Is Alex Semin A Streaky Shooter?

Alex Semin

Photo credit: Greg Fiume

I showed the other day that all hockey players, particularly goal scorers, hit slumps, and it’s possible that the skater is not actually streaky by nature, but due to some bad bounces he appears streaky. One of the criticisms of Washington’s other Alex, Alex Semin, is his shooting game lacks consistency– but is that really true?

I admit, if I had to put odds on “Is Semin a streaky shooter?”, I would say it was 70-30 he was and I think most Caps’ fans would agree. To find out I am going to set up two different simulations (each run 10,000 times and normalized to Semin’s 175 games) and compare those results to Semin’s actual performance from the 2008-9 season to present day. The first simulation is going to be for Mr. Consistency. He always shoots 14.4% every night and has the same shot distribution as Semin did for the last 2 1/2 years. The second, Mr. Streaky,  is a bit more complicated.

For Mr. Streaky I use two different shooting percentages. One for when he is hot (21.6%) and one for when he is cold (9.6%). To determine if he was initially hot or cold at the beginning of a “season”, I flipped a coin and then assigned a 90% probability that he would stay in the state he was in game after game. In other words, he has a 90% chance of staying hot if he’s hot, or cold if he’s cold and only a 10% chance to change states. He too had the same shot distribution as Semin did for the last 2 1/2 years.

Then I compared the numbers of three categories to see if Semin’s actual season stats were closer to one or the other:

  1. Number of streaks where he went 3 or more games scoring at least one goal.
  2. Number of 0 goal games.
  3. Number of multi-goal games.
Streaks Alex Semin Mr. Consistency Mr. Streaky
Games Played 175 175 175
Number of 3+ game streaks 7 5 4
Number of 0 goal games 106 115 120
Number of multi-goal games 20 12 9

First we have the number of 3+ game goal scoring streaks. Mr. Consistency saw 5-goal scoring streaks out of each 175 game block while Mr. Streaky (surprisingly) saw only 4. Semin’s actual number was 7, giving him twice as much of a chance at those numbers using the consistency model than with the streaky one.

Games with a goose egg were close for each, but Semin is four times more likely to have only 106 games without lighting the lamp using the consistency model than the one that saw hot and cold streaks.

Finally we have the multi-goal games. Semin had 20 of them, almost as much as both simulations combined, but he is more than ten times likely to have his numbers using the consistency model’s footprint than Mr. Streaky’s.

What does it all mean? For one, Semin doesn’t appear to be as inconsistent as many fans believe– at least not when it comes to scoring goals in the NHL. That’s not to say he doesn’t go through scoring droughts (or make bonehead plays in the offensive zone), but when looked at in terms of what we can expect from elite scorers just by random chance alone there seems to be some real consistency there: Semin had more goal-scoring streaks and multi-goal games than the consistency model– with fewer zero goal games to boot!

Maybe this makes his $6.7 million one-year extension a bit more palatable, maybe not, but one thing is for sure: Alex Semin is more consistent than we give him credit for.

Tagged with:
  • I’d rather have Semin on my team than Kovalchoke. ‘Nuff said right there!

  • DarkStranger44

    Good post!

  • bb

    I agree with Sakurafire up there.

  • Ian

    Neil, just FYI. When I first read this post, it blew my mind. Great work.

  • KHtaD

    So this is essentially a 2-state HMM without the learning algorithm? I’ve been looking at doing a full-blown HMM for Semin’s shooting with a learning algorithm to discern what his true states or what the data look like if he is, in fact, modelable with a hidden Markov process.

  • @KHtaD Yea. (HMM = hidden Markov model). #geeksrule


    Excellent post!

  • colcurry

    I think you need to run your simulations with other players to provide some context – do players always outperform their consistency/streaky models? Also, you might want to play around with the parameters a bit. A shooting percentage of 9.6 isn’t particularly bad, and the probability of maintaining a given state (90%) seems rather high.

    Why not look at, say, 5 game moving averages and calculate the standard error? I guess it all depends on how you define consistency.

  • Will

    This doesn’t make sense. Ignoring Mr Streaky, Semin is streakier than Mr Consistency. Mr Consistency is streakier than Mr Streaky in 2 of the 3 columns. I only have a basic understanding of Markov chains but either this data is a fluke or it is a bad model.

  • Neil – RMNB

    @Will I never said he was exactly like one or the other, just perhaps more consistent than we give him credit for.

    @Colcurry I did the same for Phil Kessel as a FanShot over at Pension plan Puppets. Turns out he is WAY more consistent that people give him credit for. Also, will have Ovechkin post later today.

    I agree, a 9.6% shooting percentage isn’t bad, but it is when comparing to an average of 14%. Plus, keep in mind we are looking for a cold shooter – not a terrible one.

    Feel free to run one yourself using a 5 game moving average, calculate the standard error and lmk what you find.

  • Pingback: What We Learned: Only one solution to this Phil Kessel trainwreck « harveyrobb()