Absolute Player Value and Salaries.

  1. You have chosen to ignore posts from Bookboy007. Show Bookboy007's posts

    Absolute Player Value and Salaries.

    Every time there's thread about guys being overpaid, I wonder if anyone has ever tried to find a way to assess a player's absolute value - what he should be paid based on his talent, effort, impact on games etc.  Anyone ever seen this?  I'd love to see someone try...because it would be a freaking gong show.

    Player salaries are driven by the market for their services.  Period.  The reason guys get paid more than we think they should is that there aren't enough competent, proven players to go around.  It's a sellers' market unless you're a goalie.

    The exception is if the Rangers are involved in any way.  Then salaries are market x2 + $3M.

     
  2. You have chosen to ignore posts from asmaha. Show asmaha's posts

    Re: Absolute Player Value and Salaries.

    There are tons of things out there like this:

    http://www.nhldigest.com/nhl-player-cost-per-point/

    and this was pretty interesting:

    http://web.dsbn.edu.on.ca/~charlotte.morrison/FOV1-000C6097/FOV1-000C6099/NHL%20SALARIES%20-%20High%202%20low%203.doc

    But nothing that brings it ALL together.
     
  3. You have chosen to ignore posts from Fletcher1. Show Fletcher1's posts

    Re: Absolute Player Value and Salaries.

    The only thing I can think of anywhere close is some tables I have seen rating players' $$/Points which is obviously deeply flawed for players that have key responsibilities other than scoring (eg. anyone other than scoring forwards).  It's not the worst way to rate the stars of the game though and gives the interesting perspective of a forward like Tyler Seguin vs. a forward like Scott Gomez.

    You could certainly take it a lot further, although I don't think you'd ever have something with much overall accuracy.

    I'm sure you could build a statistical algorithm with points, hits, blocked shots, TOI, +/-, etc. as a factor of salary.  It would still be a mess but you could probably identify the outliers on either side of high and low value production per $.  

    What is more interesting would be the trends.  I suspect that once you pay anyone over $6 million, you preclude them from being able to score in the top tier on absolute value.  It seems like the % difference between $5 million and $8 million in salary would be much greater than the corresponding difference in point production in those players, with rare exceptions.  So, do you pay someone $8 million a year to get 100 points instead of 70; or do you pay it to get a much safer chance that even in a down year that player will 70 instead of 40?

    So, build the algorithm already Book.
     
  4. You have chosen to ignore posts from 49-North. Show 49-North's posts

    Re: Absolute Player Value and Salaries.

    Back when I was at university, many, many years ago, as part of a stats project, I did a regression analysis on NHL players salaries.  It obviously had limitations, because the variables which determine salaries differ from forwards (scoring or defensive), d-men (scoring or shut-down), and goalies.  However, by using as many variables as I could think of, it was possible to develop a formula which explained a significant portion of a player's salary. 

    It's many years ago, so my memory is quite foggy, but aside from the obvious variables such as goals, total points, etc. there were some which stood out quite strongly, such as experience (I found that games played was stronger than years of service), penalty minutes (within a certain range --denoted aggressiveness, but once the # got too high, the predictive ability of the model broke down), and "I/O" switches such as "All-Star Selection", "National Team Selection", and "NHL Award Winner".
     
  5. You have chosen to ignore posts from SanDogBrewin. Show SanDogBrewin's posts

    Re: Absolute Player Value and Salaries.

    http://www.behindthenet.ca/nhl_statistics.php?ds=30&f1=2011_s&f2=5v5&f7=60-&c=0+1+3+5+4+6+7+8+13+14+29+30+32+33+34+45+46+63+67

    Minimum 60 games then plug in what they make from capgeek. Goals scored and PP points would need to be plugged in as well. Some of the stats on behind the net could be taken off but it does give you a better idea.
     
  6. You have chosen to ignore posts from BadHabitude. Show BadHabitude's posts

    Re: Absolute Player Value and Salaries.

