The Common Sense Scoring Index for 2011-12
When I resurrected Chris Hollis' Common Sense Scoring Index last season, I had no idea what I was getting myself into, nor did I have any idea how satisfying the project would be at the end of the season. Thanks to all of the commenters and emailers who helped with the game-by-game ratings, we were able to accomplish an excellent scoring system which would allow us to say with confidence which of the Red Wings' players were pulling their weight and which were more detrimental to the squad.
Going forward, we've got a slew of improvements to the index. First and foremost, instead of tallying them up at the end of the season, we will have a running tally of the subcategories which lead to total adjustments. Both points and plus/minus adjustments will have complete breakdowns laying who is doing what and how. Those categories are laid out below. Each recap will have by-game ratings and will include a season running tally.
More important than that, the stats themselves will be housed at HockeyCSSI.com. Thanks to the programming work done by ChuckyD, these stats will not only be corrected for the context of the game, but also included in the context of each player's total ice time, tracked exhaustively and able to be cross-referenced with any number of other statistics. Want to know how many of Holmstrom's screener's assists happen on power plays as opposed to 5-on-5? We'll be able to show you that. Want to know which players shift it on and assist on their own goals when the Wings are trailing? That won't be hard to look up either at HockeyCSSI.com.
Keep reading for the categories.
Points Adjustments:| Screener’s Assist | Given to a player whose body position prevented a goaltender from seeing a puck to make a save. | ||
| Third Assist | Given to a player judged to be the third person in a play worth of earning an assist for a goal. | ||
| Fourth Assist | Given to a player judged to be either fourth or fifth in a play worth of earning an assist. | ||
| Self-Assist | Awarded to a player judged to have made an assist-worthy play on a goal he scored himself. | ||
| Non-Touch Assist | These are given to a player who, by positioning, prevented a player from the other team from being able to make a play to prevent a goal being scored. | ||
| Bonus Assist | Awarded to a player on a play which has been judged that, by its virtue, is special enough to warrant giving the same player more than one assist on the same goal. | ||
| Assist Lost | Designed to keep track of times when an official assist has been taken from a player due to context. | ||
| Goal Credit | Designed to track goals awarded or lost due to the context of a scoring play. |
Plus/Minus Adjustments:
| Goal-Scored Plus | These are plus ratings awarded to a player when it was deemed that a defensive contribution he made on the ice (whether by starting or preventing transition) helped lead directly to a goal. | ||
| Coverage Minus | Given to a player who made a mistake in defensive coverage that was determined to have led directly to an opponent scoring. | ||
| Turnover Minus | Given to a player who was judged to be directly or indirectly at fault for a turnover that the opposition used to score a goal against his team. | ||
| Overall Plus | Given to a player whose overall play during the game was ascertained to have positively impacted puck possession for his team in a way that he was not properly credited for in the official stats. | ||
| Overall Minus | Given to a player whose overall play during the game was ascertained to have negatively impacted puck possession for his team in a way that he was not properly credited for in the official stats. | ||
| Penalty Plus | Given to a player who worked to force the opposition to take a penalty. | ||
| Penalty Minus | Given to a player who either committed a bad penalty of his own or made a mistake which forced his teammate to take a bad penalty. | ||
| Shift Change Plus | Created to allow for correction of official plus/minus stats to either give or take away credit on a play when a goal is scored during or after a line change in which a more deserving player was not on the ice during the scoring play. | ||
| Shift Change Minus | Created to allow for correction of official plus/minus stats to either give or take away credit on a play when a goal is scored during or after a line change in which a more deserving player was not on the ice during the scoring play. | ||
| Power Play Plus Lost | Designed as a minus category to clear undeserved plus ratings from players who were on the ice when a goal was scored between the time a power play ended and the penalized player was able to get back in the play. | ||
| Penalty Kill Minus Cleared | Designed as a plus category to clear undeserved minus ratings from players who were on the ice when a goal was scored between the time a power play ended and the penalized player was able to get back in the play. | ||
| Goal Against Minus Cleared | Given to a player who was deemed to have been in position and not-at-fault for a goal scored. These are also given to players on the ice for a bad goal against. A player may have minuses added on for specific faults, but he will have his official minus cleared to track goaltender fault. | ||
| Goal Saved Plus | Given to a player who does anything that prevents what should be considered a surefire goal. Lifting the stick of a player preparing to receive a pass on a wide-open backdoor and outright making a save on a shot are both examples of this. |
Adjustments can be made by half-points or whole and players are not limited to a maximum of one on a plus or a minus. If a play is good enough, a player can find himself with more than one assist. If a turnover is bad enough, he can find himself with multiple minuses on the same goal. Large adjustments are uncommon, but are used when called for.
