By The NumbersHockey Analytics... the Final Frontier. Explore strange new worlds, to seek out new algorithms, to boldly go where no one has gone before.
Total Hockey Rating (THoR): A comprehensive statistical rating of NHL Players
Total Hockey Rating (THoR): A comprehensive statistical rating of NHL Players
At the start of March is the MIT Sloan Sports Analytics conference - basically the Super Bowl of sports analytics conferences.
One of the hockey related research papers is "Total Hockey Rating (THoR): A comprehensive statistical rating of National Hockey League forwards and defensemen based upon all on-ice events" - and is a very interesting look at valuing players.
"Here we present a new comprehensive rating that accounts for other players on the ice will a give player as well as the impact of where a shift starts and of every non-shooting events such as turnovers and hits that occur when a player is on the ice. The impact of each play is determined by the probability that it leads to a goal for a player’s team (or their opponent) in the subsequent 20 seconds. The primary outcome of this work is a reliable methodology that can quantify the impact of players in creating and preventing goals for both forwards and defenseman. We present ratings for the top forwards and defensemen based on all events from the 2010-11 and 2011-12 NHL regular seasons."
It's pretty close to the ratings I posted here a few months ago. Only players that I think are not supposed to be there include: Timonen (I think he is one of the worst dman in the league), Tyutin, Kesler and to an extent Pavelski. All the others look to be correct. It's amazing how undervalued guys like Steen and Kennedy are.
It's pretty close to the ratings I posted here a few months ago. Only players that I think are not supposed to be there include: Timonen (I think he is one of the worst dman in the league), Tyutin, Kesler and to an extent Pavelski. All the others look to be correct. It's amazing how undervalued guys like Steen and Kennedy are.
It's pretty close to the ratings I posted here a few months ago. Only players that I think are not supposed to be there include: Timonen (I think he is one of the worst dman in the league), Tyutin, Kesler and to an extent Pavelski. All the others look to be correct. It's amazing how undervalued guys like Steen and Kennedy are.
It's pretty close to the ratings I posted here a few months ago. Only players that I think are not supposed to be there include: Timonen (I think he is one of the worst dman in the league), Tyutin, Kesler and to an extent Pavelski. All the others look to be correct. It's amazing how undervalued guys like Steen and Kennedy are.
2 Data
For this analysis we use data from the NHL‟s Real Time Scoring System (RTSS).
That explains the funky ratings right there. You can't draw any real conclusions about the validity of the rating sytem when you're feeding it funky data. Well, other than the fact the designers of the system deliberately created a system that would rely on incredibly inconsistent data, something that certainly doesn't boost their credibility.
They did take some actions to attempt to correct for home/away biases and shot locations in certain rinks (MSG), but there is still plenty of biases, inconsistencies, and a lack of standardization for the RTSS across the league. Too much human error and subjectivity in the system currently.
That explains the funky ratings right there. You can't draw any real conclusions about the validity of the rating sytem when you're feeding it funky data. Well, other than the fact the designers of the system deliberately created a system that would rely on incredibly inconsistent data, something that certainly doesn't boost their credibility.
They did take some actions to attempt to correct for home/away biases and shot locations in certain rinks (MSG), but there is still plenty of biases, inconsistencies, and a lack of standardization for the RTSS across the league. Too much human error and subjectivity in the system currently.
This is a common problem in the NHL, however, their system is fairly sound structurally. It will only improve over time as the data input becomes more accurate. It's a starting point.
That explains the funky ratings right there. You can't draw any real conclusions about the validity of the rating sytem when you're feeding it funky data. Well, other than the fact the designers of the system deliberately created a system that would rely on incredibly inconsistent data, something that certainly doesn't boost their credibility.
They did take some actions to attempt to correct for home/away biases and shot locations in certain rinks (MSG), but there is still plenty of biases, inconsistencies, and a lack of standardization for the RTSS across the league. Too much human error and subjectivity in the system currently.
Yup. All this work to create an "objective" number, and yet they base it off incredibly subjective (and inconsistent) source data. If the authors of the study wanted to create something to talk about so that analysts can refine their methods further, perhaps they made a valuable contribution. As a stand-alone study, however, it doesn't look particularly useful.
Maybe I haven't hung around long enough, but tell me, what's wrong with the NHL's real time data?
It's tracked by humans. One guy might see two players bump along the boards and call it a hit, while another won't. Repeat that x10 hits per game x82 games per season and you've got a large discrepancy.
__________________ CanadianHockey________ __ __________Sens, Oilers, and Team Canada
Additionally, games played in specific arenas are way out of whack statistically compared to the rest of the NHL. Look at the shot counts in games played in Dallas, and compare it to the rest of the NHL.
It's tracked by humans. One guy might see two players bump along the boards and call it a hit, while another won't. Repeat that x10 hits per game x82 games per season and you've got a large discrepancy.
T6his happens, to a much lesser degree in baseball say for counting defense at SS.
One guy, can't remember his name actually looks at every play or hit to the every SS to give a more consistent evaluation, at least in his eyes and it's almost universally accepted as being pretty darn good.
It's actually accepted as a better measurement of defense than what MLB uses.
I've said it numerous times Stats has more application and use in baseball and to a lesser degree in football than it does in hockey when players from both teams play offense and defense on the same shift or play, something that doesn't happen in baseball or football.
The very nature of hockey makes it less measurable than the two other sports.
I'm sorry, but if Alex Steen is ranked #1 and Tyler Kennedy is #3... it makes you wonder just how valuable this metric is.
Quote:
Originally Posted by MastuhNinks
If ever there was a time to discredit a study based on the 'eye-test', it's now.
Crap in, crap out. The metric itself may still be useful, but since they're inputting the information from RTSS, then the results aren't going to be trustworthy.
Not a ringing endorsement of a metric when Paul Martin ranks highly in it given the time table, he was quite brutal during that time...he acknowledges it, the coach scratched him, and the GM asked him if he wanted to get out of Pittsburgh because of how bad he was playing...
And then Tyler Kennedy, who has his position incorrectly listed as 'C' when I don't believe he's taken a shift at center since turning pro years ago, is just a worker bee that does well on the forecheck and the cycle...but can't pass and shoots directly into the goalie's chest from anywhere (though I do like him on a cycling, aggressive forecheck team for the third line, he should never be on any sort of "best of" list even if it's just of third liners).
If those players were the only two exceptions, then I'd give it half a look...but it's just a couple of glaring ones in an otherwise odd mish-mash...
