Colby’s Contract: the Great Debate


Somewhat surprisingly, there has been a fair amount of debate over yesterday’s signing of Colby Armstrong.

I find this interesting because  much of the criticism seems to revolve around the notion of $3m equating to more than a 15 goal career average, even though Armstrong wasn’t exactly acquired for his offensive prowess.   The main criticism seems to be, why would the team be willing to make a $9 million investment over 3 years, when similar production can likely be found at a cheaper rate?

Now, it seems most decided to stop at that point and take the easy road; that being negativity for the sake of negativity (a known idiosyncrasy of Canadian hockey fans).  But instead of screaming “WHY did they sign him?”, I propose a different question:  Why DID they sign him?

A lot of interesting and useful numbers have been presented in much of the analysis, and provide an excellent starting point for this discussion.  Here are a few of Armstrong’s key advanced stats, courtesy of and

Advanced Stats, Colby Armstrong
Rank among ATL forwards with 20+ GP
Even Strength82-19-17-3679-14-10-24
Standard +/-+6+5 (5th)
Avg TOI15:0914:47
GVTN/A4.5 (5th)
QUALCOMP0.054 (2nd)0.018 (7th)
QUALTEAM0.014 (6th)-0.119 (12th)
CORSI0.77 (1st)-1.02 (5th)
OFF.ZONE STARTS40.3 (10th)47.1 (6th)

(For the lay, CORSI can be summed up as “zone time” or “possession time” and is a very useful measure in conjunction with plus/minus. GVT, or goals versus threshold, takes into account offensive, defensive, goaltending and shootout stats to evaluate individual players against one another (e.g. value of a player above what a replacement would have contributed).  Offensive Zone Starts refers to the ratio of offensive to defensive faceoffs for which the player was on the ice).

So what do the statistics show us?

Highs and lows.  On the surface, the numbers reflect a player who has received consistent 3rd line ice-time, who just two seasons ago showed he can produce positively in both offensive and defensive aspects of the game. The numbers also show a player who struggled mightily to repeat that performance the following season.

Last season’s dip in Quality of Teammates could arguably provide some insight into Armstrong’s CORSI dropping back into the negative, and at least partially account for the decrease in offensive output.  It is noteworthy, after all, that both his CORSI and GVT ranked fifth among team forwards despite the actual numbers.

However, when compared to numbers from across the NHL, Armstrong’s advanced stats aren’t exactly anything special.  And yet the Leafs‘ brass targeted and signed him anyway. Obviously not because the stats showed he had an average year last year. Rather, they saw something in his game they liked, a fit with the type of team they are trying to build. The key word being “saw”.

What I’m getting at is, while statistical measures are an important part of the evaluation of a player, a powerful set of  tools which provide evidence to the tangible, or rather the numerically-translatable, abilities and attributes of a particular player, they don’t tell the whole story.  Statistics provide the quantitative … but what about the qualitative?  And in the excitement over numbers and visual comparables, it can be easy to allow the quantitative to dominate the analysis.  After all, the numbers are all right there, whereas detailed video footage, aside from a couple of YouTube clips here and there, isn’t quite so easy to locate.

To elaborate, the stat sheets don’t tell you how a player plays the game, and provide little evidence of his style or fit within a certain system or philosophy. To wit, they don’t tell you that Alexei Ponikarovski played on the periphery and rarely drove the net. They don’t tell you that Mikhail Grabovski likes to carry the puck through five players at a time or that Jason Blake had a thing for drawing circles around the rink with his skate blades.

More specifically, statistics don’t tell you how much time a guy spends in front of the net, his willingness to engage in puck battles along the boards or in the corners, the percentage of those battles he wins, or the emotional impact he creates off a well-timed, well-executed physical play. And they don’t take into account the immeasurables — such as spirit, work ethic, creation of momentum. Such things cannot be accounted for numerically, and yet are essential to a player’s impact on his team.

In the case of Armstrong, I would argue that what he brings to the table are the immeasurables. This is not to say those quoting the measurable/quantitative stats don’t have a point — theirs is, actually quite valid — however, I would suggest that were the team focused solely on those measurables, they more than likely would not have signed him. Let’s face it, by league standards his numbers — in all categories — were average last season at best.