Goalie Ratings:
| Bad Goal | Given each time a goalie allows the puck in the net on a shot or play that he is deemed to have been expected to make the stop. | ||
| Goal Forgiven | Credited to a goalie who was judged to have done everything right to prevent a goal, but saw one getting by him through bad luck or bad play by his defense. An example would be for one of his own players deflecting a puck past a goalie. | ||
| Head-to-Head Plus/Minus | A rating of -1, 0, or +1 given to a goalie as a measure of how well he played in comparison to the opposing goaltender. | ||
| Overall Game Rating | A rating on a scale of -3 to +3 (0 being average) given to a goaltender for his performance during a game. Ratings on the severe end of the scale should be used very judiciously. |
Bad goals will be given throughout the game and will always result in a player getting a +1 (or +0.5) worth of plus credit. In situations where a bad goal is scored and players still deserve fault, they will be given additional minuses credited with the kind of mistake they made to allow the opposition to get the shot off which led to the bad goal.
The other two ratings are a result of a hybrid system to track goalie play. The Head-to-Head Plus/Minus is built with the idea that a goalie's ultimate job is to be better than the guy on the other end of the ice. This will help us track whether a goalie over the course of the season is the kind of guy a team can count on to consistently perform at the level he needs to. The Overall Game Rating is an alternate system by which a goalie is rated against a consistent standard of how well he should perform, regardless of whether the goaltender on the other side of the rink has a good game. As the season goes on, we will get a much better idea as to how well these ratings systems work.
The most important thing to the CSSI is user feedback. Posts will generally go up the morning after games. While I strive for consistency and relative objectivity, your perspective is crucial in making sure the ratings stay consistent.
Please feel free to ask any questions about this Index or specific situations and how they would be scored. Also, please check out last year's posts (under CSSI in the Sections area of the left sidebar) for guidance on how things will be scored.
Here's to another great year of Red Wings hockey!
20 comments
|
0 recs |
Do you like this story?
Comments
Bonus points
I’m assuming these are the “Overall Plus/Minus” in the player categories. Right?
Also, you’ve got “Turnover Minus” for giveaways leading to goals against, but are you going to attempt to track takeaways not credited in the official stats that lead to goals for? For example, forcing a bad pass that ends up on the stick of a Wing who then scores?
Bonus points
I’m assuming these are the "Overall Plus/Minus" in the player categories. Right?
Yes, those are the bonus ratings from last season. The title of that section will change from “Bonus Ratings” to “Overall Adjustments”
Also, you’ve got "Turnover Minus" for giveaways leading to goals against, but are you going to attempt to track takeaways not credited in the official stats that lead to goals for? For example, forcing a bad pass that ends up on the stick of a Wing who then scores?
Forcing a bad pass that ends up on the stick of a Wing who scores will create a Goal-Scored Plus for the player who forced it and very likely a non-touch assist as well.
by J.J. from Kansas on Oct 4, 2011 1:18 PM CDT up reply actions
Of course I’ll be around to add my oaf-ish feedback for the best statistical project around.
I hate Jonathon Ericsson.
Excellent
You’re one of the feedback MVPs from last season, oaf-ish as you may be.
by J.J. from Kansas on Oct 4, 2011 1:52 PM CDT up reply actions
It would be awesome if other team’s fans joined in on this CSSI thing. There would obviously be differences in some of the subjective measurements, but it would still provide a more situational method of comparing teams.
I don’t want to speak for JJ here because the CSSI is definitely his baby. But the new software being used was written with the capability to support CSSI data entry for all teams and be able to cross reference everything. The new system that is being rolled out is never Red Wings centric. Any features, stats, graphs, or tables you see that apply to the Wings could be used by any other team as well.
well-said
We are definitely interested in the capability of having one (or more) CSSI specialists for each team in the league to get a leaguewide system going to track CSSI for everybody. The site right now is being used from something of a Wings’ perspective because we’re both Wings fans and that’s who we like, but part of the planned future for the project definitely does include the hope that we can build it for all teams so that we can have a rankings system by which we can compare players around the league.
by J.J. from Kansas on Oct 4, 2011 2:46 PM CDT up reply actions
The main issue you’d run into would be calibration, because a lot of what you’re doing is subjective and not purely numbers-driven. However, that’s not to say it can’t be done, but it would take meetings and training so that everyone was on the same page.
