Tuesday, August 30, 2011

The Best and the Worst of the 2011 NHL Offseason

The 2011 NHL offseason was an interesting one. An unprecedented growth in the salary cap (and floor) meant that GMs had more money to throw around than ever before. Unfortunately for these teams (but not for the UFAs and their agents), there was a dearth of high-end talent which resulted in numerous bad signings and an increased frequency of big-time trades. From the day before the draft when Philadelphia offloaded both Jeff Carter and Mike Richards to the draft day trades where two WC elites cemented their status for years to come, to the surprise swap of Dany Heatley and Martin Havlat, to say this offseason was hectic in terms of trades would be a drastic understatement. The goal then of this post is to address the moves we think were best and the moves we think were worst.

Best


5. Ian White signing - 2 years, $5.75 million

Triumph has already done a great job highlighting this move, so I’ll keep it brief. Frankly, I don’t know which element of this deal is more impressive from Detroit’s perspective, the cap hit or the term. Signing a defenseman with White’s offensive prowess in UFA to a deal paying him less than $3M/year is a real coup. Detroit needed somebody to run the power play with the retirement of Brian Rafalski; thankfully for them they found upgraded the player for about 50% the cost.

4. Jeff Carter trade (Columbus)

Again, this move has been well documented by my colleague, though most of that has been an analysis of Philadelphia’s haul. I want to focus on this from Columbus’ perspective. First and foremost, they get an elite forward who is likely to play Center alongside Rick Nash. The folks at Broad Street Hockey put together some great analysis on Carter. An under appreciated element of this deal is that a trade like this is the only way Columbus could’ve acquired a player of Carter’s caliber without drastically overpaying somebody as a free agent, a mistake we will cover later on in this article.

3. Brian Campbell trade (Chicago)

The reaction that most had to this trade tells us all that needs to be said. Most hockey folks were in utter disbelief that the Blackhawks could actually move Brian Campbell’s deal. 5 years left at a cap hit north of $7m for an above-average (but not elite) defenseman. Fortunately for Blackhawks fans, Brian Campbell’s biggest fan had found himself a new job in Sunrise, Florida. The newly found cap space allowed Stan Bowman to add some precious depth to an already impressive core, and if necessary, gives the Blackhawks freedom to tweak their roster before the trade deadline. While Chicago is assured to miss Campbell, they are still a Stanley Cup contender and #1 on Driving Play’s forthcoming pre-season power ranking. Looking to the future, this move will also pay-off next year and the years thereafter. Having all core players locked up for at least 3 years plus an added $7.5m in cap space puts Chicago in the discussion of the NHL’s most dangerous teams going forward. None of this would’ve been possible with Brian Campbell’s contract still on the books.

2. Semyon Varlamov trade (Washington)

This one is pretty simple. Semyon Varlamov, while a nice player, is by no means an elite goaltender. He was also on his way out of Washington, likely to the KHL. George McPhee deserves a medal for moving a player his team was unlikely to sign for what has the potential to be the first overall pick. Washington already has an impressive core, and Michal Neuvirth’s play last year made Varlamov expendable. Adding a high first round pick to that mix is extremely impressive. To be less politically correct, I’d set the over/under on 15 seconds for George McPhee holding off laughter once the trade call was finalized.

1. Tomas Vokoun signing - 1 year, $1.5 million

We believe the best move of the offseason belongs to Washington for their signing of Tomas Vokoun. While the benefactor of one of the league’s most aggressive shotcounter, there is no doubting Vokoun’s status as an elite goaltender. Adding one of the NHL’s best netminders to one of its best teams makes for a scary proposition for the rest of the league. What makes this deal all the more impressive is the low term and the low cap hit. It is unlikely Vokoun will play in Washington in 2012-13, but that doesn’t matter. What does matter is they made an appreciable improvement to their 2012 Cup equity for the absurdly low price of $1.5m. One side note - if the rumors about Colorado's interest in Vokoun are true, then they really got played this offseason, what with losing out on Vokoun, overpaying for Washington's goalie leftovers, and then seeing Vokoun sign a 1-year deal in Washington.

Worst


Honorable Mention: Florida's Offseason

Dale Tallon has always had the reputation of somebody who can spend like a drunken sailor, so news of a rising cap (more importantly, a rising cap floor) must have been well received on his end. He responded first by trading for Brian Campbell, who’s large cap hit made a nice initial dent into the amount of money they’d need to spend to reach the floor. July 1st is when the fun really started, and when it was over he gave a collection of average 2nd and 3rd liners (Tomas Fleischmann, Scottie Upshall, Tomas Kopecky, Sean Bergeinheim, and Marcel Goc) a total of $15,450,000 for next year. None of this is nearly as bad as giving Ed Jovanovski $4,125,000 a year for 4 years on a 35+ contract. Where some teams made the most of a rising cap (Washington), other teams shot themselves in the foot (Florida).

5. Philadelphia's Trades + the Ilya Bryzgalov signing - 9 years, $51 million

Again, the two trades have been covered at great length by my colleague Chase, so the plan here is to take a little bit of a different approach. I first want to address the timing of these moves. Philadelphia did about as well as they could from a return standpoint. This in specific has been covered by Chase, so I’ll merely direct you to that analysis if you have not seen it before. My issue with the trades is timing. There seems to be a pretty big disconnect between the ages of Philly’s core pieces. On one hand you re-tool for the future, with the idea that the haul of Voracek, Couturier, Simmonds, and Schenn can come close to duplicating the performance of Carter and Richards. There are two problems here: first, there is a good chance this doesn’t happen, for reasons that should be pretty obvious. Second, if it does happen, it’s unlikely to occur soon, which is important considering their best player is 36 years old and coming off of knee surgery. If Pronger stays healthy, the Flyers probably win the East and Sergei Bobrovsky is probably still their starting goaltender. That ought to tell you all you need to know about the Bryzgalov signing. It’s an overreaction and it was completely unnecessary.

4. Brad Richards signing - 9 years, $60 million

Yes, Brad Richards put up some pretty numbers the last two years, but no, he’s not as good as those numbers appear and he’s certainly not 9 years/$60 million good. For one, Richards doesn’t do much in the way of puck possession. Dallas has been outshot when Richards was on the ice for three of the last four years, and while Richards has generally played against relatively tough competition over that stretch, he has not shown the ability to drive the play forward, an asset we believe to be an important element of player evaluation. And while there is no doubting Richards is one of the league’s best playmakers, it does not make him one of the league’s best players. This overpayment, coupled with Richards’ general injury history and his non-elite even strength play, makes the signing of the star of the 2011 UFA class a bad one.

