Percentages vs. Timing

Jonathan Willis
June 03 2012 02:37PM

Image courtesy of Wikimedia Commons; own work produced by Resolute.

One of the most frequent counterarguments to any percentage statistic – whether it be special teams’ effectiveness, save percentage, or something else – is the notion that overall percentage matters less than careful timing.

It’s an inane argument.

Consider the current series between the Los Angeles Kings and New Jersey Devils. The Kings’ power play has been terrible for the entirety of the post-season, but they’ve managed to compensate for it with superb 5-on-5 play and a very strong penalty kill. During game two, CBC commentator Craig Simpson noted that the power play effectiveness was less about percentages than it was about timing – i.e., that how many goals the unit scored mattered less than when they scored them.

The problem is that it isn’t true.

For the whole series against the Devils, the Kings have been either tied or ahead by one goal. Both contests have been decided by a single goal; both have further been decided in overtime. Any goal, at any juncture, would have been significant – particularly since the Kings have twice blown one-goal leads. No special timing is required. All goals are significant.

The same thing holds true for goaltending. The game is within a single goal something like 70% of the time during the course of the regular season. This is a favourite topic of the Contrarian Goaltender, who noted the following about Jonathan Quick recently:

Nobody said big saves aren't important. The argument is in fact the opposite, that nearly all saves are very important. That makes them all pretty big saves, especially in the low-scoring environment of the playoffs. Is it possible to let in a goal and not switch momentum around? I think most people would answer that in the negative, which means that every save in a close game is a big save based on momentum and "keeping the team in the game". For a guy like Quick, who plays in a close game nearly every time out, that simply means that talking about big saves is pointless. Just talk about saves, period.

Based on playoff commentary, one could easily be led to believe that there are huge periods of time where players and teams can inflate their statistics. These are the tracks of time where teams and players score or allow unimportant goals, goals that make them look good but don’t help them win. It’s fairy tale thinking – of the 82 games in the post-season so far, 83% of them have been decided by two or fewer goals. 60% finished with the teams separated by just one tally. Of the small percentage of games that finished with a gap of more than two goals, the majority of the goal-scoring in those games took place within clutch situations – i.e., in a 3-0 victory, the first and second goals were obviously critical, as was every save up until the third goal.

The idea that a team can time its way to victory, scoring only at precise moments to ensure the turn of the tide is a ludicrous one. Their careful sense of timing would need to encompass the vast majority of every game they played.

It’s also an idea out of sync with reality. The Los Angeles Kings were one of the league’s worst teams in one-goal games this year – they won 17 of them and lost 29 (37.0 win percentage), a figure that sandwiched them between the 29th-in-the-NHL Edmonton Oilers (37.5 win percentage in one goal games) and the 30th-in-the-NHL Columbus Blue Jackets (35.5 win percentage in one goal games). If the Kings are a team that utilized careful timing to score when it matters most, they did a good job of disguising it for 82 games.

Recently at NHL Numbers

74b7cedc5d8bfbe88cf071309e98d2c3
Jonathan Willis is a freelance writer. He currently works for Oilers Nation, Sportsnet and Bleacher Report. He's co-written three books and worked for myriad websites, including the Edmonton Journal, Grantland, ESPN, The Score, and Hockey Prospectus. He was previously the founder and managing editor of Copper & Blue.