You Can’t Buy Talent, Except Sometimes You Can: The Best Free Agent Values Between 2005 And 2012

Updated: October 31, 2012 at 7:17 am by Brent Morris



In July of 2005, the NHL and NHLPA put aside their differences to sign a collective bargaining agreement which would no doubt stand for all time.  With that eternal and faultless document came a salary cap, which was to be set at $39 million for the 2005-06 season and would move for subsequent seasons in concert with league revenues.  Coming along with the cap was a decreased age for free agency – it was set at 31 for 2005, but would decrease over time to cover all 27 year old players and players who’d accrued 7 NHL seasons.  

No Panacea

Free agents used to be the province of big market teams alone, but one of the first coups of the new free agency was struck when then-star Paul Kariya inked with small-market upstart Nashville. We exited this CBA with the last major free agent signings being Alex Semin’s one year deal in Carolina and Shane Doan re-upping with Phoenix.

As teams have jumped into free agency with both feet, fans league-wide have realized what only Rangers fans seemed to experience in the late 90s and early 2000s:  free agency is not a cure-all. In fact, often it can create more problems than it solves, especially if the player’s contract makes him dead weight. I decided to try to quantify what free agent signings turned out best, using a metric developed by Tom Awad called GVT.  

GVT takes into account stats like goals, assists, plus-minus, and whips them into a measure which he claims to measure how many total goals above a replacement player a particular skater is worth. As with all hockey statistics, the usual caveats apply:  GVT is a poor universal stat, as it only uses traditional counting stats. It tends to overvalue players who play a lot on the power play (who thus rack up points) and penalize players who kill penalties (because during their ice time they rarely score).

However, it confers with it two advantages: 1) It’s era-adjusted, so that it knows how much a goal is worth in 2006 versus 2011, something which is important considering the not-unsizable disparity between the two and 2) It sure feels like certain GMs only value these numbers and ‘intangibles’, giving us some insight into which players performed best according to GMs’ largest criteria.


However, before we begin digging more into the numbers, here are the types of contracts I included in this study: 

– Only deals signed in July of 2005 or later were considered.  Deals signed in the summer of 2004 under which players had yet to play a game were ineligible.

– The player had to reach unrestricted free agency. If the player’s rights were traded and that player signed before July 1, as Scott Hartnell and Kimmo Timonen did, they are ineligible.

– The player had to sign with a new team. Players who were traded at the trade deadline and then signed long-term with their new team were not considered.  Teams tend to be able to value their own talent much better than other teams’, and I didn’t want that effect confounding my study. 

– The player had to sign a multi-year deal. One-year deals sometimes turn out very well, especially minimum-salary deals to unrecognized or uncertain talents, so they don’t seem terribly relevent.

– The player had to have accrued a career GVT above 8.0. I figured any player below this number would be unlikely to fulfill both conditions of having signed a multi-year free agent deal and having it turn out well. To give you an idea of just how low an 8.0 career GVT is, 192 players accrued a GVT of 8.0 or higher in just the 2011-12 season.

With these players in hand, I started my search. However, it would be silly just to look at GVT, so I threw in another wrinkle.  I used and other resources to find out the players’ salaries in each season. Unfortunately, the salary cap has only stayed static between two seasons – otherwise it has gone up, sometimes significantly.  That means a player signed in year X ends up being more valuable in year X+2 because his salary has stayed the same versus a rising cap (meaning his percentage of the possible total salary space has gone down). I ended up taking a player’s GVT, and dividing it by that player’s cap hit percentage of the total salary cap. I call this new statistic Goals Versus Threshold Cap Hit, or GVT$ for short.  

As an example: a player with a 5 GVT and a $1M salary, under a $50 million salary cap, would have a GVT$ of 250 or 5/ (1/50). I then averaged all the GVT$ values of each player for the duration of their contract to see who performed best over the entire span.  Seeing that some sorts of contracts tended to, at their best, turn out much better than others, I decided to separate out longer and shorter contracts, as well as contracts that still had years remaning on them.  So we get three lists, not one:


The best players under this metric are guys who signed two year contracts and ended up seriously outperforming them.  Some of them are quite obscure, too – I imagine even the most diligent NHL fan could forget Gelinas’s time in Florida or Park’s on Long Island.


