Apr 252013
 

I am hoping to get playoff stats on stats.hockeyanalysis.com but it is going to take some work especially if I am to do game by game and series by series stats including “with you” and “against you” stats. As such I have decided to start a crowd funding project at RocketHub.com (because unlike Kickstarter they support Canadians) to help justify the time I will have to put in to getting these stats up in a relatively short time frame. Below is the description of the project and what I hope to achieve and if you are interested in contributing you can do that at the project page at RocketHub.com. Your contributions are greatly appreciated and I think you will enjoy what I have planned for stats.hockeyanalysis.com.

—————————————

Hello. This is David Johnson from HockeyAnalysis.com and creator of the popular advanced hockey statistics website Stats.HockeyAnalysis.com. Much of my work on hockey analytics has been at the macro level, or more specifically evaluating players over 1, 2, or more years. This works great for the regular season and for evaluating a players overall talent level which is where my interest mostly lies but there seems to be a strong demand for more micro level stats such as how players or teams perform in a single game or over a short stretch of games (i.e. after the trade deadline, before and after a coach got replaced, etc.) and this is especially true during the Stanley Cup playoffs.

The problem is, much of my existing code base that I use for stats.hockeyanalysis.com is designed for macro level stats and to revamp it to calculate stats on a per game or per playoff series basis and make these available on the web will take a significant redesign and rewrite of large portions of the code.

My goal for this project is to make some of those changes so I can get some playoff stats up for those that are interested and down the road make per game and per groups of game data available for regular season data starting next season. Here is what I am hoping to generate for these playoffs:

  • Team stats by series and playoffs overall
  • Player stats by series and playoffs overall
  • Game by game team stats
  • “with you” stats by game, series and playoffs overall so you can see how the team performed with various pairs of players on the ice.
  • “against you” stats by game, series and playoffs so you can see which players were successful at scoring on or shutting down their opponents.
  • For each of the above I will be adding goal, shot, fenwick and corsi data (totals and possibly %’s).
  • Will add zone start data to “with you” and “against you” data as time permits.
  • Will start with just looking at 5v5 situations but will add other situations if time permits.

My intent is to start by adding playoff stats similar to the existing regular season stats and then as development progresses I’ll be adding the other features with hopefully the majority of them being added by the end of round 1 if not sooner.

I am looking for some funding so I can justify the significant time over the next few weeks that it will take to rewrite my code and make game by game playoff stats available. I figure if each of the regulars that use stats.hockeyanalysis.com contributes between $10 and $50 (larger donations certainly welcome though) it will be easy to reach my funding goal. Any additional funding beyond my goal will be devoted towards adding similar game by game features to the regular season data for the start of next season.

Apr 192013
 

Tyler Dellow has a post at mc79hockey.com looking at zone starts and defensemen and if you read it the clear conclusion is that zone starts seem to matter quite a bit. In the third chart you can see that defensemen who get the most extreme defensive zone starts have an average corsi% of 44.7% while the average corsi% for defensemen with the most extreme offensive zone starts is 53.3%. This would seem to indicate that for defensemen zone starts can impact your corsi% anywhere from -5.3% to +3.3%. This is far more significant than I have estimated myself using a different methodology so I pondered that part of the reason for this is that when you start in the defensive zone you are playing with weaker quality of teammates than when you start in the offensive zone. My reasoning is that players that get used primarily in the defensive zone are often weak offensive players as if you are a good offensive player you will be given offensive opportunities. I wanted to explore this concept further and that is what I present to you here.

Unlike Tyler Dellow I used forwards in my analysis but it is unlikely that this will have a major impact in the analysis as forwards and defensemen are always on the ice together. One difference between my analysis and Tyler Dellow’s is I used data from stats.hockeyanalysis.com where as Tyler used stats from behindthenet.ca. Behindthenet.ca includes goalie pulled situations in their data and this has the potential to greatly emphasize the impact of zone starts. I feel it is important to eliminate this factor so I have it removed from the data. I also only used 2011-12 data but that shouldn’t have a major impact on the results.

So, my theory is that players who start in the defensive zone are weaker players overall. The challenge to this is that players who start with players that start frequently in the defensive zone likely start frequently in the defensive zone themselves and thus their stats are subject to zone start effects so if they have weak stats we don’t know whether they are due to the zone starts or because they are weak players. My solution was to look at the players zone start adjusted stats that I have on stats.hockeyanalysis.com. These stats ignore the first 10 seconds after a zone face off as it has been shown that the majority of the benefit/penalty of a zone face off has largely dissipated after 10 seconds. I understand that it may seem weird to use zone start adjusted data in a study that attempts to estimate the impact of zone starts but I don’t know what else to do.

I want to also point out that I will be using ZS adjusted FF% team mates when the team mates are not on the ice with the player and this may also mitigate the ZS impact on the teammates stats. My reasoning is, if a player has an extensive number of defensvie zone starts, it is quite possible that when his team mates are not playing with him their zone starts are more neutral or maybe even offensive zone biased. It if there ever was a way to get a non-zone start impacted FF% to use as a QoT metric this is probably the best we can do.

