With less than a minute remaining in Saturday night's Red Wings-Kings game, Detroit's Niklas Kronwall took a shot that deflected off the stick of a member of the Los Angeles Kings. Like a million deflections before it, the puck flew high into the air and bounced off the netting above the glass behind the net. But in a series of events that Jordan and Bird would appreciate, the puck hit off the netting at just the right angle such that on its way down, it struck goalie Jonathan Quick in the back and caromed into the goal.

This should have been nothing more than a blooper, like a basketball player launching a half-court shot well after the buzzer, just to see if he could sink it. Play, of course, should have been stopped once the puck touched the netting, so the once-in-a-lifetime series of bounces shouldn't have had any real impact on the game, in which Detroit was trailing by a goal when Kronwall fired his shot.

But the on-ice officials didn't see the puck hit the netting, and therefore didn't whistle the play dead. And because they didn't, the goal had to stand: Even though replays immediately showed what had happened, whether a puck has gone out of play isn't reviewable. And so despite clear evidence that the goal shouldn't have counted, the referees had to rely on what they thought they saw in real time, and couldn't make it right.

In other words, it's exactly the kind of scenario that replay reviews are designed to avoid. A flaw in the NHL's system was revealed on Saturday night. But it can be fixed.

The NHL has long understood the benefits of instant replay: It first adopted a review system in 1991, after the NFL but before the NBA or Major League Baseball. The early days of replay were a far cry from the high-tech "war room" of today, with its countless high-definition replay options. Because not every game was televised in 1991, replay officials sometimes had to rely on the video recordings that teams would make for training purposes.

Replay was limited to a handful of things related to whether a goal should be counted: Whether the puck had crossed the goal line, whether it had been kicked or thrown into the goal, whether it went off an official, whether it crossed the goal line before the net was dislodged, and whether it went in before time expired at the end of a period. At the time, the NHL used video goal judges in each arena. Since then, replay has been expanded, but not by much: It can now also be used to determine whether a puck was hit with a high stick on a potential goal, and to establish that the official game clock has the right time. (Replays now all go through a centralized center in Toronto, rather than through goal judges in each arena -- a system that works so well that Major League Baseball will use a similar one when it rolls out expanded replay next season.)

The purpose of NHL replay has more or less remained the same for more than 20 years now: Other than the use of video review to determine the proper time on the clock, it's designed to make sure that good goals are allowed to stand, and bad ones aren't. So here's a modest proposal: Let's tweak the system to make sure that all the good goals really are allowed to stand, and that all the bad ones aren't.

If an event that immediately precedes a goal should negate it, the referees should be allowed to ask for a review, even if it doesn't fit one of the criteria already on the books. If a puck hits the netting above the glass and somehow bounces in, like it did on Saturday night, let a ref look it over and retroactively stop the clock at the moment the puck went out of play. Or, to give another example, if the ref believes there's incidental contact with the goalie when there is none, he should be able to see a replay from a better angle before a potential goal is waved off, in case it turns out there's no contact at all. (Currently, the on-ice officials huddle to discuss what happened in such a situation, meaning the ref that made the call can consult colleagues who may have had a poor view of the play, but not with video that might show it from a perfect angle.)

Could this create confusion? Perhaps. But even if the rule book can't anticipate every scenario, referees should be allowed to exercise some common sense. There's some wiggle room on what the exact details might look like -- but that can be worked out. For instance, if replay was expanded to include any event that directly impacted a potential goal, officials would have to judge what exactly that meant. But ultimately, such a system would allow referees to correct more types of mistakes, which means getting more calls right.

Consider an actual example from last season. This Matt Duchene goal -- in which he was a mile offside but was allowed to skate in on a breakaway because the linesman misinterpreted what constitutes a "back pass" from the defending team -- was allowed to stand, and the resulting clip was widely mocked in the days afterwards. The play couldn't be reviewed, but the call could have easily been overturned if it was. (The NHL later admitted the linesman blew the call.) Offsides calls wouldn't be universally reviewable, and ultimately a decision would need to be made on what would happen if a goal was scored 10 or 20 or 30 seconds after a missed offsides call. Still, officials should be given the leeway to review, and disallow, a goal like Duchene's.

Some things could still be tough to fix: There's nothing a ref can do after the fact if he blows a whistle too soon and the puck crosses the line shortly thereafter, since there can't be an expectation that players continue to play after the whistle sounds. (It's this logic that will keep the "intent to blow" rule on the books and unreviewable as long as a whistle and the intention to blow one remain one and the same in the eyes of the NHL.) But the league should at least fix what it can.

A more drastic overhaul of the league's replay system, one that could look at anything from penalty calls to icing disputes, could be tricky in a sport in which the flow of play is as fluid as it is in hockey. But fixing the replay system to do what it was intended to do in the first place is an easy next step. Expanding reviews of potential goals fits in perfectly with the spirit of the pretty-good-but-still-imperfect system that's already in place.

The basic argument for replay is that officials are human and make mistakes, but since the technology exists to catch those mistakes, it'd be silly not to use it. As any viewer of Saturday night's Red Wings-Kings game knows, the technology exists to show if, say, a puck has bounced off the netting above the glass, off the netminder's back, and into the goal. So let's let the referees use it.