3. Ville Leino signing - 6 years, $27 million

Where the Richards deal was more the case of a very good player being paid like an elite one, the Leino deal is a case where a slightly above average player is being paid like a very good one. After struggling to get solid ice time on a stacked Detroit roster, Leino parlayed a good 2010 playoffs and a nice 2010-11 regular season into a 6 year, $27 million dollar deal. The talent is there, but our issue with the deal is that Buffalo likely failed to delve deeper into Leino’s numbers. See, Leino’s numbers had a few things going for them that made them appear to be better than they actually were. First, he played for one of the league’s best teams. This had an impact on the quality of his minutes. Philadelphia had the fortune of having two elite centers to play against the opposing teams’ best players, allowing Leino and his principal linemates, Danny Briere and Scott Hartnell, to reap the benefit of soft minutes. Leino - Briere - Hartnell as a unit also took a disproportionate amount of faceoffs in the offensive zone (Leino’s OZone% was 62.3%). Even under this perfect storm of good teammates and soft minutes, Leino only managed to put up 53 points. Now, his raw point total may rise because of an increase in ice time and power play time, but as we know, a player’s point total is misleading without context, and as of now, it looks like Buffalo mistook a player who took advantage of a very favorable circumstance for a very good player.

2. James Wisniewski signing - 6 years, $33 million

Part of the reason the Jeff Carter trade was so great for Columbus is because it allowed them to acquire an elite player without having to overpay in terms of cap hit. While Wisniewski is not an elite player, he is still quite solid, but nevertheless, we saw the phenomena that afflicts Columbus and similarly sized and located markets take full effect, as the Jackets drastically overpaid to acquire the services of Wiz. The problem is that almost all of Wisniewski’s offensive value comes from his ability on the power play, as 60% of his points came with a man advantage. There is obviously value here, but as power plays occur less frequently, the value of power play specialists fall. Wiz is solid at even strength; he’s not much of a play driver, but in the past he’s shown the ability to play tough minutes. However, all of this doesn’t really add up to a player worth $5.5 million per year. Teams like Columbus generally have internal caps that are pretty rigid, and dedicating such a large proportion of that space to good but not great players is a recipe for sustained mediocrity.

1. Semyon Varlamov trade (Colorado)

To be honest, the first thing I did when I heard about this move was laugh. I didn’t laugh because of Semyon Varlamov, I laughed because a team that is in the middle of rebuilding just moved what will almost assuredly be a top 5 pick for an average (at best) starting goalie. Yes, Varlamov is young and now under team control for three years, no, that is not as valuable as a high first round pick. Where this move is worse than any of the signings (or the other trade) is while this move might make Colorado marginally better this year (but still nowhere near good enough to be a playoff team), it almost assuredly makes Colorado much worse in the future. The recent success of teams like the Blackhawks and Penguins show that rebuilds can happen quick with good drafting, especially in the early rounds. Colorado just traded away what used to be a top organizational asset for a player who will have no appreciable impact on their ability to contend in the future.

Monday, August 29, 2011

On Whether Or Not Europeans 'Disappear' In The Playoffs

Several days ago, I asked the good readers of Driving Play what they'd like to see in a post. 'djschoo' commented that he'd like to see whether or not Europeans 'disappear' in the playoffs. I thought this was a great question and I was so excited about figuring this out that right away I decided to put it off until tomorrow. And the next day. However, it's nearly two weeks later and I've returned with an answer.

I finally decided on a simple method to see if European players 'disappear' - I would take European forwards who played on playoff teams who scored at least .5 points per game and played in at least 40 games in a given season. I would then look at their playoff point per game totals and compare them to their regular season totals. If the Europeans did worse in the playoffs than they did in the regular season, they're clearly chokers who hate winning. Simple enough, right?

However, I saw a problem with this method. Defenses are better in the playoffs, so I had no idea by how much a player's points per game would be 'expected' to decrease. Without a touchstone, I wouldn't have any idea how much better or worse Europeans are doing compared to a theoretical average. So I used Canadian-born forwards as my touchstone, since we all know they have heart and grit and grow up dreaming about the Stanley Cup. However, I did not edit out French Canadian players, so this study is not exactly comprehensive. (That's a joke, people). Also, I only used Swedes, Finns, Slovaks, Czechs, and former Soviets as 'Europeans' - apologies to Thomas Vanek, Jochen Hecht, and Marco Sturm, but they're not covered in here, mostly out of my laziness. All numbers are courtesy of hockey-reference.com's wonderful Play Index feature.

Before I show the numbers, I do want to point this out, as I knew this would be an issue before I even began the study. Players play different amounts of playoff games based on how well their team does. I therefore knew I'd run into 'survivorship bias' - the farther a particular player X got in the playoffs, the more likely it would be that he personally did well, since his team was winning games. Plus, the more games a given player plays, the more impact he's going to have on the summed ratio of points to games. 40 games and a .5 points per game minimum is going to cover most of the top forwards on a given team. If most of the top forwards on that team do well, the team is likely going to do well in the playoffs. Let's keep that in mind as we look at the numbers.

The first table is just going to show points per game for Canadians in the regular season and playoffs, compared with European points per game in the regular season and the playoffs. I've bolded who had the smaller difference (i.e. who was 'better') each year.

YearCanada RegCanada PoffDifferenceEuro RegEuro PoffDifference
2005-060.8220.689-0.1330.8500.592-0.258
2006-070.8030.579-0.2240.8550.556-0.299
2007-080.8120.691-0.1210.8160.793-0.023
2008-090.7730.625-0.1480.8090.682-0.127
2009-100.7080.7100.0020.8440.805-0.039
2010-110.7290.697-0.0320.7590.693-0.066

We can see that the Canadians beat the Europeans 4 out of the last 6 years, in terms of having a better points per game difference. Okay, so we all knew that Europeans had no heart and intentionally tank in the first round so they can go play in the World Championships, which is what they really care about. However, this table doesn't tell the whole story. I decided to tally all the games played and all the points scored by the players in this study over this stretch to compare the total since the lockout.

