Most Read: Sports

http://www.washingtonpost.com/2010/07/06/ABMK8PP_linkset.html
Complete hockey coverage
On Twitter Twitter: kcarrera and PostSports |  On Facebook Facebook |  E-mail alerts: Sports and Redskins |  RSS   RSS Feed
Posted at 01:29 PM ET, 12/04/2010

Statistical analysis: Which Alex is more talented at scoring goals?

There is no debate that both Alex Ovechkin (named the best player by his peers three out of his five full seasons in the league, including the last two) and Alexander Semin (on pace for a career-high 55 goals this season) are two of the NHL's most talented players, but which one is more talented at scoring goals?

Goal scoring is influenced by luck -- and by "luck," I mean random chance, like when a puck hits the crossbar and makes its way into the twine, or when a deflection skips (or does not skip) past the goalie's blocker. Certainly there is skill involved, but how much more skill one Alex has over the other is what we will try and answer.

To determine how much skill goes into each Alex's end-of-year goal totals I used a Monte Carlo (random) simulation to estimate the random variance, which is just a fancy way of saying, "how close are these numbers to the average?" The larger the variance, the more the results spread away from the average. In this model, all shots on goal have the same chance of lighting the lamp, regardless of who the goalie is or which teams are playing. Then, I took a look at how these purely random goal totals differ from the actual goal totals for both Ovechkin and Semin.

Each Alex was given the same number of shots per game as he had in the actual NHL seasons he played, but the chances at a goal were based on the 9 percent average shooting percentage of the past three years. After simulating 100,000 careers for each, I looked at how the randomly-generated goal totals were distributed (calculated as standard deviation) and compared them to their real-life counterparts.

  StdDev (random) StdDev (actual) % Luck % Skill
Ovechkin 0.65 0.80 66% 34%
Semin 0.48 0.68 49% 51%
To find out the percentage attributed to random chance divide StdDev(random) 2 / StDdev(actual) 2


Ovechkin's simulation indicates that up to two thirds (66 percent) of his goals scored could be a function of luck. Semin's simulation, on the other hand, indicates his goals scored are almost a 50/50 split of luck and skill.

This would explain Ovechkin's frustration and Semin's ability to terrorize goalies almost at will.

Ovechkin, who may rely on chance as a majority of his ability to score goals, is less fortunate on fewer shots. His shooting percentage of 8.5 percent this season is almost 50 percent lower than his career average, and he is taking one fewer shot per game than he has historically - miring him in a nine-game goal-scoring drought on his way to a career low 30-goal season.

Semin, meanwhile, is seeing his superior talent get luckier than normal by shooting 19 percent (vs. 14 percent historically) in a career year when he could score 50 or more goals, proving it is better to be lucky and good.


Neil Greenberg also writes for Russian Machine Never Breaks. You can follow him on Twitter here.

By Neil Greenberg  |  01:29 PM ET, 12/04/2010

Categories:  Statistical analysis, Statistical analysis, Statistical analysis

 
Read what others are saying
     

    © 2011 The Washington Post Company