The Year’s Best Science Fiction & Fantasy 2016 (Edited By Rich Horton) was the ideal selection to quantitatively examine factors which contributed to my having a high opinion of a science fiction or fantasy short story and the range and frequency of these factors considered appropriate for the short fiction market. I read all the stories in this collection and for each of its 30 stories I recorded my overall rating, length, point of view, uniqueness of voice, first impression, and engagement. With these ratings in hands, I determined the correlation of each factor with my overall rating of the story in this collection. I determined that the strongest predictor of a high rating was my engagement with a story.
I define engagement as the number of pages left in the story after the page on which I felt engaged divided by the total number of pages in the story. This definition of engagement allowed me to track the number of pages I feel invested in the story without penalizing shorter stories. Engagement had the highest Pearson correlation coefficient of any metric which I tracked (0.67: a value which indicates a strong relationship between a high engagement and my rating of the story).
My positive rating of the opener of a story was weakly correlated with my overall positive rating of the story. Length of the story had no significant correlation on my rating. 37% of stories were told in the first point of view, 63% in the third point of view, and point of view had no impact on my average rating. 87% of stories were told in the present tense, 13% in the past tense, and tense of the story also hand no impact on my overall rating. Approximately 60% of stories were by male authors, and received an average rating of 2.9/5 stars and approximately 40% of stories were by female or non-binary authors and received and average rating of 3.3/5 stars.
The majority of the stories were less than 15 pages. The majority of my ratings for these stories were 3/5 stars with only 4/30 stories receiving a 5 star rating. Stories told in what I considered a unique voice (3/30) had a slightly lower average rating than stories in a straightforward voice (27/30) (3 vs 3.2 average rating) although the stories written in a unique voice had a much greater standard deviation, meaning that they encompassed a greater variety of ratings. With only three stories in the unique voice sample, it is difficult to tell whether this is a reliable factor to consider.
After discovering that my engagement with a story correlated with my overall opinion, I went back to examine whether any story lost my engagement after obtaining it. None did. My take away from this analysis is to consider optimizing for engagement as I write my own science fiction stories. Does your engagement with a story correlate with your overall opinion? Weigh in in the comments.