Ok, so what I did was compare a players 5v5 FF% (fenwick %) and zone start adjusted 5v5 TMFF% (zone start adjusted FF% of teammates when team mates are not playing with him) and came up with the following:

FFPct_vs_TMFFPct_by_ZS

As you can see, TMFF% does seem to vary across zone start profiles as I had hypothesized though to a lesser extent than the players zone start influenced FF% which is to be expected. So, if we subtract TMFF% from FF% we get the following chart:

FFPct-TMFFPct_by_ZS

This chart indicates that the zone start impact on forwards once adjusted for quality of teammates (as best we can) ranges from -2.5% to +2.15% which is significantly lower than the -5.3% to +3.3% estimate that Tyler Dellow came up with for defensemen without adjusting for quality of teammates and using goalie pulled situations included in the data. That said, this is still more significant than my own estimates when I compared 5v5 data to 5v5 data with the first 10 seconds after a zone start ignored. When I did that I calculated the impact on H. Sedin’s FF% due to his heavy offensive zone starts to be +1.4% to his FF% and considered this an upper bound. To investigate this further I plotted the average difference between 5v5 FF% and my 5v5 zone start adjusted FF% and I get the following:

FFPct-ZSAdjFFPct_by_ZS

The above is an estimate of the average impact of zone starts using my zone start adjustment methodology which ignores the first 10 seconds after a zone face off. This is significantly lower than either of the previous 2 estimates as we can see in this summary table:

Methodology ZS Impact Estimate
T. Dellow’s estimate for defensemen -5.3% to +3.3%
My TM Adjusted estimate for forwards -2.5% to +2.15%
My 10 second after Zone FO adjustment for forwards -0.5% to +0.41%

I am pretty sure none of what I have said above will put an end to the impact of zone starts on a players statistics debate but at the very least I hope it sheds some light on some of the issues involved. For me personally, I have the most confidence in my zone start adjustment method which removes the 10 seconds after a zone face off. My reasoning is studies have shown that the effect of a zone face off is largely eliminated within the first 10 seconds (see here or here) and also because it is the only methodology that compares a player to himself under similar playing conditions (i.e. same season, almost identical QoT, QoC and situation profiles) eliminating most of the opportunity for confounding factors to influence the results. If this is the case, the impact of zone starts on a players stats is fairly small to the point of being almost negligible for the majority of players.

 

Apr 122013
 

Even though I think the idea of ‘usage’ and ‘tough minutes’ is a vastly over stated factor in an individual players statistics they are interesting to look at as it gives us an indication of how a coach views the player. So for all the usage fans, here is another usage statistic which I will call the Leading-Trailing Index, or LT Index for short.

LT Index = TOI% when leading / TOI% when trailing

where TOI% is the percentage of the teams overall ice time (in games that the player played in) that the player is on the ice (so a 5v5 TOI% of 20% means the player was on the ice for 20% of the time that the team was at 5v5). Thus, the LT index is a ratio of the players ice time when his team is leading to his ice time when his team is trailing adjusted for the overall ice time that the team is leading/trailing. Any number greater than 1.00 indicates the player gets a greater share of ice time when the team is leading and anything under 1.00 indicates the player gets a greater share of ice time when the team is trailing.  So, any players with an LT index greater than one is used more as a defensive player than an offensive one and anything less than one they are used more as an offensive player than a defensive one. Any player around 1.00 is a well balanced player. So, looking at this seasons data we have the following player usage:

Defensive Usage

Defenseman LT Index Forward LT Index
MICHAEL STONE 1.21 BJ CROMBEEN 1.66
KEITH AULIE 1.20 MATHIEU PERREAULT 1.45
RYAN MCDONAGH 1.19 CRAIG ADAMS 1.35
PAUL MARTIN 1.16 TRAVIS MOEN 1.33
BRYCE SALVADOR 1.15 BOYD GORDON 1.26
BRENDAN SMITH 1.14 JAMES WRIGHT 1.26
SCOTT HANNAN 1.14 MICHAEL FROLIK 1.23
ANDREJ SEKERA 1.13 BRIAN BOYLE 1.22
MIKE WEBER 1.13 MATT CALVERT 1.22
JUSTIN BRAUN 1.12 TANNER GLASS 1.20
BARRET JACKMAN 1.12 MATT MARTIN 1.19
ROBYN REGEHR 1.12 RUSLAN FEDOTENKO 1.19
CLAYTON STONER 1.12 STEPHEN GIONTA 1.19
ANTON VOLCHENKOV 1.11 CASEY CIZIKAS 1.19
RON HAINSEY 1.11 JEFF HALPERN 1.18
TIM GLEASON 1.11 DAVID JONES 1.17
ROSTISLAV KLESLA 1.11 NIKOLAI KULEMIN 1.17
ROB SCUDERI 1.10 ZACK KASSIAN 1.17
NIKLAS HJALMARSSON 1.10 RYAN CARTER 1.16
NICKLAS GROSSMANN 1.10 TORREY MITCHELL 1.16

Offensive Usage

Defenseman LT Index Forward LT Index
RYAN ELLIS 0.78 DEREK DORSETT 0.77
KRIS LETANG 0.84 RAFFI TORRES 0.77
MARK STREIT 0.86 TAYLOR HALL 0.79
KYLE QUINCEY 0.87 CORY CONACHER 0.79
MATT NISKANEN 0.87 JORDAN EBERLE 0.80
JUSTIN SCHULTZ 0.87 NAIL YAKUPOV 0.82
DOUGIE HAMILTON 0.88 RYAN NUGENT-HOPKINS 0.82
VICTOR HEDMAN 0.88 RICH CLUNE 0.82
DAN BOYLE 0.89 BLAKE COMEAU 0.82
KEVIN SHATTENKIRK 0.89 KYLE PALMIERI 0.84
ALEX PIETRANGELO 0.89 BRENDAN GALLAGHER 0.84
JOHN-MICHAEL LILES 0.90 CLAUDE GIROUX 0.86
JOHN CARLSON 0.90 VINCENT LECAVALIER 0.86
P.K. SUBBAN 0.90 DREW SHORE 0.86
LUBOMIR VISNOVSKY 0.91 TJ OSHIE 0.87
CODY FRANSON 0.91 ALEX OVECHKIN 0.87
JAMIE MCBAIN 0.91 JONATHAN HUBERDEAU 0.87
ROMAN JOSI 0.92 NICKLAS BACKSTROM 0.87
JARED SPURGEON 0.93 SCOTT HARTNELL 0.87
CHRISTIAN EHRHOFF 0.93 MARIAN HOSSA 0.87