YearCDN GPCDN PCDN P/GPCDN POFF GPCDN POFF PCDN POFF P/GP
2005-06410933790.8226334360.689
2006-07372629910.8034972880.579
2007-08280722800.8123692550.691
2008-09370928670.7735263290.625
2009-10379226840.7085563950.71
2010-11383927970.7296374440.697
Totals21982169980.773321821470.667

And now, Europeans:

YearEUR REG GPEUR REG PEUR REG P/GPEUR POFF GPEUR POFF PEUR POFF P/GP
2005-06239920400.852501480.592
2006-07279223870.8553812120.556
2007-08262221390.8164113260.793
2008-09282522850.8094813280.682
2009-10206517420.8442562060.805
2010-11206215650.7592671850.693
Totals14765121580.823204614050.687

We can see that the Canadians have -.106 points per game in the playoffs as compared to the regular season, whereas the Europeans have -.136. So, Canada wins, right? Not necessarily - here's some reasons why that number could be lower for Europeans:

A: Survivorship bias. I know I went over this before, but the numbers make it more stark. In 3 of the 6 seasons, the Europeans have accounted for 300 or fewer playoff games played. That's not a lot. Also, the distribution of European players throughout the league is grossly uneven. The two years where European players beat out Canadians in terms of regular season and playoff point differential also happen to be the two years when the Detroit Red Wings made the Stanley Cup Finals. The Red Wings are heavy on European players. In essence, while there are teams without a strong European presence who make the Cup Finals, there are very few teams without at least 2 Canadian forwards in their top 6. This means there will almost always be several Canadian players who have long and great playoff runs, but that's not true of Europeans. In 2010, only one European forward qualified for this study who played in the Cup Finals (Marian Hossa), and he didn't even have a good playoff.

B: Chance. I should just have this written out for every post, but the playoffs are a huge crapshoot. Chance combined with the survivorship bias mentioned above could be the largest reasons why Canadian players are doing better relative to European players.

C: Fewer Power Plays. I don't have any evidence to support this right now, but I suspect that European players play more and do better on the power play than their Canadian counterparts. I also suspect that there are fewer power plays and better penalty-killing units in the playoffs than in the regular season. This took long enough to investigate - perhaps I will look at this in a later post.

D: European players refuse to go to the corner or in front of the net and only care about Olympic Gold. Well, we all knew this one already, didn't we? Anyway, thanks for reading - next up from me: Devils vs. Flyers playoff results from 1995-2011, Martin Brodeur vs. The Cast of Thousands - Has Brodeur really been the difference?

Tuesday, August 23, 2011

Part II: The Aftermath of the Mike Richards and Jeff Carter Deals

In the middle of the purgatory also known as the NHL offseason, the beginning of unrestricted free agency on July 1st serves as a mad dash of excitement. As fans, we are oftentimes given our first glance into what our favorite teams will look like come October. For fans of the Philadelphia Flyers, however, the first preview into 2011-2012 came 8 days prior with the dealing of Mike Richards and Jeff Carter. With the new found cap space following the trades, GM Paul Holmgren wasted little time signing goaltender Ilya Bryzgalov to a massive 9-year/$51 million contract in an attempt to shore up his goaltending situation. The Flyers had of course acquired Bryzgalov’s negotiating rights from Phoenix prior to the start of free agency, and were therefore able to lock up the goaltender should they come to terms before July 1. Once he was free to negotiate with the rest of the league’s UFAs, Holmgren decided to make a splash by signing Jaromir Jagr, Maxime Talbot, and Andreas Lilja to new contracts. Where exactly will these four acquisitions fit within the Flyer lineup? Let us turn to the numbers to try and find an answer.

Taking my familiar approach, I am going to start by looking at the per game ice time of the Flyers’ newly-acquired skaters. Setting Jagr aside, below are the totals for both Talbot and Lilja from last season per nhl.com:

PlayerGames PlayedES TOI/GameTeam RankPP TOI/GameTeam RankSH TOI/GameTeam RankTotal TOI/GTeam Rank
Talbot (PIT)8212:01100:06202:55215:049
Lilja (ANA)5213:5360:0173:30217:255

As we can see, Talbot’s ES TOI would most certainly portray him as a bottom-six checking forward – a role which he will be expected to reprise in Philadelphia. He is also certain to play an important role in the team’s penalty killing, and his hefty average time helped contribute to a Pittsburgh unit which ranked 3rd in the NHL in shots allowed per 60 minutes last season.

Moving to Lilja, it seems that he is a defensemans version of Talbot – a third pairing player who logs major minutes on the penalty kill. He should see a similar role in Philadelphia, replacing veteran Sean O’ Donnell whilst challenging youngsters Erik Gustafson and Kevin Marshall for playing time. Lilja should bolster the team’s back-end depth with another capable body, which could prove important should veteran Chris Pronger miss an extended amount of time to begin the season.

Getting back to the team’s wild card signing of the summer, where does Jaromir Jagr fit into all of this? Having not played in the NHL since 2007-08, it’s hard to project just how much ice time the future hall-of-famer will see this season as we cannot refer to any recent data. However, Jagr does figure to play a large role both at 5-on-5 and on the power play, and will most certainly see top-6 minutes in both categories. Setting Jagr aside once again, what kind of context can we give both Talbot and Lilja’s ice time? The tables below should help to clarify such a question:

PlayerES G (Team Rank)ES A (Team Rank)ES Pts (Team Rank)PP G (Team Rank)PP A (Team Rank)PP Pts (Team Rank)
Talbot (PIT)6 (T-13)12 (10)18 (11)0 (T-11)0 (T-13)0 (T-14)
Lilja (ANA)1 (T-5)6 (6)7 (6)0 (T-5)0 (T-6)0 (T-7)

PlayerCorsi ONCorsiRelFenwick %CorsiRelQoCZone Start %Zone Finish %
Talbot (PIT)3.91-2.955.30.45954.249.9
Lilja (ANA)-19.64-11.449.4-0.01050.444.6

Beginning with the man they call "Superstar", Talbot was asked to play against fairly good competition this past season and on the surface seems to have performed okay. Eric T.’s Balanced Corsi shows us that his Corsi ON score is close to what we would expect from someone given such a role, though it does fall short by about one shot. Talbot’s balanced zone shift is also a little below expected which suggests that the gritty forward’s 55.3% Fenwick score is most likely the result of his favorable zone start. Our resident stats guru JaredL took a deeper look at Talbot trying to figure out if the Flyers were getting value out of his new 5-year/$8.75 million contract:

Max Talbot

JaredL here. Like I did for Voracek, Chase wanted me to crunch the numbers on Talbot. Maxime Talbot made a name for himself by getting smoked by Dan Carcillo and scoring on Chris Osgood, two very impressive feats indeed. So he's like Jeff Carter but with four times as many goals in the Stanley Cup Finals. Wow has my material taken a beating this offseason.