    In Response to Absolute Player Value and Salaries.:
    [QUOTE]Every time there's thread about guys being overpaid, I wonder if anyone has ever tried to find a way to assess a player's absolute value - what he should be paid based on his talent, effort, impact on games etc.  Anyone ever seen this?  I'd love to see someone try...because it would be a freaking gong show. Player salaries are driven by the market for their services.  Period.  The reason guys get paid more than we think they should is that there aren't enough competent, proven players to go around.  It's a sellers' market unless you're a goalie. The exception is if the Rangers are involved in any way.  Then salaries are market x2 + $3M.
    Posted by Bookboy007[/QUOTE]

    I know a professor who does this, but on mostly stats.  Baseball is most of his clientele as they have a billion and one stats on everything.  he applied it to several things, like consumer products weighing which factor was most important to you (speed, cost, features).  I know he's been in on some NHL contracts.

     
  7. You have chosen to ignore posts from red75. Show red75's posts

    Re: Absolute Player Value and Salaries.

    Wasn't this the whole point of that number crunching conference in Boston a couple of months ago? Trying to adapt the Oakland A's evaluation of player value strategy for hockey? Bring sabermetrics into the game?
     
  8. You have chosen to ignore posts from SanDogBrewin. Show SanDogBrewin's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:[QUOTE] Trying to adapt the Oakland A's evaluation of player value strategy for hockey ? Bring sabermetrics into the game ? Posted by red75[/QUOTE]

    Oh man Book just loves talking Billy Beane metrics applied to hockey...
     
  9. You have chosen to ignore posts from red75. Show red75's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:
    [QUOTE]In Response to Re: Absolute Player Value and Salaries. : Oh man Book just loves talking Billy Beane metrics applied to hockey...
    Posted by SanDogBrewin[/QUOTE]

    Well in that case let's give him an aneurysm...

    Core Age (team)

    The sum of each skater's GVT multiplied by age on
    a team, divided by the team GVT. It is a better measure
    of the average age of a team, weighting the best
    players the most.

    Corsi

    A statistic originally invented by Jim Corsi, who was
    the goaltender coach for the Buffalo Sabres. Corsi is
    essentially a plus-minus statistic that measures shot
    attempts. A player receives a plus for any shot attempt
    (on net, missed, or blocked) that his team directs at
    the opponent's net, and a minus for any shot attempt
    against his own net. A proxy for possession.


    Corsi QoC

    A measure of competition quality using Corsi as
    its basis. While a slight improvement on goal-based
    measures, the scale can be incredibly hard to decipher
    based on the quality of teams faced throughout the
    year.

    Corsi Rel QoC

    A measure of competition quality using relative Corsi
    as its basis. It is less luck-driven than QualComp
    and more universal than Corsi QoC because it's based
    on a relative metric. The most statistically sound quality
    of competition metric currently used.

    DGVT

    Defensive Goals Versus Threshold. This stat measures
    a player's worth, in goals, at shot suppression.

    ESP/60

    Even strength points per 60 minutes. A commonly
    used statistic measuring the amount of points a player
    scores per 60 minutes of even strength ice time. For
    the 2010-11 season, anything over 1.62 ESP/60 was
    considered to be scoring at the rate of a top-six forward,
    though 1.8 ESP/60 is the common benchmark.

    ESSV%

    Even strength save percentage. Many analysts prefer
    to measure goaltenders using only their even strength
    shot-stopping since there is less year-to-year variation
    and less luck involved.

    ESTOI

    Even strength time on ice is a measure of the ice
    time a player averages per game in even strength situations.


    Fenwick

    Another possession metric, originally devised by
    Matt Fenwick of the Battle of Alberta blog. Fenwick
    follows the same concept as Corsi, but doesn't include
    blocked shots. Fenwick is considered to have better
    predictive value for future goal differential than Corsi.
    The removal of blocked shots is also valuable since
    blocked shots are a proven skill worthy of being separated.

    GVT (Individual)

    Goals Versus Threshold. Developed by Tom Awad
    of Hockey Prospectus, GVT measures a player's
    worth in comparison to a typical fringe NHL player.
    GVT has two major advantages over most metrics: it's
    measured in goals, which are easily equated to wins,
    and it is capable of comparing players across multiple
    positions and multiple eras. GVT is the summation of
    OGVT, GGVT, DGVT, and SGVT.