Balanced Usage

Defenseman LT Index Forward LT Index
MICHAEL DEL_ZOTTO 0.99 BRYAN LITTLE 0.99
ERIC BREWER 0.99 MIKE FISHER 0.99
JAKUB KINDL 0.99 MIKKEL BOEDKER 0.99
ADRIAN AUCOIN 0.99 ALEXEI PONIKAROVSKY 0.99
ALEX GOLIGOSKI 0.99 JASON POMINVILLE 0.99
ERIK GUDBRANSON 1.00 CHRIS STEWART 0.99
DREW DOUGHTY 1.00 DANIEL BRIERE 1.00
THOMAS HICKEY 1.00 RADIM VRBATA 1.00
JOHNNY ODUYA 1.00 ALEX TANGUAY 1.00
SLAVA VOYNOV 1.00 GABRIEL LANDESKOG 1.00
MATT IRWIN 1.00 JIRI TLUSTY 1.00
FRANCIS BOUILLON 1.01 COLIN WILSON 1.00
JONAS BRODIN 1.01 PATRICK DWYER 1.00
BRENT SEABROOK 1.01 JADEN SCHWARTZ 1.01
JOSH GORGES 1.01 BRANDON SAAD 1.01
DUSTIN BYFUGLIEN 1.01 LEO KOMAROV 1.01
BRENDEN DILLON 1.01 DREW MILLER 1.01
GREG ZANON 1.01 DAVID PERRON 1.01
KRIS RUSSELL 1.02 TOM PYATT 1.01

It’s amazing how much more BJ Crombeem gets used protecting a lead than when trailing. You’d have to think that score effects could have a significant impact on his stats because of this. Not really a lot of surprises there though though in the case of a guy like Derek Dorsett him being in the ‘offensive usage’ category has more with the coach not wanting to use him defending a lead than hoping he will score a goal to get the team back in the game.

 

Apr 122013
 

Now that I have added home and road stats to stats.hockeyanalysis.com I can take a look at how quality of competition differs when the team is at home vs when they are on the road. In theory because the home team has last change they should be able to dictate the match ups better and thus should be able to drive QoC a bit better. Let’s take a look at the top 10 defensemen in HARO QoC last season at home and on the road (defensemen with 400 5v5 home/road minutes were considered).

Player Name Home HARO QOC Player Name Road HARO QOC
GIRARDI, DAN 8.81 MCDONAGH, RYAN 6.73
MCDONAGH, RYAN 8.49 GORGES, JOSH 6.48
PHANEUF, DION 8.46 GIRARDI, DAN 6.03
GARRISON, JASON 8.27 SUBBAN, P.K. 5.95
GORGES, JOSH 8.25 PHANEUF, DION 5.94
GLEASON, TIM 8.21 GUNNARSSON, CARL 5.48
SUBBAN, P.K. 8.19 ALZNER, KARL 5.35
WEAVER, MIKE 7.92 STAIOS, STEVE 5.15
ALZNER, KARL 7.74 TIMONEN, KIMMO 4.95
REGEHR, ROBYN 7.72 WEAVER, MIKE 4.67

There is definitely a lot of common names in each list but we do notice that the HARO QoC is greater at home than on the road for these defensemen. Next I took a look at the standard deviation of all the defensemen with 400 5v5 home/road minutes last season which should give us an indication of how much QoC varies from player to player.

StdDev
Home 3.29
Road 2.45

The standard deviation is 34% higher at home than on the road which again confirms that variation in QoC are greater at home than on the road.  All of this makes perfect sense but it is nice to see it backed up in actual numbers.

 

 

Apr 112013
 

Stats.hockeyanalysis.com has just gotten even better! Several people have asked why I have zone start adjusted stats for team stats and it is a good question. The answer to that is that it was just easier from a programming point of view to have the same ‘situations’ for both the player level and the team level and since I was already calculating, for example, 5v5close zone start adjusted data for players it was east to add 5v5close zone start adjusted data for teams. Since it makes sense to have non-zone start adjusted data for teams it was on my todo list to get it implemented. So, now it is done, and so much more. The situations that you can access data for at both the player and team level are:

  • 5v5
  • 5v5 home
  • 5v5 road
  • 5v5 close
  • 5v5 tied
  • 5v5 leading
  • 5v5 trailing
  • 5v5 up 1 goal
  • 5v5 up 2+ goals
  • 5v5 down 1 goal
  • 5v5 down 2+ goals
  • 5v4 PP
  • 4v5 PK

In addition to all of the above, all of the above are also available in their Zone Adjusted forms except for the 5v4 PP and 4v5 PK situations. In total, there are now 24 different situations you can search for stats on.  Have at it and don’t blame me for any lost weekends (or lost productivity at work).

(As usual, if you find any issues with the new data please let me know. The stats should be correct but while I have done some testing on the new code to display the stats but it isn’t completely tested.)

 

Apr 112013
 

Every now and again someone asks me how I calculate HARO, HARD and HART ratings that you can find on stats.hockeyanalysis.com and it is at that point I realize that I don’t have an up to date description of how they are calculated so today I endeavor to write one.

First, let me define HARO, HARD and HART.

HARO – Hockey Analysis Rating Offense
HARD – Hockey Analysis Rating Defense
HART – Hockey Analysis Rating Total

So my goal when creating then was to create an offensive defensive and overall total rating for each and every player. Now, here is a step by step guide as to how they are calculated.

Calculate WOWY’s and AYNAY’s

The first step is to calculate WOWY’s (With Or Without You) and AYNAY’s (Against You or Not Against You). You can find goal and corsi WOWY’s and AYNAY’s on stats.hockeyanalysis.com for every player for 5v5, 5v5 ZS adjusted and 5v5 close zone start adjusted situations but I calculate them for every situation you see on stats.hockeyanalysis.com and for shots and fenwick as well but they don’t get posted because it amounts to a massive amounts of data.

(Distraction: 800 players playing against 800 other players means 640,000 data points for each TOI, GF20, GA20, SF20, SA20, FF20, FA20, CF20, CA20 when players are playing against each other and separate of each other per season and situation, or about 17.28 million data points for AYNAY’s for a single season per situation. Now consider when I do my 5 year ratings there are more like 1600 players generating more than 60 million datapoints.)