Player Team Years GVT$
Andrew Brunette COL 2005-2007 681.25
Craig Anderson COL 2009-2011 421.44
Ray Whitney CAR 2005-2007 345.07
Richard Park NYI 2006-2008 299.41
Steve Montador BUF 2009-2011 288.55
Matt Cooke PIT 2008-2010 276.7
Martin Gelinas FLA 2005-2007 275.21
J.P. Dumont NSH 2006-2008 241.69



Six wingers, one goalie, and a defenseman. Whitney and Brunette were both undervalued talents before the 2005 lockout – Brunette was picked in one expansion draft, then traded to another expansion team for a 5th round draft pick. Whitney was claimed on waivers, then traded for a 5th round draft pick. Nevertheless, both guys still played in the league in 2011-12 at advanced ages (38 for Brunette, 39 for Whitney). Brunette’s career is likely done, but he had 733 points in an NHL career that many would’ve said couldn’t happen despite the fact that he had 334 points in 267 AHL games. He is a tribute to the power of numbers over perception.  

Cooke, Gelinas, and Park were 3rd liners who contributed more than their salaries might’ve suggested. Dumont was signed very late in free agency because the Sabres walked away from his arbitration award – he proceeded to put up a career high in points, then beat that the next season. Montador was a combination forward/defenseman who finally stuck at defense and contributed on the power play – although, if the first year is any indication, his deal with Chicago could very well end up on the worst contracts list.


Player Team Years GVT$
Curtis Glencross CGY 2008-2011 407.64
Mikael Samuelsson VAN 2009-2012 204.82
Joe Corvo OTT 2006-2010 175.13
Filip Kuba TBL 2006-2009 164.57
Brian Rafalski DET 2007-2011* 141.64
Jaroslav Spacek BUF 2006-2009 130.15
Radim Vrbata TBL 2008-2011 126.5
Marc Savard BOS 2006-2010 123.8

* – Retired before the end of contract


Four of the contracts on this list were traded before their conclusion – Samuelsson, Corvo, Kuba, and Vrbata. Curtis Glencross is far and away the best value – what in the world was Edmonton thinking when they did not tender him a contract in 2008?   Mikael Samuelsson’s numbers are inflated by the fact that he’s typically a quarterback on the power play. Ditto Corvo.

Rafalski is one of the very few signings of a star player that worked out amazingly well – playing next to Nicklas Lidstrom certainly had to help. Radim Vrbata was terrible for Tampa Bay, but when he was traded back to Phoenix he performed superbly next to Martin Hanzal.

I fear Marc Savard’s current contract will make people forget about his old one, but he too was undervalued, having been traded twice before ending up as Ilya Kovalchuk’s center and putting up points like it was his job (since it was). One can still see what effect being a little fake artist has on the NHL old boys’ club.



Player Team Years GVT$
Michael Ryder DAL 2011-? 244.34
Ian White DET 2011-? 230.59
Marian Hossa CHI 2009-? 181.6
Toni Lydman ANA 2010-? 164.46
Marcel Goc FLA 2011-? 164.09
Jordan Leopold BUF 2010-? 163.29
Erik Cole MTL 2011-? 159.2
Dan Hamhuis VAN 2010-? 145.68


Quickly, how many goals do you think Michael Ryder has in his career? If you said 197, you’re a Michael Ryder fan, probably. Ryder’s 197 goals since 2003-04 rank 24th best over that span, and half the list in front of him will end up in the Hall of Fame. Ian White’s contract looked like the best value of the summer when it was signed, and he did not disappoint (although it helps to play with Nik Lidstrom). Erik Cole’s season, on the other hand, was a shock to many – he set career highs in goals and shots on goal at age 33.

Jordan Leopold was Lydman’s replacement in Buffalo, but it seems both have worked out quite well for their new team. Marian Hossa is one of two superstars who signed a massively front-loaded deal with a new team (the other is Brad Richards) – it’ll be interesting to see if Zach Parise and Ryan Suter appear on this list a year later, Hamhuis passed on offers from the Islanders, Pittsburgh, and Philadelphia to sign with his hometown Canucks. Marcel Goc ranked 5th and fellow Panther Tomas Fleischmann ranked just off the list – so not all the Panthers’ signings last summer went horribly wrong (yet…).


As I stated at the outset, this list would undervalue defensive defensemen, and they are certainly not represented well here.  However, another surprise is that goalies are not represented here at all besides Craig Anderson.  Goalies are typically both the top and bottom of the GVT list – they have the greatest impact on who wins and loses games.  Still, GMs are apparently not finding gigantic bargains there, at least not in multi-year deals with goalies from other teams.  We see fewer star players than I would have expected, but then again star players tend not to change teams all that often, and when they do,  the amount of suitors drive the price up considerably.  We’ll see more of that effect with Part 2 in this series:  The Worst Contracts Of The Post 2005 Lockout Era.