MASS's Player Ratings, After Week 6 (AHL)

Hello everyone, once again, if you did not see my comment earlier, I apologize for the late and incomplete post. I spent a lot of time reworking the offense calculation because of the goal support bug discovered last week. Thank you to everyone who commented and contributed to that discussion. I believe I found a fix for it so please take a look and let me know what you think. I am in a rush tonight to get this posted for you all before week 7 starts so I will not be posting the changes yet. Give me a couple days and I will edit this thread to give everyone the breakdown of the issue discovered and what I did to fix it. Thank you everyone for the feedback and support! Enjoy!

*UPDATE* 4/1/19
Issue- Some players with very low goal support (Under 2) were getting too high of offensive ratings if they were scoring about half of the points or a point per game. The issue was found originally in the NHL spreadsheet where a player with 17 points in 10 games was rated 5th because his team was scoring 1 point a game. Before the fix your G% and A% was divided by your goal support to determine the likelihood you would score a goal or assist in a one point game. Because of this set up a player with a 5 for goal support would need a G% 5x better than a player with 1 for goal support. So if the 5 goal support player had a G% of 100 meaning he scored atleast one goal every game, the other player would need a 20% to have the same rating. This stat by itself was used as a consistency stat and was not the basis for most of a player's offensive rating. However, the handful of points awarded for having a good stat for this was enough to elevate players up to 20 spots higher or lower, depending on what end of the goal support they were at and their rank in the standings.

Current Fix- The new calculation is as follows: Weight*(G%/100)*[1-(.1*GS/maxGS)] where GS means goal support and maxGS is the max goal support by any player in that group. G% means the percent chance you will score atleast one goal in any given game. (It is a consistency stat.) Weight is how many points gets contributed to your score for this stat. A perfect stat of 1 will give you the full weight score and a 0 (never scored in a game) will add 0 points to your overall score. This weight is different by position though which will be broken down below.

To describe what is happening in the equation it works like this: The player who has the highest scoring line, so best GS, is expected to have a 10% better G% than the lowest GS player to receive the same weighted score for this stat. So if the best GS player has a G% of 100% then the worst GS player would need a G% of 90% to receive the same score for this stat. This is scaled proportionally by that 10% stated above. So if the best GS player had a G% of 50% then the worst GS player would need a 45% for G% to get the same score. This same setup was done for A% as well.

This change also affected the other offense values because of how it was written in the formula so they all were adjusted accordingly. The new breakdown for offensive stats are as followed:

*This is the max points you will get added to your score for being the best in these 4 stats*

Forwards:
G- 22
A- 17
G%- 15
A%- 11

Defense:
G-8
A-4
G%-3
A%-1

I hope this clears up the issue and fix. Let me know if you have any questions, thoughts, feedback, etc. Thanks for the read.

https://docs.google.com/spreadsheets/d/1_BlNqPLgb9qNtGNRA9Hq2hUlOix5_Nc_3w_aSBA6VVo/edit?usp=sharing

 

I like this system, makes me look good. On the same note, I've had 2 average/bad weeks but still remain in the top 5. Anything standing out that's keeping me up there? @xMAssStAtiC-

 

@Rezhomey You have one of the higher takeaway counts, and a great OEE (own end efficiency), and a few points to help gain a few points to your ranking. Penalty minutes is killing you but your huge hit count is helping soften the blow.

 

Moving up in the ranks but God damn, my PK % is heinous.

EDIT: Rez, you're a superstar, stop being so modest.

 

Wish I was good

 

Don’t we all