Calculate TMGF20, TMGA20, OppGF20, OppGA20

What we need the WOWY’s for is to calculate TMGF20 (a TOI with weighted average GF20 of the players teammates when his team mates are not playing with him), TMGA20 (a TOI with weighted average GA20 of the players teammates when his team mates are not playing with him), OppGF20 (a TOI against weighted average GF20 of the players opponents when his opponents are not playing against him) and OppGA20 (a TOI against weighted average GA20 of the players opponents when his opponents are not playing against him).

So, let’s take a look at Alexander Steen’s 5v5 WOWY’s for 2011-12 to look at how TMGF20 is calculated. The columns we are interested in are the Teammate when apart TOI and GF20 columns which I will call TWA_TOI and TWA_GF20. TMGF20 is simply a TWA_TOI (teammate while apart time on ice) weighted average of TWA_GF20. This gives us a good indication of how Steen’s teammates perform offensively when they are not playing with Steen.

TMGA20 is calculated the same way but using TWA_GA20 instead of TWA_GF20. OppGF20 is calculated in a similar manner except using OWA_GF20 (Opponent while apart GF20) and OWA_TOI while OppGA20 uses OWA_GA20.

The reason why I use while not playing with/against data is because I don’t want to have the talent level of the player we are evaluating influencing his own QoT and QoC metrics (which is essentially what TMGF20, TMGA20, OppGF20, OppGA20 are).

Calculate first iteration of HARO and HARD

The first iteration of HARO and HARD are simple. I first calculate an estimated GF20 and an estimated GA20 based on the players teammates and opposition.

ExpGF20 = (TMGF20 + OppGA20)/2
ExpGA20 = (TMGA20 + OppGF20)/2

Then I calculate HARO and HARD as a percentage improvement:

HARO(1st iteration) = 100*(GF20-ExpGF20) / ExpGF20
HARD(1st iteration) = 100*(ExpGA20 – GA20) / ExpGA20

So, a HARO of 20 would mean that when the player is on the goal rate of his team is 20% higher than one would expect based on how his teammates and opponents performed during time when the player is not on the ice with/against them. Similarly, a HARD of 20 would mean the goals against rate of his team is 20% better (lower) than expected.

(Note: The OppGA20 that gets used is from the complimentary situation. For 5v5 this means the opposition situation is also 5v5 but when calculating a rating for 5v5 leading the opposition situation is 5v5 trailing so OppGF20 would be OppGF20 calculated from 5v5 trailing data).

Now for a second iteration

The first iteration used GF20 and GA20 stats which is a good start but after the first iteration we have teammate and opponent corrected evaluations of every player which means we have better data about the quality of teammates and opponents the player has. This is where things get a little more complicated because I need to calculate a QoT and QoC metric based on the first iteration HARO and HARD values and then I need to convert that into a GF20 and GA20 equivalent number so I can compare the players GF20 and GA20 to.

To do this I calculate a TMHARO rating which is a TWA_TOI weighted average of first iteration HARO. TMHARD and OppHARO and OppHARD are calculated in a similar manner. TMHARD, OppHARO and OppHARD are similarly calculated. Now I need to convert these to GF20 and GA20 based stats so I do that by multiplying by league average GF20 (LAGF20) and league average GA20 (LAGA20) and from here I can calculated expected GF20 and expected GA20.

ExpGF20(2nd iteration) = (TMHARO*LAGF20 + OppHARD*LAGA20)/2
ExpGA20(2nd iteration) = (TMHARD*LAGA20 + OppHARD*LAGF20)/2

From there we can get a second iteration of HARO and HARD.

HARO(2nd iteration) = 100*(GF20-ExpGF20) / ExpGF20
HARD(2nd iteration) = 100*(ExpGA20 – GA20) / ExpGA20

Now we iterate again and again…

Now we repeat the above step over and over again using the previous iterations HARO and HARD values at every step.

Now calculate HART

Once we have done enough iterations we can calculate HART from the final iterations HARO and HARD values.

HART = (HARO + HARD) /2

Now do the same for Shot, Fenwick and Corsi data

The above is for goal ratings but I have Shot, Fenwick and Corsi ratings as well and these can be calculated in the exact same way except using SF20, SA20, FF20, FA20, CF20 and CA20.

What about goalies?

Goalies are a little unique in that they only really play the defensive side of the game. For this reason I do not include goalies in calculating TMGF20 and OppGF20. For shot, fenwick and corsi I do not include the goalies on the defensive side of things either as I assume a goalie will not influence shots against (though this may not be entirely true as some goalies may be better at controlling rebounds and thus secondary shots but I’ll assume this is a minimal effect if it does exist). The result of this is goalies do have a HARD rating but no HARO, or shot/fenwick/corsi based HARD or HARO rating.

I hope this helps explain how my hockey analysis ratings are calculated but if you have any followup questions feel free to ask them in the comments.

 

Apr 052013
 

I often get asked questions about hockey analytics, hockey fancy stats, how to use them, what they mean, etc. and there are plenty of good places to find definitions of various hockey stats but sometimes what is more important than a definition is some guidelines on how to use them. So, with that said, here are several tips that I have for people using advanced hockey stats.

Don’t over value Quality of Competition

I don’t know how often I’ll point out one players poor stats or another players good stats and immediately get the response “Yeah, but he always plays against the opponents best players” or “Yeah, but he doesn’t play against the oppositions best players” but most people that say that kind of thing have no real idea how much quality of opponent will affect the players statistics. The truth is it is not nearly as much as you might think.  Despite some coaches desperately trying to employ line matching techniques the variation in quality of competition metric is dwarfed by variation in quality of teammates, individual talent, and on-ice results. An analysis of Pavel Datsyuk and Valterri Filppula showed that if Filppula had Datsyuk’s quality of competition his CorsiFor% would drop from 51.05% to 50.90% and his GoalsFor% would drop from 55.65% to 55.02%. In the grand scheme of things, this are relatively minor factors.

