Friday, April 30, 2010

My System for Awarding Scoring Chances

During the 2009-10 season, I have tracked scoring chances for a total of 73 Toronto Maple Leafs' games, and lifted data for three other TOR games tracked by three bloggers who'd partaken in counting scoring chances. I've been posting the results of this project over the past week or so, which you can find by scrolling down the main page or by clicking the links on the right side of this page underneath "Scoring Chances for the 2009-10 Toronto Maple Leafs".

How I Awarded Scoring Chances

I marked down a scoring chance each time a player takes a shot toward the goal that I feel was a good opportunity at scoring. During each game, I noted each scoring chance in a notebook, writing down the period, time left in the period, the team that got the chance, assigned a letter rating (see below), and whatever other notations I wanted to make. After compiling these for the whole game, I'd enter this data through Vic Ferrari's scoring chances application. Thanks to Vic, we can see who the players on the ice were for each scoring chance and the team strength (number of players on the ice versus the number opponents on the ice). As the Leafs were struggling early in the season, I was considering counting scoring chances for some games to see how they were faring,

There are quite a few factors to look for in determining a scoring chance. Several of them are

1) Positioning of the shooter on the ice when he took the shot. Was he within what I've dubbed the prime scoring area?

Prime Scoring Area: A shot taken from within this area is almost always counted as a scoring chance by anyone counting them, providing it is a half-decent shot attempt. What I consider the scoring area is from within approximately the bottom quarter-section of the face-off circles nearest the goal, starting from around the face-off circle dots, and in towards the sides of the net. Other chance counters may use a home-plate shaped area that goes from the posts to the faceoff dots and extends back to the end of the FO circles nearest the blue-line. Shots from outside may be counted at my discretion depending on other factors.

2) Shot Selection: Shot type, distance, velocity, immediacy of the shot (for example, an shot taken immediately after receiving a pass, like off a one-timer).

3) The ability of the goaltender to clearly see the puck that is coming toward the goal. If there is enough traffic in front of the goal (skaters positioned or skating about in front of the goal) or a screen (a player positioning himself directly front of a goaltender), I will award a chance off a longer distance shot if it gets through and on goal.

4) Tips/deflections, one must also watch closely for these when a player takes a shot from outside the "Prime Scoring Area" and there are players in front or near the goal.

A shot on goal or a missed shot can be a scoring chance, but not a blocked shot. In that case, the defender did his job to prevent the shot from getting through. Sometimes, however, it is difficult to see if a shot did get through or not so the occasional mistake is likely to happen when counting chances. To me, a block is when a defender gets in the way of the puck and stops it from getting on goal, slows it down enough to make it a non-threatening shot, or deflect it wide of the goal. Otherwise, the shot may be counted as a scoring chance even if the NHL's official shot counter calls it a blocked shot.

After counting a scoring chance, I'd give a letter rating of A, B, or C for the chance I just marked down.

A chance rating:
A's were awarded to what I felt were very high-quality chances. For those that didn't result in goals, only a BIG save or a missed shot prevented the puck from going in the net. Odd man rushes, breakaways, rebounds, among others usually get this grade if the shooter gets a dangerous shot off.

B chance rating:
A solid chance at scoring... a good shot, often taken within the Prime Scoring Area. A screened shot may warrant this rating depending on the difficulty of the shot. A breakaway or odd-man rush may be rated a B quality chance instead of A if the puck-handler doesn't challenge the goalie much (say he skates straight in and shoots it right into the goalie's chest or something like that). Even an immediate shot taken from a little distance outside the area may get this rating at my own discretion.

C chance rating:
A lesser-quality chance at scoring than an A or B but a decent one nonetheless, most often taken from outside the Prime Scoring Area. Usually a tipped shot from farther out or a shot through traffic/screens that make it difficult for the goalie to see the puck, or sometimes a flubbed shot from within the scoring area.

2 comments:

Olivier said...

Cool writeup; what do you intend to do with the A-B-C classifications?

Also, you must read this interview with Jim Corsi by David Staples of the Edmonton Journal.

Keep up the good work!

Slava said...

Thanks, Olivier. Some time in the next week I shall put up a couple of posts containing data on High Quality (A) chances (and there may be further analysis in the future). I figured I should get this one out of the way first before putting up the HQ chance stats. There will be one post for HQ shot attempts by Skaters and HQ chances faced by each goalie, and the other post will display HQ chances for and against while on-ice for each TOR player at Even-Strength. Interesting stuff in there, like the contrasts between Mike Komisarek (stay-at-home defender) and Dion Phaneuf (and his more up-tempo, high-risk style of play).

Thank you very much for the link to the Corsi article, looks like there's some good stuff to digest in there for sure.