The Penguins are probably the toughest team for WOWY analysis because the without-you numbers are mostly determined by who played more with Malkin, Staal and especially Crosby. The best way I can think of to evaluate a role player like Talbot is to look at how well those three did with him on the ice compared to without. The injuries last season make this tough because the sample sizes are really small. Even going back two or three seasons they aren't great but a pattern definitely emerges. Here are the numbers from the last three seasons for those centers, plus the times none of the three were on the ice, with and without Talbot. I excluded times where two of Crosby, Malkin and Staal were together.

CenterTalbotCorsiTimeRateOppCorsi
CrosbyOn13264.72.947-0.977
CrosbyOff2001972.96.0820.161
MalkinOn-25269.1-5.5740.358
MalkinOff-601583.2-2.274-0.021
StaalOn-21136.7-9.2210.165
StaalOff2342419.35.8030.685
None of 3On-651533.8-2.5430.097
None of 3Off472260.41.248-0.274

Looking only at score-tied spots it gets a bit better for Talbot, but the same unfortunate pattern is present:

CenterTalbotCorsiTimeRateOppCorsi
CrosbyOn1191.37.226-0.848
CrosbyOff123777.59.4930.224
MalkinOn-597.2-3.086-0.941
MalkinOff10601.80.9970.166
StaalOn-1851.4-20.998-0.13
StaalOff34940.22.170.687
None of 3On24610.62.3580.187
None of 3Off998766.781-0.483

As you can see from the tables, every group was worse off with him on the ice. Most troubling is that this includes the fourth line. It appears that he performed worse 5-on-5 than the average Pittsburgh fourth liner over the last three seasons. I should point out that Talbot was on the ice for about a third of the Penguins' penalty-kill time and appears to have been a positive. The Penguins were just slightly better off with him on the ice, and almost all of that time was with Staal off the ice:

Penalty KillCorsi RateGoal Rate
With Talbot-76.103-4.384
Without Talbot-77.455-5.008

I'm biased, but $1.75M of cap space seems like a lot for a guy who can adequately run the second PK unit but has been somewhere between bad and awful 5-on-5. I'll now pass it back to Chase.

Taking a gander at Lilja’s numbers will probably want to make most of you look away, and for good reason. His scores fall way short of what we would expect from somebody given his role, both in balanced Corsi and balanced zone shifts. His quality of competition certainly isn’t anything to write home about, and he still fails to break a 50% Fenwick score with a favorable zone start. All in all, Lilja’s metrics suggest that he should be used in nothing more than a depth role for the Flyers as he certainly isn’t sending the play in the right direction on his own. I do find Lilja as a rather curious signing, especially considering his 35+ contract which will count against the cap for the next two seasons barring he enters long-term injury reserve status.

Getting back to Jagr, we once again do not have any relevant data to project how he may perform in his expected role come October. We could look at how the Rangers last used him in 2007-08, but Jagr is now 39 years old and figures to have declined since his last season in North America. He has scored at close to a point-per-game rate in the KHL (a lower scoring league), but he figures to see a gross increase in his quality of competition playing for the new-look Flyers. The abundance of power play time and favorable zone start percentage that he figures to receive will help his chances of achieving the 50-point benchmark that I have seen thrown around so loosely. Though his contract is for only one season, Jagr’s age is most certainly a concern. For a player with so many question marks, a cap hit of $3.3 million seems to be a bit of an overpayment. Sure, Jagr could score 50+ points and could give the Flyers excellent top-6 production. Nevertheless, all expectations at this point cannot be justified until we finally see how the crafty Czech can perform against the best players in the world.

Having looked into where the new skaters fit into the lineup, will Ilya Bryzgalov help make up for an offense that is missing Mike Richards and Jeff Carter? Adam Kimelman of NHL.com seems to think so, but as friends of Driving Play Geoff Detweiller and Kent Wilson have written, the numbers simply do not bear this out. Giving Bryzgalov such a lucrative contract is most puzzling to me, especially when Sergei Bobrovsky had an excellent rookie season – his first in North America. The Flyers could be using some of the money that they are lighting on fire handing over to Bryzgalov to help bolster the team’s lost offensive depth, but instead are tying it up in one player who cannot help in this regard.

Unfortunately, I am once again left to conclude that the additions of Jagr, Talbot, Lilja and Bryzgalov do not stack up as a net positive for Philadelphia in the wake of the departure of Mike Richards and Jeff Carter. Both Talbot and Lilja do not seem to be carrying the water at 5-on-5 whatsoever and we do not yet know what Jaromir Jagr will provide to the offense should he compete and stay healthy. Mr. Kimelman is not the first to argue that Bryzgalov will make up for some of the offense that is no longer with the club, but as Part III of this saga will point out, Richards and Carter were an integral part of Philadelphia’s success for more reasons than just points on the scoresheet.

Monday, August 22, 2011

Triumph's Principles For Evaluating Prospects

If you're a boorish American hockey fan like me, you probably don't get much of a chance to see NHL prospects, either live or on television. Prospects tend to play in cold, scary places, with a dearth of shopping malls and an overabundance of death by frostbite. We might see ten seconds of his highlights during the NHL Draft, we might see him looking awkward in a ridiculous suit going up to the podium, and then for several years he's nothing but numbers on a page.

Whatever the case, I find myself shaking my fist and gnashing my teeth at the way prospects tend to be evaluated, especially the evaluation on a site that rhymes with Rockey's Suture. I've thought about it a bit and have come up with some rules that I use when I look at a prospect's stats. I don't know that my way is any better than their way, but I do think mine is at least more logical. Again, this is in the face of the fact that quite simply, most of us do not ever get to see these players until they are in an NHL uniform.