Don’t over value Zone Stats either

Like quality of competition, many people will use zone starts to justify a players good/poor statistics. The truth is zone starts are not a significant factor either. I have found that the effect of zone starts is largely eliminated after about 10 seconds after a face off and this has been found true by others as well. I account for zone starts in statistics by eliminating the 10 seconds after an offensive or defensive zone face off and I have found doing this has relatively little effect on a players stats. Henrik Sedin is maybe the most extreme case of a player getting primarily offensive zone starts and all those zone starts took him from a 55.2 fenwick% player to a 53.8% fenwick% player when zone starts are factored out. In the most extreme case there is only a 1.5% impact on a players fenwick% and the majority of players are no where close to the zone start bias of Henrik Sedin. For the majority of players you are probably talking something under 0.5% impact on their fenwick%. As for individual stats over the last 3 seasons H. Sedin had 34 goals and 172 points in 5v5 situations and just 2 goals and 14 points came within 10 seconds of a zone face off, or about 5 points a year. If instead of 70% offensive zone face off deployment he had 50% offensive zone face off deployment instead of having 14 points during the 10 second zone face off time he may have had 10.  That’s a 4 point differential over 3 years for a guy who scored 172 points. In simple terms, about 2.3% of H. Sedin’s 5v5 points can be attributed to his offensive zone start bias.

A derivative of this is that if zone starts don’t matter much, a players face off winning percentage probably doesn’t matter much either which is consistent with other studies. It’s a nice skill to have, but not worth a lot either.

Do not ignore Quality of Teammates

I have just told you to pretty much ignore quality of competition and zone starts, what about quality of teammates? Well, to put it simply, do not ignore them. Quality of teammates matters and matters a lot. Sticking with the Vancouver Canucks, lets use Alex Burrows as an example. Burrows mostly plays with the Sedin twins but has played on Kesler’s line a bit too. Over the past 3 seasons he has played about 77.9% of his ice time with H. Sedin and about 12.3% of his ice time with Ryan Kesler and the reminder with Malhotra and others. Burrow’s offensive production is significantly better when playing with H. Sedin as 88.7% of his goals and 87.2% of his points came during the 77.9% ice time he played with H. Sedin. If Burrows played 100% of his ice time with H. Sedin and produced at the same rate he would have scored 6 (9.7%) more goals and 13 (11%) more 5v5 points over the past 3 seasons. This is far more significant than the 2.3% boost H. Sedin saw from all his offensive zone starts and I am not certain my Burrows example is the most extreme example in the NHL. How many more points would an average 3rd line get if they played mostly with H. Sedin instead of the average 3rd liner. Who you play with matters a lot. You can’t look at Tyler Bozak’s decent point totals and conclude he is a decent player without considering he plays a lot with Kessel and Lupul, two very good offensive players.

Opportunity is not talent

Kind of along the same lines as the Quality of Teammates discussion, we must be careful not to confuse opportunity and results. Over the past 2 seasons Corey Perry has the second most goals of any forward in the NHL trailing only Steven Stamkos. That might seem impressive but it is a little less so when you consider Perry also had the 4th most 5v5 minutes during that time and the 11th most 5v4 minutes.  Perry is a good goal scorer but a lot of his goals come from opportunity (ice time) as much as individual talent. Among forwards with at least 1500 minutes of 5v5 ice time the past 2 seasons, Perry ranks just 30th in goals per 60 minutes of ice time. That’s still good, but far less impressive than second only to Steven Stamkos and he is actually well behind teammate Bobby Ryan (6th) in this metric. Perry is a very good player but he benefits more than others by getting a lot of ice time  and PP ice time. Perry’s goal production is a large part talent, but also somewhat opportunity driven and we need to keep this in perspective.

Don’t ignore the percentages (shooting and save)

The percentages matter, particularly shooting percentages. I have shown that players can sustain elevated on-ice shooting percentages and I have shown that players can have an impact on their line mates shooting percentages and Tom Awad has shown that a significant portion of the difference between good players and bad players is finishing ability (shooting percentage).  There is even evidence that goal based metrics (which incorporate the percentages) are a better predictor of post season success than fenwick based metric. What corsi/fenwick metrics have going for them is more reliability over small sample sizes but once you approach a full seasons worth of data that benefit is largely gone and you get more benefit from having the percentages factored into the equation. If you want to get a better understanding of what considering the percentages can do for you, try to do a Malkin vs Gomez comparison or a Crosby vs Tyler Kennedy comparison over the past several years. Gomez and Kennedy actually look like relatively decent comparisons if you just consider shot based metrics, but both are terrible percentage players while Malkin and Crosby are excellent percentage players and it is the percentages that make Malkin and Crosby so special. This is an extreme example but the percentages should not be ignored if you want a true representation of a players abilities.

More is definitely better

One of the reason many people have jumped on the shot attempt/corsi/fenwick band wagon is because they are more frequent events than goals and thus give you more reliable metrics. This is true over small sample sizes but as explained above, the percentages matter too and should not be ignored. Luckily, for most players we have ample data to get past the sample size issues. There is no reason to evaluate a player based on half a seasons data if that player has been in the league for several years. Look at 2, 3, 4 years of data.  Look for trends. Is the player consistently a higher corsi player? Is the player consistently a high shooting percentage player? Is the player improving? Declining? I have shown on numerous occassions that goals are a better predictor of future goal rates than corsi/fenwick starting at about one year of data but multiple years are definitely better. Any conclusion about a players talent level using a single season of data or less (regardless of whether it is corsi or goal based) is subject to a significant level of uncertainty. We have multiple years of data for the majority of players so use it. I even aggregate multiple years into one data set for you on stats.hockeyanalysis.com for you so it isn’t even time consuming. The data is there, use it. More is definitely better.