    GVT (Team)

    Goals Versus Threshold for a team is equal to the
    team's goal differential. By definition Team GVT
    = sum of Player GVT - Replacement Level (which
    GVT defines as 1.5 goals per game or 123 goals per
    season).

    Net Penalties/60

    The difference between the penalties a players draws
    and the penalties a player takes. This difference is recorded
    as a rate per 60 minutes of ice time. Net Penalties
    per 60 gives insight into which players are giving
    their teams extra power play opportunities. Net Penalties/
    60 tends to be higher for forwards and lower for
    defensemen, given their roles on the ice.

    OGVT

    Offensive Goals Versus Threshold is the portion of
    GVT measuring a player's ability to generate goals.
    OGVT measures a player's contribution to goal creation.


    PDO

    Created by Vic Ferrari, PDO is the sum of a player's
    on-ice save percentage and on-ice shooting percentage.
    PDO is an excellent way to measure "puck luck"
    or good fortune as it regresses heavily to the mean of
    100 (sometimes shown as 1000). For example, a player
    with a PDO of 103.4 is likely to see his luck drop
    next year, affecting his plus-minus or point totals. A
    player with a PDO of 97.1 will likely have a "bounceback"
    year purely by getting a few more bounces go
    his way.

    PPP/60

    A rate stat that measures the number of points a player
    records per 60 minutes of power play time. Even
    strength and power play rates are split out to give
    better insight into which situations a player excels at.
    Much more liable to fluctuate than ESP/60 based on
    much less playing time involved, so it's better to observe
    a player's PPP/60 over several seasons.

    QualComp

    A measure of competition quality using relative
    plus-minus as it's basis. A number higher than 0 indicates
    a higher than average level of competition. One
    drawback to QualComp is that it is goal-based, incorporating
    more puck luck than Corsi-based methods.

    QS

    A Quality Start is a measure of whether a goaltender
    "gave his team a chance to win". In order to record a
    quality start, the goalie must have a save percentage
    over .912 or a save percentage between .885 and .912
    while allowing less than three goals. The reason Quality
    Starts are so important is that teams win 77.5%
    of the games in which their goalie records a Quality
    Start.

    QS%

    Quality Start Percentage shows the ratio of a goaltender's
    games in which they recorded a quality start.
    A quality start rate of 60% is considered elite, while
    anything below 40% is quite poor.


    Regression to the mean

    A statistical term that refers to the phenomena of
    statistical measures returning to their average value
    over time. For example, if Player A's career shooting
    percentage is 8.5% and he shot 11.3% last season, he
    will likely regress closer to 8.5% the following year.

    Relative Corsi

    A player's Corsi value in comparison to his teammates.
    Relative Corsi is expressed as the player's
    Corsi minus the team Corsi rate. A positive value indicates
    a player who is better than the team average and
    a negative number is a player who is worse than the
    team average.

    Replacement level

    The level of performance of an easily obtainable
    player, such as an AHL call-up. Replacement level is
    considered the minimum performance that teams can
    achieve, therefore many statistics (such as GVT) express
    the value of players as value over replacement
    level.

    SGVT

    Shootout Goals Versus Threshold is a measure of
    a skater's or goaltender's value in the shootout. It is
    based on the player's ability to score or prevent goals
    in the shootout versus a league average player in the
    same situation.

    Sheltered minutes

    A term referring to the type of ice time a player is
    given by the coach. Players given "sheltered minutes"
    are generally matched up against easy competition
    and/or given a very high zone start ratio.

    Translations

    Converting a player's scoring data from other
    leagues to an NHL equivalent, using scoring changes
    of those who made the move previously as a basis. For
    example, an AHL translation would be what a player
    would score in the NHL, based on what happened to
    previous AHL players with similar statistics when
    they went to the NHL.

    UFO%

    An advanced faceoff metric, Ultimate Faceoff Percentage
    is calculated from even strength, non-empty
    net situation, road faceoffs only, and adjusted for
    strength of competition.

    WOWY

    With or Without You is a style of analysis used to
    determine which players benefit the most from their
    linemates and which players are driving play.