Just to be clear, I am assuming that I am talking about rating prospects according to statistics found on hockeydb.com and theahl.com, as well as junior hockey and college hockey websites.

Principle #1: The most recent season is the most relevant, with descending importance to be placed on each season previous to the most recent.

This seems self-explanatory, but I see so much prospect evaluation that overrates performances from 3 years ago, or underrates what happened last year.

Principle #1B: Don't overvalue a fluky season

While the most recent season is the most important, a season that appears to be an aberration may just be that. A large jump in numbers may not indicate a large jump in ability. We can't always get shooting percentage data for the lower levels, but an inordinately high shooting percentage always makes me wary.

Principle #2: Draft position matters, but not really - the further away we get from that player's draft, the less relevant their draft position is.

Players taken near the top of the draft tend to outperform players taken later in the draft. This is not an absolute principle, however - one year after a draft, we can re-evaluate all the players who participated in that draft. Even after one year, there will be some players who have underperformed their draft position, and some have outclassed the players around where they were taken. After several years, the differences will become more stark. I see too much player analysis that relies on the fact that a guy was taken at a high position 3 years ago. So what! That information shows that he was likely worthy of being thought of as a top player, then. But what matters more is what he's done since then.

Principle #2b: Don't sleep on undrafted free agents

Hockey is being played all over the Northern part of the world - it's become more and more popular. Sometimes scouts just plain don't see a particular guy and he goes undrafted. Or, more likely, an athlete simply improved from the time he was draft-eligible. There are plenty of examples of players who were not drafted who have gone on to excel in the NHL. Undrafted free agents should be evaluated like everyone else; they're not second-class citizens simply because they weren't one of the top 210 players when they were 18 or 19.

Principle #3: Make sure to take into account a player's age in relation to where he is playing

Prospects play at all different levels and against all different sorts of competition. It is extremely difficult to evaluate the play of an 18 year old player in junior hockey, versus an 18 year old in college, versus an 18 year old in the NHL or AHL, versus an 18 year old in the KHL, and so on. Regardless, take into consideration whether the prospect is playing in a men's league, or if he's 'old for his league' (e.g. college senior, overage junior player). Just playing on a men's team at 18 can mean a prospect is better than a similar prospect dominating the junior league in the same country.

Principle #3b: If a prospect is 'old for his level', he should be dominating

This seems simple enough, but is often missed - a college senior or overage junior player, if he is to be considered a legitimate NHL prospect, should be crushing. He should be one of the top players in the league. Most of his fellow prospects have moved on to more difficult levels - he needs to demonstrate that he is capable of taking that next step.

Principle #4: The strength of a prospect's team matters

Player A scored 60 points in 60 games, but he's the 5th highest scorer on the team, which did very well. Player B is the same age and plays the same position, but played in a different league. He scored 20 points in 40 games, but he was 2nd on the team, which did poorly. Who's a better player? It's obviously unclear, but unless Player A is on a team with some really great prospects, I'm inclined to downgrade his performance. A rising tide lifts all point-scorers, and likewise, an ebb tide lowers all point-scorers.

Principle #5: Defensemen are more difficult to evaluate than forwards

Since the main method of evaluating these prospects is point scoring, we know that points don't even come close to telling the whole story with a defenseman. Neither does plus/minus. In general, more points are obviously better than fewer, but plenty of defensemen have had their offense dry up by the time they reached the NHL. Plus/minus is too team-dependent to have much meaning. Ice time is a better measure than plus/minus, but that's not always available.

Principle #6: Give up on evaluating goalies, it's impossible

This is mostly tongue-in-cheek, but goalies seem to me largely unpredictable. Occasionally there's a goalie who blows away the competition; those are easy to spot. The rest, who has any idea? There's always a Lundqvist or Backstrom that comes out of nowhere to be a star. I pay almost no attention to draft position when it comes to evaluating a goalie.

Principle #7: Do NOT evaluate a player just based on his upside!

If you head over to that site that rhymes with Rockey's Suture, so much of prospect evaluation is based on this nebulous concept of 'upside'. What does this mean, exactly? I'm not sure - they have that silly numbers/grade system also, as if they can predict how a player will grow at the NHL level. Regardless, their reviews of prospects are always glowing, and they talk about how this player 'should' be a second-pairing defenseman in three years, etc. Yet three years later, a lot of these prospects are not in the NHL and no longer considered prospects. Remember to consider 'downside' as well - a prospect may hit a plateau in their growth. Not all of your team's prospects will hit their 'upside'. In fact, most of them won't. Don't be pessimistic, be realistic.

Well, that about wraps 'er up. The future always looks brighter than the present, it's part of why we're sports fans. Still, let's not go inking 19 year olds into the starting lineup in three years. Let's also not fall into the trap of thinking that because a player is a 'prospect', that means he is 'unproven' and 'could fail', even if he's been dominant at all levels prior to this one. That happens, but it's also rare. There's a thoughtful way to evaluate players one has never seen, and it's not what a player 'should be' based on his draft position. It's based on what a player 'could' be based on a reasonable range of outcomes.

Friday, August 19, 2011

On Luck, Skill and Sample Size in Shooting Percentage

The roles of skill and luck in shooting are an important and often misunderstood part of hockey analysis. This is particularly true when analytical and traditional fans get together. A typical discussion might go something like:

A: Steven Stamkos got 91 points with 45 goals at the age of 21. He's definitely getting 100 points next year, and could easily score 50 goals a season as he improves.
B: Yeah, but he was lucky to make 16.5% of his shots. That's not sustainable and his numbers will probably go down next year, even if he does actually improve.
A: So shooting is all luck? You clearly know nothing about hockey and should actually watch some games instead of just sitting at your computer all day coming up with fancy stats that don't mean anything.

Turns out both guys have a point. Stamkos should probably expect his stats to drop next season because, in addition to goal numbers trending downward for the league as a whole, he's probably not burying 16.5% of his shots. On the other hand, I person B probably should watch more hockey, it is a great game.

A lot of the confusion comes from incorrect either/or thinking. Scoring on a high percentage of your shots is a result of both luck and skill. I'm not just putting down the traditional fans here. Those of us in the analytical community, myself included, are prone to bad thinking as well. While most people ignore or underrate the importance of luck in hockey in general and shooting in particular, we tend to go too far the other way, chalking everything up to luck and ignoring the skill aspect.