WOWY’s are where it is at

In my mind WOWY’s are the best tool for advanced player evaluation. WOWY stands for with or without you and looks at how a player performs while on the ice with a team mate and while on the ice without a team mate. What WOWY’s can tell you is whether a particular player is a core player driving team success or a player along for the ride. Players that consistently make their team mates statistics better when they are on the ice with them are the players you want on your team. Anze Kopitar is an example of a player who consistently makes his teammates better. Jack Johnson is an example of a player that does not, particularly when looking at goal based metrics.   Then there are a large number of players that are good players that neither drive your teams success nor hold it back, or as I like to say, complementary players. Ideally you build your team around a core of players like Kopitar that will drive success and fill it in with a group of complementary players and quickly rid yourself of players like Jack Johnson that act as drags on the team.

 

Apr 052013
 

Yesterday HabsEyesOnThePrize.com had a post on the importance of fenwick come playoff time over the past 5 seasons. It is definitely worth a look so go check it out. In the post they look at FF% in 5v5close situations and see how well it translates into post season success. I wanted to take this a step further and take a look at PDO and GF% in 5v5close situations to see of they translate into post season success as well.  Here is what I found:

Group N Avg Playoff Avg Cup Winners Lost Cup Finals Lost Third Round Lost Second Round Lost First Round Missed Playoffs
GF% > 55 19 2.68 2.83 5 1 2 6 4 1
GF% 50-55 59 1.22 1.64 0 2 6 10 26 15
GF% 45-50 52 0.62 1.78 0 2 2 4 10 34
GF% <45 20 0.00 - 0 0 0 0 0 20
FF% > 53 23 2.35 2.35 3 2 4 5 9 0
FF% 50-53 55 1.15 1.70 2 2 1 10 22 18
FF% 47-50 46 0.52 1.85 0 0 4 3 6 33
FF% <47 26 0.54 2.00 0 1 1 2 3 19
PDO >1010 27 1.63 2.20 2 2 2 6 8 7
PDO 1000-1010 42 1.17 1.75 1 0 5 7 15 14
PDO 990-1000 47 0.91 1.95 2 1 3 4 12 25
PDO <990 34 0.56 1.90 0 2 0 3 5 24

I have grouped GF%, FF% and PDO into four categories each, the very good, the good, the mediocre and the bad and I have looked at how many teams made it to each round of the playoffs from each group. If we say that winning the cup is worth 5 points, getting to the finals is worth 4, getting to the 3rd round is worth 3, getting to the second round is worth 2, and making the playoffs is worth 1, then the Avg column is the average point total for the teams in that grouping.  The Playoff Avg is the average point total for teams that made the playoffs.

As HabsEyesOnThePrize.com found, 5v5close FF% is definitely an important factor in making the playoffs and enjoying success in the playoffs. That said, GF% seems to be slightly more significant. All 5 Stanley Cup winners came from the GF%>55 group while only 3 cup winners came from the FF%>53 group and both Avg and PlayoffAvg are higher in the GF%>55 group than the FF%>53 group. PDO only seems marginally important, though teams that have a very good PDO do have a slightly better chance to go deeper into the playoffs. Generally speaking though, if you are trying to predict a Stanley Cup winner, looking at 5v5close GF% is probably a better metric than looking at 5v5close FF% and certainly better than PDO. Now, considering this is a significantly shorter season than usual, this may not be the case as luck may be a bit more of a factor in GF% than usual but historically this has been the case.

So, who should we look at for playoff success this season?  Well, there are currently 9 teams with a 5v5close GF% > 55.  Those are Anaheim, Boston, Pittsburgh, Los Angeles, Montreal, Chicago, San Jose, Toronto and Vancouver. No other teams are above 52.3% so that is a list unlikely to get any new additions to it before seasons end though some could certainly fall out of the above 55% list. Now if we also only consider teams that have a 5v5close FF% >50% then Toronto and Anaheim drop off the list leaving you with Boston, Pittsburgh, Los Angeles, Montreal, Chicago, San Jose and Vancouver as your Stanley Cup favourites, but we all pretty much knew that already didn’t we?

 

Apr 012013
 

I have been on a bit of a mission recently to push the idea that quality of competition (and zone starts) is not a huge factor in ones statistics and that most people in general over value its importance. I don’t know how often I hear arguments like “but he plays all the tough minutes” as an excuse as to why a player has poor statistics and pretty much every time I do I cringe because almost certainly the person making the argument has no clue how much those tough minutes impact a players statistics.

While thinking of how to do this study, and which players to look at, I was listening to a pod cast and the name Pavel Datsyuk was brought up so I decided I would take a look at him because in addition to being mentioned in a pod cast he is a really good 2-way player who plays against pretty tough quality of competition. For this study I looked at 2010-12 two year data and Datsyuk has the 10th highest HART QoC during that time in 5v5 zone start adjusted situations.

The next step was to look how Datsyuk performed against various types of opposition. To do this I took all of Datsyuk’s opponent forwards who had he played at least 10 minutes of 5v5 ZS adjusted ice time against (you can find these players here) and grouped them according to their HARO, HARD, CorHARO and CorHARD ratings and looked at how Datsyuk’s on-ice stats looked against each group.

OppHARO TOI% GA20
>1.1 46.84% 0.918
0.9-1.1 34.37% 0.626
<0.9 18.79% 0.391

Lets go through a quick explanation of the above table. I have grouped Datsyuk’s opponents by their HARO ratings into three groups, those with a HARO >1.1, those with a HARO between 0.9 and 1.1 and those with a HARO rating below 0.9. These groups represent strong offensive players, average offensive players and weak offensive players. Datsyuk played 46.84% of his ice time against the strong offensive player group, 34.37% against the average offensive player group and 18.79% against the weak offensive player group. The GA20 column is Datsyuk’s goals against rate, or essentially the goals for rate of Datsyuk’s opponents when playing against Datsyuk. As you can see, the strong offensive players do significantly better than the average offensive players who in turn do significantly better than the weak offensive players.

