2010/11 Draft Review Part 1: Absolute Inter-round variance

In this series, we'll look at how valuable draft picks are by round and to each manager. Who does the best job of drafting, and how likely is any given manager to find a starter in each round?

PART ONE

No draft pick comes with the guarantee of success. Injury and shifting roles can derail the fantasy production of even the best in the game. For us, that manifests in hits and misses, successes and failures in our drafts - a binary designation we sometimes have to wait a full 25 weeks to assign.

Well, it's been almost 48 weeks since we drafted our squads for the 2010/11 Definitely Offside season. It's about time we start finding the winners and losers.

Our draft comes with the wrinkle of having 96 players kept from season to season, effectively condensing the variance from the top of the draft board to the bottom. That said, the first overall pick was Taylor Hall, and the last pick was Igor Makarov. You know, the 24 year old with a point in the AHL for every year of his life.

We begin our analysis with that question of condensed variance. What is the difference in value in draft picks of different rounds? Is there an appreciable advantage to having a 1st round pick over a 2nd, or a 12th rounder over a 13th?

Now to tickle my favorite g-spot - GRAPHS!

Above: The average fantasy points scored by the skaters selected in each round

In general we see a downward trend in average points per round, suggesting we selected the best players early and the worse players later.  Duh, right?  Well, it means we did a solid job of predicting fantasy point production as a group.  Imagine comparing it to a draft in which players are randomly assigned to rounds.  If that meant anything to you, you know we did a pretty good job.

The consequence of our collective success is that each successive round appears to hold a lower likelihood of selecting a high impact player than the round previous.  In the graph above, the relation is not linear (Round 5 holds a disproportionate number of injuries, for example), reflecting the reality that the actual difference between the last pick of one round and the first pick of the next is just one player.  Segregation by round is thus artificial.  Moreover, round by round averages hide the success of each manager within the round.  These are limitations I will return to in future posts on our draft.  For now, lets stay at a macro-level and maintain our imposed round-by-round analysis for the sake of clarity.

Goalies score more points than player, so including goalies in any average would considerably skew the graph.  I've separated them out below.

Above: The average fantasy points scored by skaters in each round again (blue) next to the
average fantasy points scored by goalies selected in each round (red)

Only 15 total goalies were selected, and as you can see, no goalies were selected in rounds 1, 4, 5, and 13.  Thus, their averages aren't that meaningful.  Still, we see that the round you select a goalie is not a clear indicator of the quality of goalie you get. In fact, the best goalie drafted (Carey Price) was selected in round 11, along with Dan Ellis and Antti Niemi. The best four goalies selected were taken in rounds 3, 6, 9 and 11.

In other words, in a very small sample size (15), we did a poor job of predicting the fantasy outcomes of goalies.  For Shame!

Upcoming in this series:
- Part 2: Proportional Inter-Round Variance
- Part 3: Inter-Round Variance Using Thirdiles and Interquartiles
- Part 4: Drafted versus Undrafted Skaters and Goalies
- Part 5: Inner-Round Variance by Round
- Part 6: Inner-Round Variance by Manager