    Zone Start%

    A ratio showing the percentage of a player's non-
    neutral-zone shifts that were started in the offensive
    zone. Zone starts use faceoffs as a proxy for all shifts.
    Players with a ZS% higher than 54% could be considered
    sheltered or deployed offensively while players
    with ZS% south of 46% can be considered to be deployed
    defensively or doing the "heavy lifting".


    All terms from Hockey Prospectus
     
  10. You have chosen to ignore posts from 50belowzero. Show 50belowzero's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:
    [QUOTE]In Response to Re: Absolute Player Value and Salaries. : Well in that case let's give him an aneurysm... Core Age (team) The sum of each skater's GVT multiplied by age on a team, divided by the team GVT. It is a better measure of the average age of a team, weighting the best players the most. Corsi A statistic originally invented by Jim Corsi, who was the goaltender coach for the Buffalo Sabres. Corsi is essentially a plus-minus statistic that measures shot attempts. A player receives a plus for any shot attempt (on net, missed, or blocked) that his team directs at the opponent's net, and a minus for any shot attempt against his own net. A proxy for possession. Corsi QoC A measure of competition quality using Corsi as its basis. While a slight improvement on goal-based measures, the scale can be incredibly hard to decipher based on the quality of teams faced throughout the year. Corsi Rel QoC A measure of competition quality using relative Corsi as its basis. It is less luck-driven than QualComp and more universal than Corsi QoC because it's based on a relative metric. The most statistically sound quality of competition metric currently used. DGVT Defensive Goals Versus Threshold. This stat measures a player's worth, in goals, at shot suppression. ESP/60 Even strength points per 60 minutes. A commonly used statistic measuring the amount of points a player scores per 60 minutes of even strength ice time. For the 2010-11 season, anything over 1.62 ESP/60 was considered to be scoring at the rate of a top-six forward, though 1.8 ESP/60 is the common benchmark. ESSV% Even strength save percentage. Many analysts prefer to measure goaltenders using only their even strength shot-stopping since there is less year-to-year variation and less luck involved. ESTOI Even strength time on ice is a measure of the ice time a player averages per game in even strength situations. Fenwick Another possession metric, originally devised by Matt Fenwick of the Battle of Alberta blog. Fenwick follows the same concept as Corsi, but doesn't include blocked shots. Fenwick is considered to have better predictive value for future goal differential than Corsi. The removal of blocked shots is also valuable since blocked shots are a proven skill worthy of being separated. GVT (Individual) Goals Versus Threshold. Developed by Tom Awad of Hockey Prospectus, GVT measures a player's worth in comparison to a typical fringe NHL player. GVT has two major advantages over most metrics: it's measured in goals, which are easily equated to wins, and it is capable of comparing players across multiple positions and multiple eras. GVT is the summation of OGVT, GGVT, DGVT, and SGVT. GVT (Team) Goals Versus Threshold for a team is equal to the team's goal differential. By definition Team GVT = sum of Player GVT - Replacement Level (which GVT defines as 1.5 goals per game or 123 goals per season). Net Penalties/60 The difference between the penalties a players draws and the penalties a player takes. This difference is recorded as a rate per 60 minutes of ice time. Net Penalties per 60 gives insight into which players are giving their teams extra power play opportunities. Net Penalties/ 60 tends to be higher for forwards and lower for defensemen, given their roles on the ice. OGVT Offensive Goals Versus Threshold is the portion of GVT measuring a player's ability to generate goals. OGVT measures a player's contribution to goal creation. PDO Created by Vic Ferrari, PDO is the sum of a player's on-ice save percentage and on-ice shooting percentage. PDO is an excellent way to measure "puck luck" or good fortune as it regresses heavily to the mean of 100 (sometimes shown as 1000). For example, a player with a PDO of 103.4 is likely to see his luck drop next year, affecting his plus-minus or point totals. A player with a PDO of 97.1 will likely have a "bounceback" year purely by getting a few more bounces go his way. PPP/60 A rate stat that measures the number of points a player records per 60 minutes of power play time. Even strength and power play rates are split out to give better insight into which situations a player excels at. Much more liable to fluctuate than ESP/60 based on much less playing time involved, so it's better to observe a player's PPP/60 over several seasons. QualComp A measure of competition quality using relative plus-minus as it's basis. A number higher than 0 indicates a higher than average level of competition. One drawback to QualComp is that it is goal-based, incorporating more puck luck than Corsi-based methods. QS A Quality Start is a measure of whether a goaltender "gave his team a chance to win". In order to record a quality start, the goalie must have a save percentage over .912 or a save percentage between .885 and .912 while allowing less than three goals. The reason Quality Starts are so important is that teams win 77.5% of the games in which their goalie records a Quality Start. QS% Quality Start Percentage shows the ratio of a goaltender's games in which they recorded a quality start. A quality start rate of 60% is considered elite, while anything below 40% is quite poor. Regression to the mean A statistical term that refers to the phenomena of statistical measures returning to their average value over time. For example, if Player A's career shooting percentage is 8.5% and he shot 11.3% last season, he will likely regress closer to 8.5% the following year. Relative Corsi A player's Corsi value in comparison to his teammates. Relative Corsi is expressed as the player's Corsi minus the team Corsi rate. A positive value indicates a player who is better than the team average and a negative number is a player who is worse than the team average. Replacement level The level of performance of an easily obtainable player, such as an AHL call-up. Replacement level is considered the minimum performance that teams can achieve, therefore many statistics (such as GVT) express the value of players as value over replacement level. SGVT Shootout Goals Versus Threshold is a measure of a skater's or goaltender's value in the shootout. It is based on the player's ability to score or prevent goals in the shootout versus a league average player in the same situation. Sheltered minutes A term referring to the type of ice time a player is given by the coach. Players given "sheltered minutes" are generally matched up against easy competition and/or given a very high zone start ratio. Translations Converting a player's scoring data from other leagues to an NHL equivalent, using scoring changes of those who made the move previously as a basis. For example, an AHL translation would be what a player would score in the NHL, based on what happened to previous AHL players with similar statistics when they went to the NHL. UFO% An advanced faceoff metric, Ultimate Faceoff Percentage is calculated from even strength, non-empty net situation, road faceoffs only, and adjusted for strength of competition. WOWY With or Without You is a style of analysis used to determine which players benefit the most from their linemates and which players are driving play. Zone Start% A ratio showing the percentage of a player's non- neutral-zone shifts that were started in the offensive zone. Zone starts use faceoffs as a proxy for all shifts. Players with a ZS% higher than 54% could be considered sheltered or deployed offensively while players with ZS% south of 46% can be considered to be deployed defensively or doing the "heavy lifting". All terms from Hockey Prospectus
    Posted by red75[/QUOTE]
    I nodded off at PDO.
     
