According to the CBA, a player signed to a multi-year contract can sign a new contract beginning on July 1 the year before his current contract expires. A player signed to a one-year contract is eligible to sign a new deal beginning on January 1 before his deal expires.
Unfortunately, the Internet isn't that good at logging contracts before 2008, but the thing that's interesting to me is that we're not just seeing star players sign deals a year before their current one expires - 'star' players who've done this recently include Zdeno Chara, Tyler Myers, Chris Pronger, and Marc Savard. We continue to see it even though there are plenty of cautionary tales for both player and team - Marc Savard took his brutal blindside hit from Matt Cooke later in that season, and has only managed to play 25 games under his new deal. It's very likely that Savard will never play again. We've also seen the player get screwed over - Jeff Carter was traded from Philadelphia to Columbus before the 11 year deal he signed with Philly even took effect. It's very likely that had Carter not signed that deal, he would not have been worth the 1st round pick and Jakub Voracek that Philadelphia acquired in exchange for Carter - Columbus would have had no idea whether he wanted to stay there. Now he's stuck there.
What's more interesting to me is that this trend has trickled down to less important players. I already mentioned Peverley, but the Capitals locked up Jason Chimera for 2 more years a few weeks ago. Why? I don't know. Chimera played 13 minutes a game last season. Still, the Capitals saw fit to give him a 2 year deal worth 1.75M per season that doesn't take effect until next year.
Let's look at the pros and cons of this move from the team's perspective.
Get The Player Locked Up - This is self-evident, but it needs mentioning. There's no need to scour the trade market or free agency at season's end, you know this guy is going to be on your team.
Perhaps Get A Discount? - The earlier a player is signed, the more vague next season's cap number will be. If a player is signed in July of the year before his contract expires, he and the team will likely be negotiating a contract based on the current year's salary cap number. The cap has gone up every year, and only once by an insignificant amount - the team is likely gaining some sort of savings. Remember, for teams that have effectively unlimited budgets (Toronto, Vancouver, Montreal, Philadelphia, New York R., Washington, Pittsburgh, Chicago, Detroit, Buffalo), this is huge - it's not a question of how much money you spend, but how much money you spend relative to the cap. If a guy is theoretically worth 1/30th of the cap today, and also worth 1/30th of the cap next season, signing him at 1/30th of the cap now means that he will be more valuable next season when the cap rises - there's going to be $X additional dollars available in the budget.
Give The Player Additional Trade Value - There's a sizable portion of the NHL that has trouble attracting free agents. Either the team is bad or plays in an undesirable place. It's disingenuous for teams to sign players to contracts they have no intention of honoring, but it's a practice I expect to see more and more often. If you're a team that struggles to sign players, getting a guy who's been extended is almost certainly more favorable than getting a rental.
As More Teams Adopt This Policy, The Price of All Available Free Agents Rises - This seems a tad paradoxical, but bear with me. Each early signing takes another potential body off the market. It also incrementally raises the price for each body that ends up on the market - teams always need players, and since there are so many teams who effectively don't have a salary cap, those teams are always looking to improve if they have available dollars. Furthermore, a team that loses a player to free agency will often try to replace that player via free agency. Demand always outstrips supply when it comes to free agency.
This leaves teams in a bind - they may want to risk that their potential UFA player gets to market for fear of overpaying him. On the other hand, they may already see that potential replacements either aren't on the market now, or quite possibly won't be on the market when July 1 rolls around. Furthermore, even if their replacements are on the market, they may well be overpriced because of demand. One would think that the market would correct for this, that players would recognize that they figure to make a lot more money by waiting until July 1 and playing all the bidders against one another. However, it's become clear in the salary cap world that not every player is out for the biggest dollar amount - many are willing to give up the promise of potential dollars if they're being compensated fairly and they're comfortable on the team, in the city, etc. The players who are unhappy in their current situation are the ones who figure to reap the greatest financial rewards.
Risk. We've already seen more than one of these deals go bad. I don't think the Bruins are being significantly affected by the Savard signing, but it can still be a pain to work around these things, as it can dry up the amount of cap room a team has at the deadline. Yes, it's true, if you have a player on injured reserve, you can replace his salary with People yelled and screamed about the Flyers never getting anyone at the deadline from 2008 to 2010, but that was for a simple reason - they had all kinds of dead IR money on their cap preventing them from banking cap space and thus discouraging them from making any but the most minor of moves. That's of course not the only risk - the other risk is that the player's performance declines precipitously. or that they suffer a calamitous injury before their current contract is up. Tom Poti was signed to an extension by the Capitals during last season's training camp and has played 21 games since then. It's thought that he may have to retire.
That's really the only downside, though, and for teams who figure to spend near the cap, it's not much of one. I expect to see more and more of these signings going forward.
I can see why teams are locking up young players like Tavares, Van Riemsdyk, and Myers to contracts like this - it's entirely possible that those players will improve greatly this season, making their value difficult to ascertain. The player and the team are each taking some risk there (with the exception of Van Riemsdyk since that contract is insane). However, with older players, the risk is more on the team than on the player - the player probably isn't passing up very much money by signing early. We've already seen some of these deals go bust before they start - it will be interesting to see which teams are either emboldened by success or dissuaded by failure with these sorts of contracts, and who's still signing them several years from now.