Shooting for a high percentage is skill based. Two of our favorite writers, JLikens of objectivenhl fame and Gabe Desjardins from BTN and arcticicehockey have written several articles on the subject.

In case their articles are not convincing enough, let's consider two of the best players in the game: Henrik Sedin and Sidney Crosby. While not necessarily known as snipers, these players have all the skills that one might think lead to their team putting a high percentage of shots in the net. Both have elite vision, passing ability, hands, positioning and, in one case, telepathy. We would all expect their teams to have better shooting percentages when they are on the ice than when they are sitting on the bench or worse. The numbers bear this out. Here is a chart with their teams' performances at even strength with both goalies in net from the last four seasons combined. The stats are courtesy of noted Driving Play reader Vic Ferrari's timeonice scripts, which you can find information on how to use here.

TeamGoalsShots On GoalShooting %
Penguins, Crosby on Ice236216010.9%
Penguins, Crosby off Ice41453477.7%
Canucks, Henrik on Ice273262410.4%
Canucks, Henrik off Ice35947097.6%

As you can see, the Pens with Crosby shot 3.2 percentage points higher than they did without him. While some of it may be variance, with the number of shots they took with him on, that's a difference of 69 goals or more than 17 goals per season due to better shooting. The Canucks shot 2.8 points higher with Henrik on the ice, a difference of over 73 goals, more than 18 per season, when you consider how many shots they took with him on. For the statistically minded, these shooting-percentage differences are very very very significant. To give you an idea, it varies field to field but the most common benchmark is for there to be less than a 5% chance of results this extreme, or more so, due to variance alone. That's a 1-in-20 chance. For Crosby, there is a 0.00045% chance, or less than 1 in 222,000. For Hank there is a 0.00235% chance of results that extreme due to randomness alone - less likely than 1 in 42,000. Again, 1 in 20 is the usual mark. The data confirm what anyone would guess from watching a few games - Henrik Sedin and Sidney Crosby help their teams shoot better. (Note: if you are a hater and/or think that it's the likes of Alex Burrows and Pascal Dupuis who are driving these results, feel free to be wrong. The point of this is to provide evidence of shooting skill and clearly someone has it when these two are on the ice.)

Let's now look at the role of luck on shooting percentage. To do this, I will run simulations comparing the results of a typical team that shoots well and one that does poorly. In this article on objectivenhl, which is worthy of being linked again, JLikens finds that the average team shoots at an 8.1% clip 5-on-5, with a standard deviation of 0.48%. Going by this, a team that is good at shooting, let's say 7th or 8th best in the league, would have a true 5-on-5 shooting percentage of something like 8.42%. On the other hand, a team that is bad at shooting, say 7th or 8th worst in the league, would be expected to score on about 7.78% of their shots.

Let's see how things shake out. Below is a chart giving the results of 10,000 simulations for various numbers of shots where team A has a true shooting percentage of 8.42% and team B shoots at 7.78%. The first two columns tell you the given time period and number of shots for each team. The next three columns tell you how often the team good at shooting outshot the bad (column 3), the bad team outshot the good (4) and how often they had an equal shooting percentage (5). The last two columns give what percent of the time someone looking at the data, and not knowing the underlying percentages, would get statistical signficance at the 5% level. Notice that in the last column, the statistical test would reveal that B is significantly better at shooting than A despite their shooting skill actually being over half a percentage point worse.

Time periodNumber of ShotsA scores moreB scores moreGoals scored equalA > B SignificantB > A significant
One Period831.9%28.7%39.4%1.3%1.1%
One Game2442.6%36%21.4%6%4.5%
1/4 Season50062.3%33.2%4.5%10.5%2.2%
1/2 Season1,00068.6%28.5%2.8%13.4%1.4%
1 Season2,00076.4%21.7%1.9%17.9%0.8%
2 Seasons4,00085%14%1%27.8%0.2%
3 Seasons6,00090.1%9.3%0.6%36.1%0.1%
4 Seasons8,00093.6%6%0.4%43.7%0.1%
5 Seasons10,00095.6%4.1%0.3%50.3%0%

You probably didn't find the results surprising for that first row, representing a period of play. The most common outcome, happening about 40% of the time, is that the two teams remain tied, most often at 0. The team that shoots better due to getting higher-quality shots, hitting the corners better and so on is only slightly more likely to be the one that is ahead if you know that one of them is. Less than 32% of the time will the better team find themselves ahead after a period in which both get the league average 8 shots, whereas they'll be behind almost 29% of the time.

Lower on the chart it gets more troubling, especially for us bloggers. The most common sample point for analysis is half a season. Generally the best way to study the persistence of something is to split the season in half, typically first half vs second half or even-and-odd numbered games, and compare the two samples. This works well because teams should be the same or very similar. If you study something over multiple seasons you aren't getting the same teams every year due to player and coaching changes. In half a season, the team near the top in shooting skill has only about a 2 in 3 chance of outscoring the team near the bottom with the same number of shots. There is also little chance, roughly 13%, of finding that the better team is significantly better at shooting if you were looking at the data. Even over a whole season of shooting data, there is a 1 in 4 chance that the worse team will get better results. It isn't until we get several years worth of shooting results that it tilts heavily in favor of the better shooting team and that's not realistic because teams change so much each offseason and the simulations assumed the same percentage each season.

As you can see, luck plays a huge role for all reasonable sample sizes. This is the fundamental reason why shooting stats are better than goals. Luck is less of a factor for number of shots taken than number of shots made, so they are more reliable indicators of skill over samples of a season or less. If over a season there is a 1 in 4 chance that a good-shooting team is outshot by a bad-shooting team then it's tough to say that a team's results are due to skill and not just random luck.

In a future installment I will look at how persistence is affected by sample size.

On The Strange Results Of The Winnipeg Thrashers

Some people may have forgotten this, but in December of 2010, the Thrashers were primed for a playoff berth. Mainstream journalists sat up and took notice. After their game against the Maple Leafs on December 20th, the Thrashers were 19-11-5, with a point percentage of .614. This was finally the year for them - they'd gotten rid of Ilya Kovalchuk, acquired Dustin Byfuglien, and the team was better off. We know what happened next; they went 13-25-7 over their remaining 45 games and finished with the 6th worst overall record in hockey. Then they moved to Winnipeg.