But they signed him anyway. They set an upper limit to which they were willing to spend, and spent it willingly, fully aware of every statistic available.

Now why would that be? Could it be those immeasurables, those qualitative attributes which can only be evaluated by watching a guy play, and by analyzing the tape? Dare I say the dreaded word, “intangibles”? (For the record, I hate that word. Terribly overused.)

Don’t get me wrong, I am not in any way shape or form knocking the use of advanced statistical measures. They are a wonderful, practical, and extraordinarily useful tool. But they are but one tool in the toolbox, and do not (read: cannot) complete the picture of a player’s full ability to impact a game within his role. I won’t go so far as to suggest the tale of the tape is a more important tool; but I will say it is — in my estimation — of equal importance.

Let’s put it this way: what exactly are the adjectives most commonly used to describe Armstrong? Grit. Heart. Work Ethic. Character. Leadership. Intangibles. All attributes which are nearly impossible to quantify into some form of numerical measurement.  And all attributes that can only be detected by actually watching the player, be it live action or video footage.

Coincidentally, what are some of the adjectives Burke has used to describe the types of players he wants on his team (other than “truculent” and its synonyms)?   Exactly.

In Burke’s philosophy, immeasurable attributes may not show up in the statistical output but are every bit as important to building a winner as the measurable output.  And those very immeasurables, sorely lacking among the Leafs‘ forward corps the past few seasons, are the reason Armstrong was made a top priority.

Well. Now that my take on the criticism of the player is out of the way, let’s take a peek at the contract. Hoo, boy. Here goes:

Prior to the opening of FA, the general sentiment seemed to be that Armstrong would likely go somewhere in the range of 2m to 2.5m, depending on the number of bidders. I would imagine we can all agree on that.

As we all know, a team identifies the players in whom they are most interested based primarily on the feedback of its scouting staff in accordance with the philosophy of the General Manager. And in most cases, an upper limit is set in terms of how much the team is willing to spend. I’m pretty sure we can all agree on that too.

Armstrong made an interesting comment when TSN interviewed him, to the effect of the Leafs offering the most money. That tells me that if the rumours of the expected bidding to go to 2.5m on the high end were in fact true, the Leafs would have upped the ante to ensure they got their man. Hence the 3m price tag at the end of the day. Admittedly, that’s pure speculation on my part … but the logic works.

Here’s where it gets tricky. As much as I respect a team for getting a player they clearly had high on their wishlist from the outset, I would say that 3m as an upper limit on what they were willing to spend was, yes, probably a bit too high. I have a feeling that if he’d signed at 2.5m, this conversation would be far more muted as we’d all be busy talking about how he produces more than Malholtra for the same rate. Unfortunately, that is not the case and here we are.

The flipside is, FAs do come with a premium. It’s the same story every year, and by that very token we shouldn’t really be surprised he was overpaid. Why should this year be any different than July 1sts past?

Hence the dilemma.  And the debate.

Do I like the player? Yes.

Do I like the money? No. He is definitely getting overpaid. Few will argue otherwise.

Do I understand why the dollar figure went as high as it did? Yes. Nature of the beast that is July 1st.

Overall, am I good with the deal? Yes … because it is only 3 years for a guy who is 27 going on 28, which allows the team a certain degree of future flexibility. While I would have preferred fewer dollars, $3 mil in and of itself is by no means a cap killer … unlike, say, having 26m-plus (aka 44% of the cap) tied up in 7 defensemen.

The bottom line is, the statistics show us a player who at best can have a decent two-way impact on the third line, and at worst is unlikely, based on trends, to be terribly far into the negative in any one category.  Is that worth $3 million per? Probably not.  However, the tale of the tape shows a player whose qualitative attributes (the “immeasurables”) fill some sorely needed gaps among what has been, the past few seasons, a rather lackadaisical forward group.

That’s still not worth $3 million per, but I find it renders the contract much easier to swallow.

Looking forward to your thoughts as always,

[email protected]