The Pittsburgh Penguins reached the midway point of the 2013-14 season earlier this week and we took a look at the 2nd Quarter Results. But now we are going to take a look at that 2nd Quarter performance and compare it to how well they did in the 1st Quarter of the season. You may recall that earlier this season we did the same thing with the 1st Quarter Grades, comparing the player's performance to that of their recent seasons.
This time, however, we are comparing the 2nd Quarter to the 1st Quarter in order to get an idea of what players are improving throughout the year, the exception being two players, Despres and Ebbett, who had not played during the 1st Quarter but had previous season data available to use for comparison. As mentioned last time, the grades provided are not indicative of players being better than one another, but simply a matter of comparing the player to his own past performance in order to estimate growth. So if one player gets a B and another gets a D it does not necessarily mean that one is better than the other, but rather that the B player improved on his 1st Quarter performance while the D player has gone slightly downhill.
In order to assign grades I first had to calculate a number of different advanced stats from both the 1st and 2nd Quarter of the season. I then took the 2nd Quarter and 1st Quarter results and calculated the Percent Change, which is New Value minus Old Value divided by Old Value. I then took that Percent Change and added 75% to it because on the grading scale 75% is the average midpoint grade. From there I was then able to assign letter grades for each category based on the percentages and calculate an overall grade for each player based on the Grade Point Average scale.
Last time I had gotten some complaints about the letter grades appearing too low, so this time around I opted to use the Canadian grading scale, which is a bit of a wider range than the traditional US scale. Using this scale 100-80 is in A, 80-70 is a B, 70-60 is a C, 60-50 is a D, and anything below 50 is an F. That makes the players' grades look better, but keep in mind that it is simply a way to assign a value to whether or not the player has improved from one Quarter to the next. So a player with an A or a B has improved on their performance from the 1st Quarter and a player with a D or an F has declined from their 1st Quarter results.
We looked at the same advanced stats for Forward and Defensemen that we did last time, and like last last time weighted the individual stats more heavily than we did the on-ice team stats. The individual stats are based on 5-on-5 data while the team stats are based on 5-on-5 Close data. So for 4 grade points we have the individual performance represented by Goals per 60 minutes, Points per 60 minutes, Individual Fenwick Events per 60 minutes, and Fenwick Shooting Percent. For 3 grade points we have the on-ice team performance represented by Goals For per 60, Goals Against per 60, Goals For Percent, Fenwick For Percent, Fenwick Shot Percent, and Corsi Save Percent.
For goaltenders I added some more data this time around, having access to overall, 5-on-5, and Close data for the goaltender stats. The only goaltender stats I considered individual stats and as such assigned 4 grade points to were the Save Percent data, the remainder of the stats are on-ice team stats and as such were 3 grade points. The data we use is Points Percent, overall Goals Against Average and Save Percent, 5-on-5 Goals Against Average and Save Percent, 5-on-5 Close Goals Against Average and Save %, and Shutout Percent.
In addition, I have included the percent change of PDO and Zone Start Percent just as a reference. So a player whose PDO has increased significantly would somewhat explain why their performance may have increased this quarter, while a player with a low PDO could indicate that we don't have to be concerned about a drop in performance because the dumb luck stats suggests that they are due to bounce back. A player with a low PDO who has already improved this Quarter is that much more impressive, while a player with a high PDO but lower output is a bit of a concern. Likewise, a player who sees more time in the O-Zone this Quarter should be expected to improve his numbers, while a player that has seen a decrease in O-Zone starts could be forgiven for a slight decrease in output.