Start with the situation, not the percentage
Power plays and penalty kills describe what happens when one team has a skater advantage. A power play is the attacking team’s opportunity; a penalty kill is the short-handed team’s attempt to prevent a goal and, when possible, clear the puck or create a counterattack. These units matter, but their headline percentages are summaries of past results—not direct forecasts of what will happen next.
A power-play percentage usually means power-play goals divided by power-play opportunities. A penalty-kill percentage generally means successful kills divided by times shorthanded. Check how a source defines each figure: counting conventions can differ, especially for situations such as a goal scored shortly after a penalty begins. Comparisons are most useful when the figures cover the same period and use the same definitions.
The percentages also depend on opportunity. A team with fewer power plays may have a high conversion rate on a small number of chances; another may have more chances but a lower rate. Neither figure, by itself, tells you how often a team draws penalties, how dangerous its attacks are, or how reliably it avoids taking penalties. Separate the quality of the special teams from the frequency of the situations they get.
Look behind the conversion rate
A power play can produce a goal through several routes: a screened shot from the point, a pass across the slot, a rebound, or a shot from close range. Those chances are not equivalent to a low-traffic shot from a distance. When available, look at shot location, whether a shot reached the net, rebounds, passes that force the goalie to move, and sustained possession in the offensive zone. These indicators help explain whether goals came alongside repeatable pressure or whether a short stretch of finishing had an outsized effect.
The same principle applies to penalty kills. A strong-looking kill rate may reflect effective pressure on puck carriers, clearances, blocked lanes, and limited access to dangerous areas. It may also reflect saves on chances that could become goals in another game. Consider whether the short-handed team consistently prevents clean entries and central shots, rather than relying only on the final result.
Shot-quality measures are useful but imperfect. Different providers may define expected goals or scoring chances differently, and even a well-built model estimates likelihood rather than certainty. Treat those metrics as another view of performance, not as a definitive score. When possible, compare more than one signal: goals, shot locations, rebounds, and the way chances develop.
Make comparisons on equal terms
A repeatable comparison starts by setting the question. Are you assessing a team’s power play against an opponent’s penalty kill, or trying to understand how often special-teams situations might occur? Those are related, but they are not the same task. For the first, focus on unit performance. For the second, also consider each team’s penalties drawn and taken.
Next, use a consistent time window and note the number of opportunities behind the rates. A longer sample can smooth out some random variation, while recent games may better reflect a changed lineup or tactical approach. Neither window is automatically superior. One practical method is to review the season-long picture first, then check a shorter recent period for a clear change. If the two differ, investigate why rather than choosing whichever number supports a preferred conclusion.
Then compare the matchup’s components. Ask whether the attacking unit generates dangerous chances and whether the opposing kill limits them. Consider faceoffs, entries, puck movement, and the ability to clear the zone. A team’s overall rate can be shaped by the opponents it has faced, so avoid treating two percentages as perfectly comparable without context.
Finally, keep the expected number of opportunities in view. A matchup may feature capable special teams but few penalties, limiting their role. Conversely, a penalty-prone team can create more chances for an opponent’s power play. Penalty tendencies are not guarantees, though: officiating, game flow, and player decisions vary.
Use percentages as evidence, not predictions
Small samples can swing sharply. A few goals or saves may move a rate substantially, especially when opportunities are limited. Finishing and goaltending also add variation: a well-designed power-play sequence can fail to score, while a less dangerous chance can go in. A recent streak is worth noticing, but it does not establish a new level of performance on its own.
Game state matters, too. Teams may take different risks when leading or trailing, and late-game penalties can produce unusual situations. Special-teams numbers may also be affected by personnel changes, tactical adjustments, or a schedule that shifts who is available. If you lack reliable information about those factors, leave the uncertainty visible rather than filling gaps with assumptions.
A useful final check is to write down what supports your comparison and what could undermine it: the rates, opportunity counts, chance quality, penalty tendencies, and sample size. If those signals conflict, lower your confidence. No percentage can remove the randomness of a single game, so use this process to organize evidence—not to promise an outcome.
