clock menu more-arrow no yes mobile

Filed under:

Realistic expectations and the truth about randomness

How realistic is a consistent forty goal scorer?

Marianne Helm

There are some criticisms to Winnipeg Jets winger Evander Kane and his game that are legitimately founded. He is not a perfect player by any means and has his faults; however, there are many criticisms that I believe extend from unrealistic expectations or perceptions of what is the norm, what is exceptional, and how difficult some things are to obtain.

The lofty 40 goal plateau

Forty goal-scorer is a term thrown out commonly by both supporters and critics of Evander Kane. Supporters suppose "if only Kane had better linemates, he'd be a 40 goal scorer" and critics will detract from Kane's accomplishments for not reaching that plateau.

How reasonable is it to expect a 40 goalscorer from a player?

Well over the last three seasons, a consistent 40 goal scorer would need to put up at least 103 goals total due to the shortened 2012-13 season. Only Steven Stamkos (114) and Alex Ovechkin (121) have reached that level, with neither player posting above the 40 goal (or 23 for pro-rated short season) in all three times. The next closest in goal totals is Corey Perry, whose 95 goals averages out to just slightly better than a 36 goal scorer per a season.

If looking at per game pace instead, a 40 goal scorer would need to pace at 0.49 goals per game or higher. We find again that over the last three seasons only Stamkos (0.68) and Ovechkin (0.59) have paced above that. There are two players who have matched that pace in James Neal and Evgeni Malkin, although neither one of them have matched it in all 3 seasons.

A thirty goal scorer is essentially what the old 40 was in the previous NHL era. A 30 goal scorer pro-rated with the short season would be 78 or more goals over the last 3 seasons. Yet, only 15 players have totalled 78 or more over the last three seasons combined.

Looking at pace, the per production of a 30 goal scorer with full seasons played would be 0.37 goals or more per game. Only 32 players have averaged that or better. Evander Kane sits at 34th in the NHL for per game pace with 0.36 goals.

The fact that Kane is just 0.01 goals per game short from a 30 goal scorer pace is extraordinarily impressive at his young age, especially given Kane's cast of linemates while predominately playing on the second power play unit. Now, Kane is not on the primary power play unit for a reason. Andrew Ladd is a superior power play option currently and is a better choice for the team; however, there is no denial that any players production will improve with more ice time on the power play.

The ugly myth of the consistent player

The truth is that consistency -or lack thereof- in goalscoring tells you more about a players goal totals and the natural process of randomness and variance than it does about a player, their desire, or their character.

If a 0.5 goal per game scorer (which noted previously, there are only two of) won't normally score five goals in a ten game set. Some sets will have more, while others will have less. In addition, any set with five goals will not likely have five games with goals and five without. The more normal situation is those five goals being likely scored between two to four games, which means there are most likely six to eight games without goals in a set of ten for a 0.5 goal a game scorer. The cherry on top to all this is that those games without goals are not likely to be spread out evenly but will come in bunches, as will the games with goals.

And voilà, randomness creates a streaky goalscorer even out of the best of players. The effect of randomness on creating streaks becomes even more intensified as a player averages less goals per game.

It's merely a psychological fact that our brain inherently looks for patterns and has a poor perception of what random distribution truly looks or feels like.

A famous example of this effect was the shuffle function on Apple's iPod line. Originally the iPod's shuffle function was truly random. Can you guess what happened? They had to change the programming due to public pressure from massive amounts of complaints of the shuffle function seeming to be non-random. Too often songs were being played from the same artists or even the same album in succession of two or three in a row, sometimes even in the same order as on the album. When Apple made the change,Steve Jobs was quoted: We're making it less random to make it feel more random.

Another famous example is a little party trick for number nerds. Give someone two pieces of identical paper, a pen and a coin. Tell the person to write an order of 50 pretend coin-tosses in a row, without you seeing the paper. Then have the person actually do the experiment, again without your knowledge and writing the results down. Far more often than not (unless your friend has heard this trick or knows numbers), picking out the real and fake coin-toss is easy. The real coin-toss will likely have the largest streak. The fake coin-toss will have very few large streaks and be almost patterned like with a very even distribution.


In a lot of ways, the change from past to modern era NHL has been a bit tougher for Jets fans to grasp, as some have been at least partially removed since 1996. Between 1990 and 1996, there were 103 forty or more goal player seasons. That's about 17 a season. Over the last three years there has only been 52 thirty or more goal player seasons, plus 16 players who scored 30 or more pro-rated goals in the 2012-13 lockout shortened season (which is actually easier to do in a shortened season). That's about 22 a season.

In addition, randomness is not intuitive to humans. It's not natural for us to view randomness as not being evenly distributed, just like it is natural for us to impose a pattern when trying to create randomness. Streaks will happen. For the most part, the consistent player is less of a skill and more of a chance; a chance that becomes more likely as you score more goals, but goals are becoming rarer and rarer in the new NHL.