New +/- System

Over the past few days I decided to take a look at the best variables from plus/minus and corsi to create my own “plus-minus” system. Only scoring plays will be looked at (I just don’t have the resources to do it for every chance/shot etc.) but I attempt to eliminate the variables. For instance, Claude Giroux blocks a shot at the point, chases it down, buries the break-away attempt. Why should anyone else get credited with a “plus-1” when Claude was accountable for 100% of that goal? On the defensive side, if Chris Pronger’s pass gets intercepted when he’s going point-to-point, why should any forwards get a “minus-1” for that play? The answer to both above questions is “they shouldn’t.” Statistics are very quantitative and the reason that my system is different is because it takes general hockey strategy and variability into account. Let’s look at an example from the first even strength goal of the season scored in Boston on October 6th by Jake Voracek. Take a look at the goal.

                Traditional hockey stats would award a goal to Voracek, a primary assist to Meszaros, and a secondary assist to Simmonds with all five players (Carle and Briere included) getting “plus-1.” Here’s how I broke it down:

·         Voracek was responsible for 50% of that goal with his being in good position for the rebound and driving it home.

·         Wayne Simmonds drive to the net and pass to Meszaros is good for 20%.

·         Meszaros’ drive from the point gets 20%.

·         Briere drawing Chara gets 5% and Matt Carle gets 5% for the (not seen in the video)original drive into the zone.

There was a lot going on during that play, lots of players were involved and they all deserve some form of credit. Now lets take a look at a different kind of example, the first goal from that game  (on the power play, but that doesn’t matter because I break the stats up into 5-on-5, 5-on-4, and 4-on-5).

·         Outlet by Pronger to Jagr gets Pronger 10%.

·         Jagr’s sweet pass to Giroux: 20%

·         Giroux handling the pass, turning Chara inside out and finishing: 70%

This play shows why this rating system can be implemented and effective because it shows who really did the scoring and who really messed up the defensive assignments. Here's a defensive example: (go to 1:38)

·         Giroux: He gets 10% for his weak attempt to clear the puck by batting it with his stick. Probably should have gone for the body in an attempt to tie up the puck in the slot.

·         Jagr: Mishandles Coburn’s outlet leading to the turnover. When Coburn over pursues to the opposite side of the net, Jagr is not able to cover quick enough, leaving the scorer with the easy put back. 45%

·         Coburn: 50%. Handcuffs Jagr on the outlet. Over pursues leaving Sedin all alone with no one to cover for him.

·         Timonen: He was there, good effort on the block, but defenders should get some responsibility when there is a goal in front: 5%.

·         vanReimsdyk: 0%

This is the textbook example defensivly because there were people at fault (Coby/Jagr) and people not at fault (JVR). Now that you understand the system, here’s the data for the past four games in all three situations:

That's the ES raw data, here is the cleaned up version:

Now the power play:

Here is the raw 5-on-4 data:

And now the raw 4v5:

And the clean 4v5:

If anyone is good at pivot tables I think you can extract a lot more data out of my raw numbers. I can send this information to whoever would like it and I do plan on keeping these numbers for the foreseeable future to see if any patterns do develop. Other than that, tell me what you guys think?

UPDATED AFTER WASHINGTON GAME

Value= Raw +/-'s added up for net value

ESTOI/60= Assuming that is 5-5 so Claude's 4 on 4 goal last night would not count ***(how do you guys feel about adding 4 v 4 to the ES data)***

IPM/60= Individual Plus/Minus per 60. Making value relative to time on ice