There is a game that hockey coaches will play with their teams as a reward for a focused and hard-working practice called "hockey poker." To set the game up, two teams are formed. Then there will be five shifts. During those five shifts, both teams will be forced to put out different numbers of players for each one. On one shift, a team may put out five players.

On another, a team may put out two. Over the five shifts, the teams must each use all five numbers between one and five. So, much of the game's strategy applies standard hockey tactics boiling down to playing better than the other team. But some of the fun comes from trying to gain an advantage by mixing up the number of players you put out and making sure you take advantage when you have the "power play."

This article is not about that game.

Instead, it is about what we can learn from poker when thinking about hockey strategy. Poker is a game of incomplete information, meaning there will be definite facts about what is happening as each hand plays out that some players are aware of, and others are not. The primary way this occurs in poker is that players can see their hands, but not other players' hands.

It seems that this makes the game harder. However, in games against other players, the difficulty of winning always comes from the skill of the opposition. Just like you, in poker, they have facts about the game they are not aware of. How can you best take advantage of that?

Bluffing is an excellent example of a way to use other players' lack of knowledge of your hand to your advantage. For those unfamiliar with poker, bluffing is where a player makes a bet as if they have a strong hand when they have a weak hand. In fact, in the most straightforward way, the purpose of bluffing is to give a weak hand a chance to win.

How so? If a hand gets played out to the end, players will show their cards, and the player with the strongest hand wins "the pot," which is all of the bets made by all of the players during the hand. If a player believes that they have a weak hand (at least weaker than their opponent), they will not want to get to this stage. Instead, they will either want to give up and save their money ("chips") for a better hand (by declining to match the bet of another player and choosing to "fold" their hand), or take a stab at winning by betting as if they have a high hand so the other player will give up.

However, this is not the only purpose of bluffing. Consider the following maxim from the world of poker below. 

Many people view poker as a game of interpersonal dynamics. They feel it is the player with the most social savvy that will be the winner. This can certainly be true. In this view, it's a game where you try to get a "read" on the other players so you can decide to bluff when you know they will fold and avoid bluffing when they know the other player has a strong enough hand that they can't be made to fold.

Yet, this other concept, namely that if you never get caught bluffing, you aren't bluffing often enough, flies in the face of this idea. Why is it believed then? One reason this myth carries weight is that if you get good at poker, you'll want to play where the big money is, and that means other good players. These players know how to keep their thoughts from showing on their faces, so that level of a perfect read isn't plausible. But there is another reason.

Imagine you have a high hand, and you are betting it in a straightforward manner (basically, making large bets because you believe you have the best hand). If you have been caught bluffing on a past hand, there will be doubt in the other players' minds about whether you have a high hand or are just pretending to. They can't merely be confident that you do indeed have a high hand and decide to stop matching your bets and fold. As a result, there will be cases where you win more money on good hands like that. So, by losing money on a bluff that didn't work, you can gain more money later.

Yet, being caught bluffing feels embarrassing. Because of that feeling, it is hard to fathom that to be a high-level player, you literally have to get caught bluffing. But it is true.

From a different angle, we can see this as a "right play" to make in any given situation. This is the play that you expect to, on this hand right now, allow you to have the highest "expected" amount of chips at the end of the hand which is the average amount of chips you'd have at the end in all the possible ways the hand could play out after your action. Since it is a game of incomplete information, you can't say you should make the play that will lead to "the maximum amount of chips at the end of the hand", as your lack of information means you can't know that for sure. Instead, one should use the idea of expected chips to account for the possibility that for any play you make, you can be wrong about what the other players will do based on what hands they have.

