Ten years ago last month, NHL fans got their first taste of the shootout. The 2003 All-Star Game remained tied after three periods and a scoreless overtime, and though a regular-season match would have ended at that point in a tie, the All-Star rules called for a shootout to determine a winner. The All-Star Game is just an exhibition, after all, so a shootout made for a novel, crowd-pleasing way to settle things. But some two years later, to the dismay of purists, the shootout made its regular season debut. Arenas buzzed as fans got their first look at the new normal: alternating penalty shots to determine the winner of meaningful hockey games. I'll admit, I reluctantly accepted the shootout upon its introduction -- not because I necessarily liked the idea of settling games with a skills competition, but because I loathed ties. It goes against the very nature of competition, which is supposed to ultimately determine a winner and a loser. It's why these games exist; to end the game without a winner always felt like a grand waste of everyone's time.

But after seven-plus years of regular-season shootouts, the novelty has worn off. The most exciting part of an overtime game, in fact, is the five-minute four-on-four period, which invariably involves lots of open ice and fast, back-and-forth action. The shootout has become almost anti-climactic, but there's no reason a close hockey game can't peak at the very end, while still being decided by the full range of hockey activities -- not just shooting and goaltending, but also passing and hitting and, you know, defense. So even if the shootout isn't going anywhere -- its biggest selling point is that it sometimes still provides bonkers, highlight-worthy goals from shootout artists like Pavel Datsyuk -- let's dream of what could be in an alternate universe in which the shootout is outlawed. Let's build a better overtime.

My dream no-shootout scenario is based on the following three beliefs:

1. Ties are dumb, so we need a system that results in a winner.

2. It's okay to make a hockey game a little longer. Fans and media types often shudder at the idea of extending the game, but a typical regulation hockey game lasts about two and a half hours, which is pretty reasonable, at least compared to baseball and football games that routinely last longer than three hours. The NHL can afford to institute a system in which regular-season games could potentially last a bit longer, as long as players aren't pushed to the point of exhaustion.

3. The points system needs to go. No more degrees of losing, as is the case under the current system, where a loss in overtime earns you a point. (And certainly no degrees of winning, as is sometimes suggested, where a teams gets three points for a regulation win, and two points for a "lesser" win in a game that extends beyond the traditional five-minute overtime period.)

Currently, the best part of an NHL overtime is the five minutes of four-on-four action that immediately follows the end of the third period. Instituted in 1999, it was designed to open up the ice and increase the likelihood of a sudden-death winner, back during the days of ties. Problem is, the way overtime is formatted now, it can't extend much longer: The ice is already choppy at the end of regulation, and to play much more than five minutes would mean finishing the game under unacceptable ice conditions.

So in my dream overtime, there'd be an intermission following the end of regulation to allow the Zambonis to resurface the ice. It can be shorter than the intermissions between regulation periods, since there's no need for human pucks races or car giveaways that hinge on an impossibly difficult center-ice shot. An intermission will add to length of the game, but it does so at a time when the game is, by definition, exciting (or at least, close and at a critical juncture). Fans in the building won't leave, and those watching at home would be crazy to change the channel.

Overtime would begin, as it does now, with five-minutes of four-on-four. But if that doesn't yield a goal, they'd take advantage of the clean ice surface to then play five more minutes, this time at three-on-three. The chances of a game-winning goal would go up even higher in this scenario, as the three-on-three would mean even more open ice. Is it a gimmick? Yes. But it still gives the teams a chance to win the game during something that resembles real hockey (with offense, defense, etc.), albeit with a sped-up game flow.

These periods, of course, can't go on indefinitely, as they do in the playoffs -- not only for the players' sake, but for the fans'. An occasional hockey game with a maximum length of, say, three and a half hours, is acceptable. But a hockey game that could go on indefinitely needs to be reserved for the postseason, when the stakes or so high that fans will endure five-hour marathons if that's what it takes to determine a winner. It's why the shootout exists in the first place: In order to guarantee games don't end in ties without continuing play until someone scores a sudden-death goal you need a quick-ish way of determining a winner, and a shootout, for all its flaws, accomplishes that. But a shootout isn't the only option.

Since we already live in a world with a gimmicky overtime system, why not create a better gimmicky system -- one that puts teams in real hockey situations, rather than simplifying the game down to shooters and goaltenders? And so after five minutes of four-on-four and five minutes of three-on-three, I'd implement a system similar to the one used in college football, only in this case, teams would alternate power plays until a winner was determined.

In a world where time was no concern, each team could be given a two-minute power play, and they could alternate until, in a given round, one team scores and the other doesn't. But this format runs the risk of taking a while if teams respond to goals with goals of their own, and match penalty kills with kills of their own. There's a simple solution, though: If the first team fails to score and the second team lights the lamp, the second team wins. (Any shorthanded goal would also end the game.) But if the first team scores, the second team would not only have to score themselves; they'd have to score faster. So if Team A scores 50 seconds in its power play, team B would get just 50 seconds of power play time. If they beat the clock, they win. Otherwise, they lose. In other words, the game would end, one way or the other, in the first round in which either teams scores. And to increase the chances of a goal, each round would begin with an offensive-zone faceoff (just like a real power play), and subsequent rounds could be played with more open ice: a round of five-on-four, then a round four-on-three, then rounds of five-on-three until there's a winner.

Think of the atmosphere in an arena during one of these OT power plays: A home team trying to kill off a power play and secure a victory while a crowd loses its collective mind each time the puck is cleared past the blue line. Or, alternately, the crowd watches as the home team, after a defensive stand, gets its turn on the power play with a chance to win the game with a goal. Or perhaps the home crowd knows its team not only needs a goal to match the opposition's, but needs to get that goal in less time. Any one of these is preferable to the shootout. The drama is still manufactured -- the format is designed to produce tension -- but at least the outcome is the result of a team effort, rather than a series of individual efforts.

(Oh, and if team with a really effective power play wants to stall in order to reach that part of the game, they'll have to do it during chaotic periods of four-on-four and three-on-three play. In other words, it's a dangerous strategy.)

The final element of this dream overtime involves raising the stakes in each game in order to maximize excitement. As it stands now, a win is marginally better than an overtime loss, which is itself marginally better than a regulation loss. But without a points system, there'd be a bigger difference between winning in overtime and losing in overtime. (It would also make the standings easier to understand. Since standings usually don't include a team's points-per-game number or its points percentage, math would no longer be required to figure out whether Team A is really ahead of Team B if it trails them by two points but has two games in hand.)

Of course, all of this is purely hypothetical. The shootout isn't going anywhere, if only to provide occasional material for highlight reels: Patrick Kane's slow-motion fakes from earlier this month, or Kaspars Daugavins using the tip of his stick to carry the puck towards the goal before being denied on a spin-o-rama attempt. But a guy can dream, can't he?

***

Joe DeLessio is a senior producer at New York Magazine's website, nymag.com. Follow him on Twitter @joedelessio.