But something strange happened along the way - by 'advanced metrics', the team got better, even as it did worse. Here's a look at Atlanta/Winnipeg's first half and second half even-strength Fenwick with the score tied by player, with a minimum of 10 total games. Fenwick % is shots on goal + missed shots on goal by Team X (here, Atlanta) divided by the total number of shots and missed shots taken. A player's Fenwick % is shots on goal + missed shots FOR while he is on the ice divided by total shots on goal + missed shots by both teams. We're only looking at the results while the score is tied because teams change their strategies when ahead or behind, which fouls up the numbers. (All numbers here courtesy of timeonice.com)

PlayerGP1st Half FenwickGP2nd Half FenwickDifference
Andrew Ladd390.513400.5150.002
Dustin Byfuglien400.5390.5580.058
Johnny Oduya370.425400.5320.107
Chris Thorburn360.452400.5320.08
Anthony Stewart390.479360.456-0.023
Ron Hainsey350.461400.5210.06
Bryan Little330.496410.5310.035
Tobias Enstrom400.477310.5260.049
Nik Antropov310.46400.5140.054
Evander Kane340.467340.5450.078
Zach Bogosian300.456380.5140.058
Alex Burmistrov360.398320.5920.194
Eric Boulton280.496310.5030.007
Rich Peverley390.474180.4880.014
Brent Sopel330.48190.444-0.036
Niclas Bergfors290.488220.5130.025
Tim Stapleton80.469340.5430.074
Fredrik Modin230.379100.5350.156
Jim Slater320.465
Ben Eager310.37510.333-0.042
Blake Wheeler230.598
Mark Stuart220.561
Patrice Cormier20.238180.4650.227
Rob Schremp150.592
Freddy Meyer70.34470.50.156
Radek Dvorak120.608
Ben Maxwell120.516


We see that in the first half, Atlanta was well into the negative - only two players managed to hit 50%. Their goal differential, however, was +3 despite a 46.3% Fenwick percentage. In the second half, the story was reversed - few players were in the red. Yet their goal differential with the score tied was -2 in spite of a .529 Fenwick %. We know that Fenwick % with the score tied is a better predictor of future results than Goal %, so by these measures, Atlanta/Winnipeg could be looking at a resurgence next year.

A nice chart contributed by JaredL shows the relationship between Fenwick % and Goal % as the season progressed:




We see the Fenwick % rising as the Goal % drops. What could cause the Fenwick to jump? I can think of three things that would cause the improvement:

A: Personnel Changes - The Thrashers made a few moves towards the end of the year, they brought in Radek Dvorak, Mark Stuart, and Blake Wheeler while they shipped out Brent Sopel , Niclas Bergfors, and Rich Peverley. Wheeler and Dvorak's 2nd half Fenwick while tied definitely beats Bergfors's and Peverley's.

B: Coaching Adjustments - It was Craig Ramsay's first year coaching the Thrashers, and perhaps the players had not figured out his system until the second half.

C: Player Improvement - Dustin Byfuglien played some defense for the Blackhawks last year, but this was his first year playing defense full-time. Promising youngsters Zach Bogosian, Evander Kane, and Alex Burmistrov had not played very much in the NHL. Burmistrov's jump was especially impressive.

But what of the drop in goals? I can think of two reasons for that:

A: Blind Luck - The Thrashers simply didn't get the bounces. Over such a small sample, chance will always be a factor. No one said that hockey was fair.

B: Changing Strategy - What if the Thrashers were responding to their difficulty in scoring goals by simply firing more pucks at the net? It's possible, but I doubt very much that it would result in such a wild change in Fenwick.

Still, this change in goal differential involving score tied Fenwick is one thing, but you don't get to a 14-19-6 second half record without other things going wrong, and it seems like just about everything else did. Here's a look at their Special Teams split into first and second halves:

Special TeamsPower PlayPenalty Kill
First Half20.9%80.9%
Second Half14.0%74.3%


And here's a graph showing Fenwick shooting percentage, both for and against, for the season:


We can see, again, that the opponent's shooting percentage improves while Atlanta's gets worse.

So who are the Winnipeg Jets going to be next season? It's difficult to say. They moved to a different city and switched coaches, but the personnel are going to remain pretty much intact. The team is still in the Eastern Conference despite moving to Winnipeg, which will lead to increased travel. They've yet to sign Zach Bogosian. Frankly, I don't know. For our upcoming series on Driving Play predicting the 2011-12 season, I inexplicably ranked them as #15 in the Conference - last overall. I doubt they'll make it there, but in spite of their second half Fenwick, I still think it will be a long winter in Winterpeg.

Thursday, August 18, 2011

With Or Without You: Patrick Sharp


            The usage of Patrick Sharp has been a hot topic in Hawkland for the better part of two years now.  Sharp was used primarily as the 2nd line Center during Chicago’s cup run, and placing him on the wing of Kane and Toews confused many Hawks fans last year, especially down the stretch, when David Bolland was injured.  Loading up the top line then meant that Michael Frolik or Tomas Kopecky would play Center alongside Marian Hossa.  Center depth is still somewhat a concern for the Hawks, and this same conversation about where to put Sharp in the line-up lingers on.  Our goal is to use possession numbers to identify whether loading things up by playing Sharp with Kane and Toews actually resulted in an appreciable difference in the team’s ability to drive the play. 



            The results show that Chicago’s pure possession numbers take a slight hit when Sharp plays with Kane and Toews compared with when somebody else plays with the pair.  Sharp – Toews – Kane were together for 427 Even Strength minutes last year.  When those three were on the ice, the Hawks’ Corsi/60 was 15.155 against a Quality of Competition of -0.734.  Kane and Toews as a pair played 930 minutes with any other teammate not named Patrick Sharp, their Corsi/60 and Corsi Quality of Competition were 19.4 and 0.458, respectively.   What makes the decision to load up even more questionable is the impact on Marian Hossa.  Mr. Hossa has played 618 even strength minutes with none of Kane, Toews, or Sharp.  The Hawks still had a positive Corsi in that sample, though it was much lower, coming in at a Corsi/60 of 4.7. 