Regardless, there is a play that would yield the most expected chips that hand, and this can be considered the "right play" for that situation. You often won't be able to know for sure what that play is, but it exists. Yet, a poker player should not make that right play every time. Why? As we saw above, doing this makes you predictable. Bluffing in a situation where the other player likely has a strong hand is often not the "right play", but it should be done, albeit rarely. First, every so often, it will turn out that the other player was bluffing, and you'll wind up winning the hand. Second, you'll get caught once in a while, which will help you get paid off when you have a strong hand later. And, it's not just bluffing. In other situations, the right play will have to be eschewed once in a while to ensure that you are generally unpredictable.

Yes!

Let's consider a simple situation to make the point.

Minnesota Wild center Mikko Koivu has a leading shootout success rate since it has been instituted in the NHL to end tie games in a timely fashion. He'll employ a few different approaches to the net to mix things up, but a significant majority of the time, he uses what he believes to be his best move.

How does this work? He gathers speed before touching the puck, coming in a maybe 75% of top speed. Once he reaches the puck, he heads out wide to his forehand side (left side as he is a lefty) and controls the puck on the forehand in a position where a shot could come at any time. The width he gains to the left side is enough to get him roughly in line with the face-off dots before he curves back in angling toward the net. As he gets to about the hashmark level, he pushes the puck toward the goalie. As the puck gets about two feet in front of his feet, he pulls it across to the backhand using the forehand toe and then with tremendous skill and supple touch, lifts the puck over the goalie's left shoulder to the top right corner of the net.

Every goalie he faces expects this move. In fact, they seem to fear getting beat by it because of the very fact that people know that the goalie knows it's coming, and it is embarrassing to get beat by something you knew was coming.

But, did the goalie really know it was coming? 

No. The goalie knew it was likely. But, goalies cannot just wait for him on the right (from Koivu's perspective) post because Koivu would just shoot the puck in the open half of the net in that case. But, goalies can cheat. They can load up on the right leg so that when he starts the expected version of the move, they can push off super explosively and get over to cover the top right corner. Do goalies do this? Probably. But they can't completely commit to that plan either, because he does shoot on the forehand sometimes. And he can pull back to the forehand after pulling to the backhand as opposed to going immediately upstairs.

We can assume that both of those alternative options are lower-percentage for him than the one he uses most (otherwise, he'd use them more). So, what would be the "right play"? The right play would give him the best chance of scoring on this time down the ice. Thus, the standard move to the backhand and up to the top right corner would be the "right play". But, if did that every time, the percentage of time that it would work would go down. So, in order to get the best result over the long term, he has to mix in other options that aren't the right play. He has to avoid being predictable.

I recall another Minnesota Wild player in a radio interview the day after a tough loss. The decisive goal was scored on a rush by the opposition that started when this particular Wild player had the puck along the wall by the hash marks in the offensive zone. During the interview, the player analyzed his own play indicating that he should not have pushed the puck to the point from that spot because the coverage was a little too close. The coverage was able to pressure the point man into not being able to handle the puck, starting the rush that led to the game-winning goal.

I have no objection to the idea that a mistake was made there. I did object to the player's next comment, though. He said that the "right play" was to cycle that puck back into the corner. Notice I did not say that I was correct to object. He may have been right in that there really was only one right play there. Indeed there are some situations in hockey where there is only one correct play. For example, imagine your teammate sets you up with a back door tap-in. Should you put the puck in the net in that spot, or should you "mix it up" and be "unpredictable"? Just put the puck in the net! That situation has only one right play.

But in most situations in hockey, you are a long way from the net, and you are balancing offensive opportunity and defensive risk. In these situations, there is often a "best" play. Just as often, figuring out what would be best in the heat of the moment is impossible, and players go for the first "good" play that comes to mind. However, even if the best could be figured out, for most situations, one will see in the game, it should not be used every time.

When balancing offensive opportunity, defensive risk, and deciding what to do, there are often several good options. The best option should be used the most, but hockey teams gain considerable advantage by being unpredictable to opponents (especially if teammates have a good feel for each other as players so they can remain somewhat predictable to teammates). So, they need to periodically use the second, third, and maybe even fourth-best options to make sure the other team has to respect all possibilities!