Now, let’s look at how Datsyuk does offensively based on the defensive ability of his opponents.

OppHARD TOI% GF20
>1.1 35.39% 1.171
0.9-1.1 35.36% 0.994
<0.9 29.25% 1.004

Interestingly, the defensive quality of Datsyuk’s opponents did not have a significant impact on Datsyuk’s ability to generate offense which is kind of an odd result.

Here are the same tables but for corsi stats.

OppCorHARO TOI% CA20
>1.1 15.59% 15.44
0.9-1.1 77.79% 13.78
<0.9 6.63% 10.84

 

OppCorHARD TOI% CF20
>1.1 18.39% 15.89
0.9-1.1 68.81% 18.49
<0.9 12.80% 22.69

I realize that I should have tightened up the ratings splits to get a more even distribution in TOI% but I think we see the effect of QoC fine. When looking at corsi we do see that CF20 varies across defensive quality of opponent which we didn’t see with GF20.

From the tables above, we do see that quality of opponent can have a significant impact on a players statistics. When you are playing against good offensive opponents you are bound to give up a lot more goals than you will against weaker offensive opponents. The question remains is whether players can and do play a significantly greater amount of time against good opponents compared to other players. To take a look at this I looked at the same tables above but for Valtteri Filppula, a player who rarely gets to play with Datsyuk so in theory could have a significantly different set of opponents to Datsyuk. Here are the same tables above for Filppula.

OppHARO TOI% GA20
>1.1 42.52% 1.096
0.9-1.1 35.35% 0.716
<0.9 22.12% 0.838

 

OppHARD TOI% GF20
>1.1 32.79% 0.841
0.9-1.1 35.53% 1.197
<0.9 31.68% 1.370

 

OppCorHARO TOI% GA20
>1.1 12.88% 19.03
0.9-1.1 78.20% 16.16
<0.9 8.92% 14.40

 

OppCorHARD TOI% GF20
>1.1 20.89% 15.48
0.9-1.1 64.94% 17.16
<0.9 14.17% 19.09

Nothing too exciting or unexpected in those tables. What is more important is how the ice times differ from Datsyuk’s across groups and how those differences might affect Filppula’s statistics.

We see that Datsyuk plays a little bit more against good offensive players and a little bit less against weak offensive players and he also plays a little bit more against good defensive players and a little bit less against weak defensive players. If we assume that Filppula played Datsyuk’s and that Datsyuk’s within group QoC ratings was the same as Filppula’s we can calculate what Filppula’s stats will be against similar QoC.

Actual w/ DatsyukTOI
GF20 1.135 1.122
GA20 0.905 0.917
GF% 55.65% 55.02%
CF20 17.08 17.09
CA20 16.37 16.49
CF% 51.05% 50.90%

As you can see, that is not a huge difference. If we gave Filppula the same QoC as Datsyuk instead of being a 55.65% GF% player he’d be a 55.02% GF% player. That is hardly enough to worry about and the difference in CF% is even less.

From this an any other study I have looked at I have found very little evidence that QoC has a significant impact on a players statistics. The argument that a player can have bad stats because he plays the ‘tough minutes’ is, in my opinion, a bogus argument. Player usage can have a small impact on a players statistics but it is not anything to be concerned with for the vast majority of players and it will never make a good player have bad statistics or a bad player have good statistics. Player usage charts (such as those found here or those found here) are interesting and pretty neat and do give you an idea of how a coach uses his players but as a tool for justifying a players good, or poor, performance they are not. The notion of ‘tough minutes’ exists, but are not all that important over the long haul.

 

 

Mar 202013
 

I generally think that the majority of people give too much importance to quality of competition (QoC) and its impact on a players statistics but if we are going to use QoC metrics let’s at least try and use the best ones available. In this post I will take a look at some QoC metrics that are available on stats.hockeyanalysis.com and explain why they might be better than those typically in use.

OppGF20, OppGA20, OppGF%

These three stats are the average GF20 (on ice goals for per 20 minutes), OppGA20 (on ice goals against per 20 minutes) and GF% (on ice GF / [on ice GF + on ice GA]) of all the opposition players that a player lined up against weighted by ice time against. In fact, these stats go a bit further in that they remove the ice time the opponent players played against the player so that a player won’t influence his own QoC (not nearly as important as QoT but still a good thing to do). So, essentially these three stats are the goal scoring ability of the opposition players, the goal defending ability of the opposition players, and the overall value of the opposition players. Note that opposition goalies are not included in the calculation of OppGF20 as it is assume the goalies have no influence on scoring goals.

The benefits of using these stats are they are easy to understand and are in a unit (goals per 20 minutes of ice time) that is easily understood. GF20 is essentially how many goals we expect the players opponents would score on average per 20 minutes of ice time. The drawback from this stat is that if good players play against good players and bad players play against bad players a good player and a bad player may have similar statistics but the good players is a better player because he did it against better quality opponents. There is no consideration for the context of the opponents statistics and that may matter.

Let’s take a look at the top 10 forwards in OppGF20 last season.

Player Team OppGF20
Patrick Dwyer Carolina 0.811
Brandon Sutter Carolina 0.811
Travis Moen Montreal 0.811
Carl Hagelin NY Rangers 0.806
Marcel Goc Florida 0.804
Tomas Plekanec Montreal 0.804
Brooks Laich Washington 0.800
Ryan Callahan NY Rangers 0.799
Patrik Elias New Jersey 0.798
Alexei Ponikarovsky New Jersey 0.795

You will notice that every single player is from the eastern conference. The reason for this is that the eastern conference is a more offensive conference. Taking a look at the top 10 players in OppGA20 will show the opposite.

Player Team OppGF20
Marcus Kruger Chicago 0.719
Jamal Mayers Chicago 0.720
Mark Letestu Columbus 0.721
Andrew Brunette Chicago 0.723
Andrew Cogliano Anaheim 0.723
Viktor Stalberg Chicago 0.724
Matt Halischuk Nashville 0.724
Kyle Chipchura Phoenix 0.724
Matt Belesky Anaheim 0.724
Cory Emmerton Detroit 0.724

Now, what happens when we look at OppGF%?