Now that we’ve established that splitting the pair up makes the most sense, the question becomes how to split them up.  The two primary options are either:

·      Other – Toews – Hossa / Other – Sharp – Kane

OR

·      Other – Toews – Kane / Other – Sharp – Hossa


Toews was paired with Hossa and somebody not named Patrick Sharp or Patrick Kane for 259 EV minutes.  The pair did well for themselves, putting up a Corsi/60 of 13.7.  This is all the more impressive when realizing this pair was generally out against the opponent’s toughest competition.  The Quality of Competition rating was 1.76. In other words, Toews and Hossa took on tough competition and still dominated possession.  Sharp and Kane were also impressive, with a Corsi/60 of 20.8.  The pair generally faced weaker competition.

The other alternative is pairing Toews with Kane and then Sharp with Hossa.  Toews and Kane (and no Sharp/Hossa) put up a Corsi/60 of 19.4 in 929 EV minutes.  The pair faced tough competition, though it was not nearly as tough as the minutes that Toews and Hossa played.  Sharp and Hossa have been together for 570 minutes, putting up a Corsi/60 of 15.5.  It is interesting to note the drop in the quality of Marian Hossa’s opponents when he plays with Toews compared with Sharp.  The Quality of Competition rating of Toews – Hossa is 1.76 compared to -0.7 when the Slovak winger is paired with Sharp. 

            The numbers ultimately bear out that splitting up Sharp and Toews is the optimal solution given Chicago’s current line-up.  From there, decisions on personnel get a bit murky, though there are some important implications.  First, is that none of Sharp’s minutes at Center can be considered tough.  This is interesting, as the defensive reputations of his two potential right wingers are quite different, yet the quality of opposition has not impacted who plays the right side on Sharp’s line.  We do, however, see a big drop in the quality of opposition when we compare Toews’ minutes with Kane to Toews’ minutes with Hossa.  The captain faced tough minutes regardless, though the 0.458 quality of competition when playing with Kane was a relative cake walk compared to the quality of competition he faced when paired with Hossa. 

            My personal belief is that the Blackhawks should pair Sharp with Kane and Toews with Hossa.  We can see that any line that Sharp centers will generally get softer minutes; given this, why not load up in both directions?  A line based Toews and Hossa features two elite two-way forwards who have shown the ability to crush territorially despite playing absurdly tough minutes; Sharp and Kane would reap the benefits of the other pair’s tough minutes.  It is also important to note that a Sharp – Kane pairing is considerably better than a Sharp – Hossa pairing in terms of puck possession even though the quality of minutes are effectively the same.

            In the end, this, to quote The Wire’s Marlo Stanfield, ‘sounds like one of them good problems’.  As a team that likes to play with the puck (and does so better than everybody else), the Blackhawks have an embarrassment of riches in top end talent.  The only thing that can undermine that would be playing Patrick Sharp on the left wing.
            

Tuesday, August 16, 2011

Hey! Speak Up!

It's been a roller coaster ride over the last 50 days since Driving Play was founded. We've experienced the dizzying highs, the terrifying lows, the creamy middles. We've gotten way more hits than we thought we would, and the right people are noticing what we're doing over here.

Still, one thing that's a little disappointing is the lack of comments on the posts. Are people not interested? That's cool - the last two articles I wrote were on Anton Stralman and Jack Hillen; I don't expect many people to be interested in that. Are people afraid to comment, maybe? A fair amount of thought goes into the analysis here, but that doesn't mean your rebuttal has to be a dissertation, complete with WOWY analysis. We're interested in hearing what the readers have to say, even if it's not full of numbers and jargon. At the very least, it could give us ideas about future posts.

So, I'm going to open up the floor: What would the readers like to see a post about? I can't promise anything, but I'm curious about what you guys are interested in.

Sunday, August 14, 2011

Part I.V: The Aftermath... Jakub Voracek and Rick Nash WOWY Analysis

Chase asked me to do some number crunching on Voracek. To be honest, I found the results surprising. The numbers indicate that Voracek is highly underrated and was a great pickup. It appears that he, not Rick Nash, was Driving Play at even strength in Columbus.

Here is how Columbus did at even strength with both goalies in net with both Nash and Voracek on the ice, one of them on and neither of them:

2010-2011 ESCorsiTime (mins)Corsi/60OppCorsi
Both67690.25.8240.499
Nash only-6405.1-0.889-0.136
Voracek only71414.310.2830.399
Neither-112453.9-0.2690.214

It surprised me that in the 20 periods worth of ice time he was without Voracek, the guy with the sixth most shots taken in the league (including PP) was negative, albeit slightly. It's too small of a sample to take the exact numbers seriously, but it's interesting to note that when Voracek was off the ice not only was Nash a negative-Corsi player but Columbus actually did better without him than with him. Voracek, on the other hand, did quite well without Nash. He did so against tougher competition.

This pattern also held if you look only at situations where the score was tied:

2010-2011 tiedCorsiTimeCorsi/60OppCorsi
Both55256.912.8440.192
Nash-13155.3-5.0230.002
Voracek15140.56.4080.66
Neither-52946.9-3.2950.021

Let's expand the sample by throwing in 2009-2010, when they spent more time apart. Here's all 5-on-5 minutes:

2009-2011CorsiTime (mins)Corsi/60OppCorsi
Both59713.24.9630.528
Nash only-611507.2-2.4281.325
Voracek only651421.12.7440.773
Neither-2534184.4-3.6280.497

and when the score was tied:

2009-2011 TiedCorsiTime (mins)Corsi/60OppCorsi
Both55262.812.5590.265
Nash-20584-2.0551.439
Voracek34508.94.0090.646
Neither-1641585.8-6.2050.468

Nash had tougher competition in his time without Voracek than vice versa, but Voracek's numbers are far better. I'll write a lot more on quality of competition later and come up with something more precise, but according to my rough calculations it looks like Voracek was somewhere between 4 and 4.5 Corsi shots better per 60 when you take the tougher opposition into account. Maybe a little higher if you only look at tied-score minutes.

Before the flood of angry emails from Jackets' fans comes in, I'd like to point out that this is all for 5-on-5 play. It appears that Nash was carrying the water on the Power Play, maybe the fans yelling "SHOOT" for the entire damn power play are smarter than we think, and he also played about 19% of Columbus' PK time. In both cases, the Jackets had better Corsi numbers when he was on the ice than off.

Expect Voracek to play a big role for the Flyers at 5-on-5 this season.