How to Research NHL Player Props

A forward can play a strong game, drive play for twenty minutes, put the puck on net five times, and finish with nothing on the scoresheet. A teammate can play eleven quiet minutes and collect a point because his stick touched the puck before a goal. Hockey is the sport where the gap between how well someone played and what the box score says is widest, and that gap is where most prop research goes wrong.

The NHL regular season opened on Sept. 29, 2026, which means the first weeks of the schedule offer exactly the thin, noisy samples that mislead people most. This guide lays out the variables that actually carry information, in the order worth studying them, and a routine for checking them before a game.

Why hockey outcomes are low count and high variance

Most hockey props settle on small integers. A player's goals in a game are almost always zero or one. Points, which the Hockey Reference glossary defines for players as goals added to assists, are usually zero, one, or two. When the realistic outcomes are that few, the difference between a hit and a miss is a single event, and single events in hockey are shaped by deflections, posts, a goalie's glove, and which teammate touched the puck last.

That has two consequences. First, recent results are a weak guide. A three game point streak is three binary outcomes, and a league full of players generates streaks like that constantly by chance. Sample size in player prop analysis works through why short windows mislead, and hockey is an extreme case of it. Second, the line itself carries most of the meaning. A prop set at half a point is really a question about the probability of zero, which is a very different question from whether the player is good.

Reading a distribution covers this in general terms. In hockey the distribution of almost every counting stat has a large spike at zero, and the useful research question is how much weight sits there.

Ice time is the first input

Every counting stat a skater records happens while he is on the ice, so time on ice is the exposure term that everything else multiplies. A top line forward and a fourth liner can have similar shooting habits per minute and wildly different totals because one plays nearly twice as long. Coaches decide ice time game by game, and it moves with injuries, line shuffles, the score, and penalties.

Season average ice time is a starting point, not an answer. Recent games under the current line combinations tell you more, because a player who was promoted to a scoring line last week is a different prop from the player whose season average you are looking at. The questions to ask are which line he is on tonight, whether he plays on either power play unit, and whether he kills penalties, which adds minutes that rarely produce offense.

The variables that carry information

Shots on goal

Shot volume is the most stable skater input. A shot on goal is one that either scores or is stopped by the goalie. A shot that hits the post, sails wide, or is blocked by a defender does not count. That definition matters because a player can attempt many shots and be credited with few if defenders block his lanes or his attempts miss the net. A later article in this series goes deeper on shot attempts, ice time, and deployment.

Points

Points combine goals, which depend heavily on finishing luck, with assists, which depend on who is on the ice when a teammate scores. That makes point props sensitive to linemates and to special teams, since a player on the top power play unit shares in the goals it scores. A point prop is best read as a question about the offense a player is attached to, not only about the player.

Goalie saves

Saves are a volume stat. The Hockey Reference glossary defines saves as shots against minus goals against, and save percentage as saves divided by shots against. So a goalie facing a heavy shot load can record many saves in a loss, and a goalie behind a team that suppresses shots can play well and record few. The first questions are whether he starts and how many shots the opponent generates.

The power play

A power play created by a minor penalty ends if the team with the extra skater scores. These stretches of play are where teams concentrate their best offensive players and where scoring chances cluster. Whether a player is on the first unit is one of the largest single inputs to his scoring props, and it can change without any change in his even strength role.

Goalies and availability

The goalie decision shapes both sides of a game. For skater props, the opposing goalie affects how many shots become goals. For goalie props, whether a particular goalie starts decides whether the prop exists in any useful sense. Teams rotate goalies, and the Wikipedia entry on goaltenders notes that a backup is called on to start some games so the starter can rest, and may be forced in at any time if the starter is injured or plays poorly.

Skater availability works like it does in other sports, with one hockey specific twist: a scratch or an injury rarely removes minutes from the lineup, it reshuffles them. When a top six forward is out, someone moves up a line and possibly onto the power play. Reading an injury report covers how to treat status information honestly, and the hockey version of the second order effect is usually a line promotion.

A research routine before puck drop

1. Confirm the lineup: line combinations, power play units, and which goalies are expected to start for both teams.
2. Estimate ice time from recent games under the current role, not the season average.
3. For shot props, estimate shot volume per minute from a sample large enough to mean something, then multiply by expected minutes.
4. For point props, identify the linemates and the power play unit, and ask how much offense that group generates.
5. For goalie props, confirm the starter, then estimate the opponent's shot volume before thinking about save percentage.
6. Check the schedule: a second game on consecutive nights often changes the goalie decision.
7. Compare the line with your estimate of the probability, especially the probability of zero, and pass when the role is unsettled.

The order matters for the same reason it does in basketball. Our NBA research guide argues that minutes come before production per minute because an error in exposure propagates into every stat. Hockey has the same structure with a larger random component on top. Getting ice time and deployment right will not make a one point prop predictable, but getting them wrong makes every other adjustment meaningless.

Base rates matter more in October

Early season samples are small for everyone, and hockey samples are smaller than most because the events are rare. A player who scores on a quarter of his shots in the first two weeks is almost certainly not a quarter of his shots player. The Hockey Reference glossary defines shooting percentage as goals divided by shots on goal, and it is the textbook example of a rate built on too few events to trust in the short run.

The discipline that helps is the one described in base rates in prop research: start from what players in the same role typically produce, and move toward the individual only as far as the evidence supports. Shot volume and ice time stabilize faster than goals and shooting percentage, so weight them more in the first month and let the finishing numbers earn their way in.

What this framework cannot do

This article is a framework for your own research, not a preview of a product. It can help you ask better questions, and it cannot remove the randomness that decides most hockey props. A careful estimate that a player records a point in about half of his games is still wrong half the time, and that is not a failure of the estimate.

Keep that in view when results arrive. Judge the process by whether the inputs were right, not by whether a puck hit the post. Set limits before the slate rather than during it, and if the activity stops feeling like research, help is available through the National Council on Problem Gambling and our responsible gaming page.