One thought to deal with that issue is that opposing reviewers would have to agree on ALL adjustments. Anything that could not be agreed upon would be put up for vote among all the CSSI reviewers. So lets say the Wings play the Ducks and the dipshit from Anaheim tries to give Perry a Plus for punching a baby. JJ could dispute that adjustment and the play would go up for review to the larger community. Eventually, I think a common standard would evolve.
Haha, I like that we both used Corey Perry getting a bullshit adjustment as the example.
by J.J. from Kansas on Oct 4, 2011 3:06 PM CDT up reply actions
That is my biggest concern
I can maintain a comfortable level of objectivity (I feel) because I’m comparing Red Wings to Red Wings. While there are Wings players I like more than others, I feel that I didn’t give anybody any extra breaks which weren’t well-deserved.
However, knowing that there’s somebody on the other side of the scoresheet who is going to be ranking his team’s favorite players is going to affect things. If I see a Ducks’ CSSI scorer give Corey Perry an overall +1 in a game where I think Pavel Datsyuk outplayed him, am I giving Datsyuk +1.5 based solely on that?
We’re thinking of ways to cope with this (including meetings, training, and allowing editors and superusers to collaborate peacefully to come up with good scores all the way around). As the collaborative effort grows, the system will ensure group objectivity.
The thing about the naysaying that it all depends on my opinion of a play and my subjectivity is a cop-out for people to dismiss something without giving it their own look. Overall, my numbers are less subjective than official numbers, because we showed last season that scoring can be done consistently over the course of a season in a way that allows context to enter the equation in a way that official numbers do not.
by J.J. from Kansas on Oct 4, 2011 3:05 PM CDT up reply actions
A "standard" for subjectivity applied
would be tough across multiple inputs – but it would be way cool if other teams were involved. We could look at similar stats across multiple teams – wow, complete analysis paralysis…I might not ever get any work done!!
However it would likely mean involving team(s) that we don’t have major hate for so that the standard could be applied evenly – (so no Hawks, Sharks, Ducks or Penguins I guess….)
That was my assumption (and understanding based on how the website reads).
As far as subjectivity we already have that in a bunch of stats (hits? takeaways?) that are subjective, so I don’t think that the difference in adjustments made between different teams will cause any trouble in comparing the stats after the other teams get used to how to use the adjustments… probably the first season’s worth of stats will look a little funny at times, but after that they will probably be relatively consistent even without agreement between the two CSSI specialists.
That said, if you can get the specialists to do their best to be objective (as objective as possible being fans and all) the CSSI numbers will probably be more useful and consistent faster.
And if nothing else the CSSI stats will still be able to be used to show relative strengths and weaknesses between different players on the same team. So perhaps any graphs or comparisons that occur between teams should concentrate more on relative comparisons instead of absolute number to number comparisons.
So perhaps any graphs or comparisons that occur between teams should concentrate more on relative comparisons instead of absolute number to number comparisons.
Exactly, and that’s worst-case.
This is basically how QualComp and QualTeam work. They’re not horribly useful for comparing people on different teams (although they can be used with caveats), but they’re great for comparing players on one team.
by J.J. from Kansas on Oct 4, 2011 4:56 PM CDT up reply actions
If you wanted to increase the specalist’s workload you could have each one(s) do the CSSI for each team and then you can have more accurate conversion factors between the different team’s numbers. if the sample size gets large enough it could allow you to normalize the ratings for every team.
by sprout42 on Oct 4, 2011 6:00 PM CDT via mobile up reply actions
Hmm, that wasn't worded very clearly
I meant each person does the ratings for both teams during the same game.
by sprout42 on Oct 4, 2011 6:02 PM CDT via mobile up reply actions
WOOT!
i’m excited.
OT: is a BOLD predictions post coming up this week? i’m dying to get mine out there—i promise you they will be bolder than last year’s, all while toe-ing the line between bold and ridiculous.
You wanna tell me that to mah face?!
YET ANOTHER Detroit Red Wings blog.
by uvgt2bkdnme on Oct 4, 2011 6:02 PM CDT via iPhone app reply actions
haha
BOLD predictions on Thursday morning. Get ready.
by J.J. from Kansas on Oct 4, 2011 6:12 PM CDT up reply actions

by 














