Is My Writing Improving? Observed Vs. Expected

A couple of months ago I got a little frustrated with my progress in my writing, having what appeared to me to be a lengthy lull in publications.  This got me thinking about probabilities and how many submissions I should have to make for each acceptance, etc., which is something I've written about at length before.

But one thing that occurred to me was that it would be nice to have some kind of metric on an ongoing basis so I could see how well I'm performing compared to the expected number of submissions required for each acceptance.  So I think this is pretty cool: as I've noted before, The Grinder posts acceptance rates, which can loosely be used to approximate a submission's probability of acceptance for a given market.  Summing up these probabilities over time gives a rough indicator of how many acceptances should be expected during that time, which can then be compared to the actual number of publications that occurred.

Here's my "over-under" for the past two years since I started submitting to science fiction markets:

Year	Quarter	Subs	Exp	Obs	Exp (Yr)	Obs (Yr)	Obs-Exp
2012	Q4	9	0.6359	0
2013	Q1	9	0.6796	0
2013	Q2	8	0.9949	1
2013	Q3	4	0.1591	0	2.4695	1	-1.4695
2013	Q4	1	0.025	0
2014	Q1	9	1.108	1
2014	Q2	3	0.3819	0
2014	Q3	10	0.9089	1	2.4238	2	-0.4238
Total			4.8933	3

The "Subs" column is the number of submissions I made to science fiction markets during each quarter.  The "Exp" column is the sum of the acceptance rates for all the markets I submitted to during each quarter.  The "Obs" column is the actual number of publications I had during each quarter.  So based on the submissions I've made so far, I should have had four stories accepted for publication.  I've had three, which puts me below the expected value for the average writer who logs submissions on The Grinder.  However, subsetting it by year, that one missing publication can (in my mind) be attributed to the learning curve that I went through that first year of submissions.  The "Obs-Exp" column captures the difference between the observed and expected values for each year, which I'm calling the over-under.  My over-under for the first year was -1.4695, which put me a whole publication behind where I should have been.  But in my second year, my over-under was only -0.4238, meaning I actually hit my target of two publications based on the markets I submitted to.  Which means I've improved as a writer (either in the writing itself or in selection of better fitting markets for the stories submitted).

I think this is a really good way to measure improvement in writing performance over time.

Now I just have to exceed the expectations long enough to make up for that missing publication....