With 20 games played by all teams in the NHL, we like to take a look to see how forecasters projections are faring thus far.   

It is early on in the season and lots can and will change from now until all 82 games have been played. 

However, it is fun to take a quick look to see how the forecaster's projections you entrusted with your team is doing compared to other forecasters. 

For this progress reports we have only looked at forecasters' Pts Rank, Pts TotalPPG Rank, and PPG Total projections.  We would like to have updates on all the statistical categories that we are keeping track of but unfortunately we do not have the time to do it!  We should be able to include those in our Half Season Progress Report. 

Below you will find figures that include the forecasters being evaluated, their rank in comparison to each other, their total points (based on our calculations, max 100), and their average error in their projections. 

Results for Rank projections are posted in orange.

Results for Total projections are posted in blue. 

We personally believe that rank projections are more important in comparing projections (see rationale), however most people do care for projection total accuracy more.  We have also incorporated feedback from a few forecasters themselves and have include a comparison of Points Per Game projections this season. 

Another change for this season, is the total number of projections that we are going to be comparing for each forecaster.  For the past two seasons we have used the Top 300 projections of all forecasters. This season we are going to provide rankings on the Top 100, Top 200, and Top 300 projections.  As we always say, "Mo' Data, Mo' Interesting"... and in reality it was not "Mo' Problems" for us as it only took a minor tweak of the Excel spreadsheets. 

Some items to consider:

  • To keep the GP sample size, we collected stats from all the teams once they played 20 GP (for Pittsburgh that was last week, for Ottawa it was on Wednesday).
  • Inside Hockey projected a Pts Range for players. We took the median number of the range and used that for comparison. 
  • Pete Jensen and Inside Hockey had less than 300 total projections for skaters and were not included in the 300 player projections rank. 
  • If forecasters did not specify GP in their projections it was assumed to be 82 games. 
  • If a projection was made for a player who hasn't played a game, a Pts Rank Error of 300 was used. 300 was chosen because it represents the total number of player projections in the 300 comparison.  For example if a forecaster projected Mike Fisher to score in the top 300, they will get receive a pts rank error of 300 (not the total number of players in the league).

 

 

Pts Total Projection Rankings:

Top 300 

Top 200 

Top 100 

 

Pts Rank Projection Rankings:

Top 300

Top 200

Top 100 

 

Top 25

We have also looked at the projections from the league perspective.   Below is a chart of the top 10 point getters at the 20 gp and how accurate forecasters projections were for them. 

 

 

Forecaster Results Summary Sheets

 

Article Last updated October 21, 2016.

In our first analysis of the 2016/17 regular season projections we chose to take a look at the most disagreed and agreed upon player Pts projections by all forecasters.  This means that for a player projection to be considered for this analysis, there must have been a projection made by every single forecaster the 11 who chose to evaluate this season for the player. Out of the 768 players (not including goalies) that are in our projection database, there were 216 players in which a projection was made by every single forecaster.

We have created a fantasy hockey pool for 2016/17 season using the 6 forecasters that we evaluated in the previous year.

Below are two figures which display the results of the six forecasters' predictions for the top 5 players drafted in 2015.   Figure 1 displays the