  11. You have chosen to ignore posts from red75. Show red75's posts

    Re: Absolute Player Value and Salaries.

    And here's Hockey Prospectus rankings of the best players and their relative worth for the 2011-2012 season.
    http://hockeyprospectus.com/sortable/

    Top ranked Bruin is Bergeron, with only 4 forwards - Malkin, Stamkos, Giroux and ovalchuk ranked ahead of him. Seguin is the 7th top forward, with Gaborik between him and Bergeron.
     
  12. This post has been removed.

     
  13. You have chosen to ignore posts from Bookboy007. Show Bookboy007's posts

    Re: Absolute Player Value and Salaries.

    Problem is that all of these measures are meaningless in terms of translating to dollars you pay.  You still need a benchmark salary and any real measure you choose like the mean salary or the median of "fringe players" or whatever - it's still determined by the market. 

    That's where I think Billy Beane was all smoke and mirrors.  He was able to identify statistical measures that weren't highly valued in terms of dollars but that had an impact on the field, so he could chase those guys who weren't likely to get bid up in the market, keep costs down, and compete.  The illusion is that you could win that way; that the factors he was able to identify could outweigh the factors a sack of hammers could figure out and a sack of hammers with a NY wallet could pay the moon for. The market is what it is for a reason; fans may be dumb as rocks (present company excluded) but even they understand that it's worth more as both competitive value and entertainment value to have a Kirk Gibson who you know is a threat to bomb in the 9th to win a Series game even if he doesn have a lousy modified OBP or, for us old-time stats people, a .212 Batting average.

