owners.
