Mass's Advanced Player Rating System

Hello everyone!! A few of you have been requesting that I add the advanced stats into my rating system and I have finally done that. That is an additional 15 stats in which most are used for the new ratings. I will break that down below for everyone. If you don't care about the description and don't want to give feedback there is a link at the bottom you can skip to if you want to check out the player stats. For everyone else, feedback will be crucial so I can get this ready for s10. With all the added stats I had to do a lot of rough guessing on percentages in my formula so it needs plenty of tweaking. I want to thank everyone who has enjoyed what I'm trying to do so far and I want to thank all of those who have given great feedback both positive and negative. Everything you all say, I do my best to consider when I tweak the equations and it wouldn't be this in depth if it wasn't for the community

ADDITIONS AND FORMAT CHANGES
So when you follow the link, you will realize that the spreadsheet looks a lot different. For starters I have nicely colored it so it isn't as big of an eye sore. To the right of the original stats you will find all of the advance stats that are now added. The advance stats are a little confusing and may not make any sense but no worries. At the bottom where you would normally select positions there is now a STAT KEY tab that nicely describes what each advance stat is and how its calculated. If the descriptions need more examples or a better explanation please let me know and I will correct it. To the right of the player's advance stats is where you will now find their rating just like before. New to the spreadsheet is what is to the right of the player rating. Starting next season we will be able to track the ratings of all players weekly. The stats will still be cumulative but we will be able to see if anyone is playing better or worse as the season goes. I'm working on a coloring system to show that but it isn't working yet. Lastly, you will see that players are no longer separated by specific position. Forwards are now grouped and defensemen are now grouped. This could be good and bad. Before anyone complains that they liked it better separate, this was not my choice. There is some sort of bug on the LG site where I can not separate advance stats by specific position. Once that gets sorted, I will gladly separate the positions again. With this issue, it means that the face off stat is not being used for centers. I agree that it is an important stat but it does not work well in the formula for the wingers who haven't taken a face off or are 1 for 1. Of course, this will be fixed once I can separate positions again.

FORMULA CHANGES
This is going to be complicated as a fair warning. You may have to read this a couple times to get a good understanding. I will do my best to explain it.

Before I go into the positions and their weighted categories, I want to first explain how I possibly came up with my first percentages you will see lower in this post. What I did first was take all of the different stat categories and separate them into 3 categories; Offense, Defense, and Team Play. Some of the stats like giveaways can be tossed into really any of the 3 so there can be debate but really it doesn't matter. Now that the percentages are set I can just tweak them as needed but here is how it breaks down.

Offense
G%- 50%
A%- 30%
OZE (Offensive Zone Efficiency)- 12%
P2E (Powerplay Percentage)- 8%
TOTAL=100

Defense
OEE (Own End Efficiency)- 30%
SAF (Shots Against Factor)- 10%
PPKE (Personal Penalty Kill Efficiency)- 15%
GA (Goals Against)- 10%
TA- 30%
B- 5%
TOTAL=100

Team Play
+/- 10%
PIM- 40%
H- 10%
GA (Giveaways)- 40%
TOTAL=100

Other categories like Goal Support (GS), Goals, and Assists are also used in the equation but its not weight based. They are used to very minutely scale a player's offense up or down based on how many points they actually have in comparison to the league. It will keep players who have 27 points in 27 games much lower than players who have 54 points in 27 games like before. However, players who only blow out bad teams and get shut down by good ones will have a much lower rating than a player who puts the same amount every game and has slightly less points. So this improved formula will now reward a better rating to consistency over spontaneous performance.

Scalars for Positions
so now that the stats are divided into 3 categories, I then assigned weights to the categories based on importance to the position the player plays. These are of course just a basic starting point to look at and determine what we need to tune. They are as followed:

Forwards: Offense- 80%, Defense- 10%, Team Play-10%
Defense: Offense- 10%, Defense- 80%, Team Play-10%

To understand how these percentages work with the stats above I will give an example.
For forwards offense is 80% currently. Inside the offense category above, G% is 50% which is half of the offense. So to understand how many points a player can get for that stat you take 80%*50/100=40% So if a player is the best at scoring goals consistently they will receive 40 of their possible 100 from that stat. This is done for every stat in the formula until you get a final score. Another example is team play makes up 10% of a forwards score and penalty minutes make up 40% of team play currently. So, therefore 4% of a players score comes from Penalty minutes. If a player has 0 penalty minutes, they would receive 4 points for that stat. So as you can see with all the weights and percentages it breaks these stats down nicely. Now the hardest part is finding a nice balance in all the stats which is where the community comes in.

As I continue to adjust and make improvements that I think you all will like, I will definitely need the communities help to adjust these weights and percentages. So for those who read the entire thing, I say thank you and I look forward to all feedback good and bad you all have. Once again, special thanks to everyone who has supported so far and who plan to support me further

To view the player rating system follow the link below:
https://docs.google.com/spreadsheets/d/1_BlNqPLgb9qNtGNRA9Hq2hUlOix5_Nc_3w_aSBA6VVo/edit?usp=sharing

 

Love your weekly lists!

My 2 cents: 80% seems like a lot for forwards and defense, maybe 60-20-20 or 70-15-15?

Some D have a positive impact on offense and vice versa. It would probably make me look better at the same time! ☺️

Keep it up! Entertainment at its best

 

May want to convert any of the time based stats (ex: own end efficiency, shot against freq, etc.) to seconds instead because excel and google sheets like to convert that into like a time of day (ex: 2:05 turns in into 2:05 AM/PM). I run into this with my own stat sheets I do and just convert everything to seconds and it makes the sheet look a lot nicer.

 

There are ways you can manually set it to display how you'd like if that's a consistent issue you are having, you can even manually customize how it displays in a box but I do agree with you. I should format it into sec so it's easier to read.

 

Why is pims 40% while +\- is 10%?

 

Penalty minutes reflect a player's discipline and ability to not put their team in a tough spot and plus/minus to me isn't a stat a player can control well as it reflects the team play around them. Keeping it low also helps the players who are carried on good teams have glorified ratings because of one stat. Of course 8 can tweak this some if people dont agree.

 

i am still bad