    Get out your irony goggles - some people like to make this sports stats thing waaaaay more complicated than it really is.
     
  14. You have chosen to ignore posts from Not-A-Shot. Show Not-A-Shot's posts

    Re: Absolute Player Value and Salaries.

    Watch game.

    Like player.

    Try to get player.

    Don't hang team in process.

    Repeat.

     
  15. You have chosen to ignore posts from Bookboy007. Show Bookboy007's posts

    Re: Absolute Player Value and Salaries.

    Watch lots of games.  Still like player....  Then go get player.

    There have been too many players over the years who have looked like world-beaters where I've thought, man, put that guy in the top six, play him with some better players, and he'll light it up!  Then you realize you watched the one game that week where he played hard and the one game that month where he had any success.  Call it the Petr Kalus rule (to name one among a million).
     
  16. You have chosen to ignore posts from asmaha. Show asmaha's posts

    Re: Absolute Player Value and Salaries.

    Something to be said about the numbers, but then there's this:

    35 Phil Kessel Toronto Maple Leafs
    47 Shea Weber Nashville Predators
    49 Jamie Benn Dallas Stars
    51 Anze Kopitar Los Angeles Kings
     
  17. You have chosen to ignore posts from LordByron77. Show LordByron77's posts

    Re: Absolute Player Value and Salaries.

     I think it all depends on the "fit" of the team.

    Semin's value as offensive is high, however defensively, not so much.
    Ryder did well with the B's during their cup run. Exceeded many expectations with Dallas!
    Kovalchuk and his ridiculous contract?
    Bergeron is the heart of the B's! That makes his value MUCH higher!

    Do all these guys deserve the moon? No. But with many teams trying to find a player to fit their system and their cap room ... that's the real question
     
  18. You have chosen to ignore posts from dezaruchi. Show dezaruchi's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:
    [QUOTE]Watch lots of games.  Still like player....  Then go get player. There have been too many players over the years who have looked like world-beaters where I've thought, man, put that guy in the top six, play him with some better players, and he'll light it up!  Then you realize you watched the one game that week where he played hard and the one game that month where he had any success.  Call it the Petr Kalus rule (to name one among a million).
    Posted by Bookboy007[/QUOTE]
    Is that the kid who scored on his first 3 shots in the NHL? I remember Boston had a few kids whose names started with a "K" and that one of them scored everytime he got the puck to the net in his first 3 games.
     
  19. You have chosen to ignore posts from SanDogBrewin. Show SanDogBrewin's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:[QUOTE]In Response to Re: Absolute Player Value and Salaries. : Is that the kid who scored on his first 3 shots in the NHL ? I remember Boston had a few kids whose names started with a "K" and that one of them scored everytime he got the puck to the net in his first 3 games.Posted by dezaruchi[/QUOTE]

    Dmitri Kvartalnov ?
     
  20. You have chosen to ignore posts from Bookboy007. Show Bookboy007's posts

    Re: Absolute Player Value and Salaries.

    Kalus, Krejci, Karsums.  They kept the right one, but yes, Kalus had a golden touch for his first few games, which was the worst thing for him because he was, as I recall, a bit of a head case.
     
  21. You have chosen to ignore posts from kelvana33. Show kelvana33's posts

    Re: Absolute Player Value and Salaries.

    In Response to Re: Absolute Player Value and Salaries.:
    [QUOTE]In Response to Re: Absolute Player Value and Salaries. : Is that the kid who scored on his first 3 shots in the NHL? I remember Boston had a few kids whose names started with a "K" and that one of them scored everytime he got the puck to the net in his first 3 games.
    Posted by dezaruchi[/QUOTE]

    It has to be him, cause thats the first thing that came to my mind as I read the name. I think he was the one traded for Fernandez.
     
  22. You have chosen to ignore posts from islamorada. Show islamorada's posts

    Re: Absolute Player Value and Salaries.

    Was it Kalus for Manny's fanny?
     

Share