Player Team OppGF%
Mike Fisher Nashville 51.6%
Martin Havlat San Jose 51.4%
Vaclav Prospal Columbus 51.3%
Mike Cammalleri Calgary 51.3%
Martin Erat Nashville 51.3%
Sergei Kostitsyn Nashville 51.3%
Dave Bolland Chicago 51.2%
Rick Nash Columbus 51.2%
Travis Moen Montreal 51.0%
Patrick Marleau San Jose 51.0%

There are predominantly western conference teams with a couple of eastern conference players mixed in. The reason for this western conference bias is that the western conference was the better conference and thus it makes sense that the QoC would be tougher for western conference players.

OppFF20, OppFA20, OppFF%

These are exactly the same stats as the goal based stats above but instead of using goals for/against/percentage they use fenwick for/against/percentage (fenwick is shots + shots that missed the net). I won’t go into details but you can find the top players in OppFF20 here, in OppFA20 here, and OppFF% here. You will find a a lot of similarities to the OppGF20, OppGA20 and OppGF% lists but if you ask me which I think is a better QoC metric I’d lean towards the goal based ones. The reason for this is that the smaller sample size issues we see with goal statistics is not going to be nearly as significant in the QoC metrics because over all opponents luck will average out (for every unlucky opponent you are likely to have a lucky one t cancel out the effects). That said, if you are doing a fenwick based analysis it probably makes more sense to use a fenwick based QoC metric.

HARO QoC, HARD QoC, HART QoC

As stated above, one of the flaws of the above QoC metrics is that there is no consideration for the context of the opponents statistics. One of the ways around this is to use the HockeyAnalysis.com HARO (offense), HARD (defense) and HART (Total/Overall) ratings in calculating QoC. These are player ratings that take into account both quality of teammates and quality of competition (here is a brief explanation of what these ratings are).The HARO QoC, HARD QoC and HART QoC metrics are simply the average HARO, HARD and HART ratings of players opponents.

Here are the top 10 forwards in HARO QoC last year:

Player Team HARO QoC
Patrick Dwyer Carolina 6.0
Brandon Sutter Carolina 5.9
Travis Moen Montreal 5.8
Tomas Plekanec Montreal 5.8
Marcel Goc Florida 5.6
Carl Hagelin NY Rangers 5.5
Ryan Callahan NY Rangers 5.3
Brooks Laich Washington 5.3
Michael Grabner NY Islanders 5.2
Patrik Elias New Jersey 5.2

There are a lot of similarities to the OppGF20 list with the eastern conference dominating. There are a few changes, but not too many, which really is not that big of a surprise to me knowing that there is very little evidence that QoC has a significant impact on a players statistics and thus considering the opponents QoC will not have a significant impact on the opponents stats and thus not a significant impact on a players QoC. That said, I believe these should produce slightly better QoC ratings. Also note that a 6.0 HARO QoC indicates that the opponent players are expected to produce a 6.0% boost on the league average GF20.

Here are the top 10 forwards in HARD QoC last year:

Player Team HARD QoC
Jamal Mayers Chicago 6.0
Marcus Kruger Chicago 5.9
Mark Letestu Columbus 5.8
Tim Jackman Calgary 5.3
Colin Fraser Los Angeles 5.2
Cory Emmerton Detroit 5.2
Matt Belesky Anaheim 5.2
Kyle Chipchura Phoenix 5.1
Andrew Brunette Chicago 5.1
Colton Gilles Columbus 5.0

And now the top 10 forwards in HART QoC last year:

Player Team HART QoC
Dave Bolland Chicago 3.2
Martin Havlat San Jose 3.0
Mark Letestu Columbus 2.5
Jeff Carter Los Angeles 2.5
Derick Brassard Columbus 2.5
Rick Nash Columbus 2.4
Mike Fisher Nashville 2.4
Vaclav Prospal Columbus 2.2
Ryan Getzlaf Anaheim 2.2
Viktor Stalberg Chicago 2.1

Shots and Corsi based QoC

You can also find similar QoC stats using shots as the base stat or using corsi (shots + shots that missed the net + shots that were blocked) on stats.hockeyanalysis.com but they are all the same as above so I’ll not go into them in any detail.

CorsiRel QoC

The most common currently used QoC metric seems to be CorsiRel QoC (found on behindthenet.ca) but in my opinion this is not so much a QoC metric but a ‘usage’ metric. CorsiRel is a statistic that compares the teams corsi differential when the player is on the ice to the teams corsi differential when they player is not on the ice.  CorsiRel QoC is the average CorsiRel of all the players opponents.

The problem with CorsiRel is that good players on a bad team with little depth can put up really high CorsiRel stats compared to similarly good players on a good team with good depth because essentially it is comparing a player relative to his teammates. The more good teammates you have, the more difficult it is to put up a good CorsiRel. So, on any given team the players with a good CorsiRel are the best players on team team but you can’t compare CorsiRel on players on different teams because the quality of the teams could be different.

CorsiRel QoC is essentially the average CorsiRel of all the players opponents but because CorsiRel is flawed, CorsiRel QoC ends up being flawed too. For players on the same team, the player with the highest CorsiRel QoC plays against the toughest competition so in this sense it tells us who is getting the toughest minutes on the team, but again CorsiRel QoC is not really that useful when comparing players across teams.  For these reasons I consider CorsiRel QoC more of a tool to see the usage of a player compared to his teammates, but is not in my opinion a true QoC metric.

I may be biased, but in my opinion there is no reason to use CorsiRel QoC anymore. Whether you use GF20, GA20, GF%, HARO QoC, HARD QoC, and HART QoC, or any of their shot/fenwick/corsi variants they should all produce better QoC measures that are comparable across teams (which is the major draw back of CorsiRel QoC.