Ditching the Average: How a Pure Count System Changed My Classroom
My Origin Story : Part II of III
Jul 18, 2026
As my Fall 2024 Biology data demonstrated, keeping the mathematical average while changing your scale is a profound structural trap. I thought that by calculating a GPA-style average of a student’s proficiency scores, I was being progressive.
I wasn’t.
I was just using a slightly different version of traditional grading math to force alternative learning into a traditional report card box.
When my state End of Course (EOC) data exposed that massive mismatch between my classroom averages and actual student proficiency on test day, I knew I had to drop the math entirely and look at pure performance metrics. As Sharona Krinsky and Robert Bosley frequently note on The Grading Podcast,
“Grading is the (mis)use of mathematics.”
Fixing the Rubric Rigor : Spring 2025
Before I could completely abandon the mathematical average, I had to fix how I was assessing my students. My weekly quizzes of 5 simple recall questions weren’t giving me enough data, and my GPA scale felt inflated.
So, in the Spring of 2025, I took 22 General Biology students and moved back to a 4-point scale. I also stopped avoiding rubrics. Instead of vague language, I built clear rubrics around single learning outcomes (like the Scientific Method) and evaluated them across specific sub-criteria — such as graphing, variables, and hypotheses.
This is an example of an assessment with 12 “proficiency points” across 3 learning outcomes based on quality rather than random points:
- Score 4: 10/12
- Score 3: 8/12
- Score 2: 6/12
- Score 1: < 5
This rubric overhaul aligned with a second, longer visit from Rick Wormeli — following his initial visit during my Physics experiment the previous fall — this time presenting to our entire school. The staff room was filled with heavy, structural debates. The loudest push back centered on a classic secondary dilemma: separating student content knowledge from student behavior. Without a Learning Management System (LMS) that cleanly isolates behaviors from academic performance, it is incredibly difficult to convince traditional high school teachers to get on board with grading reform. They worry we aren’t preparing kids for the strict deadlines of college.
But I was armed with actual classroom data. I showed Rick my Physics retake data, demonstrating the clear jumps in student performance once they were granted the time to actually understand the systems. Rick looked over my numbers and told me right then and there that I needed to publish my findings.
That conversation gave me the confidence to step out of the shadows and tell this story.
More importantly, I shifted the content away from strictly standard memorization and toward a mix with basic application. Instead of asking students to spit back a vocabulary definition, they had to actively apply the terms to a real scientific scenario.
The structural shift paid off instantly. That semester, my 22 students completed 107 reattempts. While 30 scores stayed the same and 19 went lower, 58 attempts showed clear improvement.
But the real victory came when the state data arrived: this time, my classroom data finally aligned with the state test. Only 3 out of my 22 students had scores that mismatched their gradebook proficiency — a massive leap forward from the 8-student mismatch the semester before.
Now, it was time to change the gradebook.
Setting the “Buckets” : Fall 2025
Two major shifts occurred in my classroom after a summer spent listening to The Grading Podcast and diving headfirst into grading reform. First, I completely removed averaging from my gradebook. Second, I replaced it with a pure Count System (often called the Bucket Method).
This system doesn’t care about a mathematical trajectory; it simply counts the total number of proficiency scores a student achieves over the course of an 18-week semester.
I decided to pilot the count system in my Fall 2025 General Biology course with 23 students. To organize my learning targets, I looked toward the Advanced Placement (AP) Biology curriculum. Because AP organizes content cleanly by unit and distinct learning outcomes, I selected 20 core topics that were appropriate for a general education framework.
Once I had my 20 topics set, I established a transparent, two-step requirement for the final A–F report card marks, completely bypassing fractions and decimal points:
The Biology Count Matrix
- To earn an A: 16/20 standards Proficient (3+), and 10/20 standards at a 4.
- To earn a B: 15/20 standards Proficient (3+), and 6/20 standards at a 4.
- To earn a C: 13/20 standards Proficient (3+), no 4s required.
I carried over the assessment design upgrades I started in the spring. For example, rather than a multiple-choice vocabulary quiz on tonicity, a quiz would feature a series of beaker diagrams. Students had to label the conditions (Hypotonic, Hypertonic, Isotonic), draw an arrow signifying the net flow of solutes, and write a short-answer defense explaining why they made that scientific choice.
By the end of the term, my 23 biology students completed 89 reassessments. 64 of those attempts resulted in an improved proficiency score, 19 remained exactly the same, and 6 showed a decline.
Get Addison Hilkert’s stories in your inbox
Join Medium for free to get updates from this writer.
Yet, a surprise setback emerged: 8 of my 23 students did not pass the EOC. It was a frustrating roadblock that forced me to start questioning the cognitive depth of my assessments. While my data focuses on pass rates in terms of numbers it is important to note that that many of the students who did not pass the standardized test did show enormous growth in science based on past data projections collected their 8th grade year and prior.
Moving to Physics: The Logistical Bottleneck
With the Biology count system running, I simultaneously decided it was time to implement a pure count framework within my 51 Dual-Enrollment Physics students. I wanted to be incredibly specific about the exact content and skills I expected my students to leave with. Using the AP Physics scaffolding, alongside the university’s required Thermodynamics unit, I mapped out a master list of 49 individual content topics.
Each topic was assessed independently using a highly structured 6-question blueprint:
- 2 questions targeting graphing skills (slope, area, and linearization).
- 2 questions targeting mathematical skills (proportional reasoning and algebraic expressions).
- 2 questions targeting conceptual argumentation.
Scoring was straightforward: 5/6 = 4, 4/6 = 3, 2/6 or 3/6 = 2, and 1/6 or 0/6 = 1.
I made it through exactly 28 topics before the sheer reality of managing 49 individual reassessment pipelines for 51 different human beings began to crush my system. Even though I had heard warnings on The Grading Podcast a hundred times before, I thought that my experience with complex Excel spreadsheets would allow me to avoid the logistical roadblock.
It was utterly impractical. I was drowning in paperwork.
I had to pivot mid-semester. To salvage the system without discarding the data, I combined the remaining isolated topics into cohesive topics, flattening the final gradebook blueprint into 40 streamlined physics topics.
Our final report card requirements for Physics mirrored the Biology bucket logic:
The Physics Count Matrix
- To earn an A: 32/40 at a 3+, with 20 at a 4.
- To earn a B: 28/40 at a 3+, with 12 at a 4.
- To earn a C: 24/40 at a 3+.
Managing the Anxiety of the “F”
When you completely remove points and averages from a high-stakes course, you trigger a massive wave of student anxiety.
The first nine weeks of Dual-Enrollment Physics are notoriously difficult. Under a traditional system, a student who fails the first two tests gets hit with a mathematical anchor that drags down their GPA for the rest of the year. Under my count system, because they hadn’t accumulated enough proficient “counts” yet in the first few weeks, their grade in the online portal looked like a D or an F.
I faced intense student panic. They were tied to a system where self-worth is calculated by a running total. I had to explicitly tell them: “I know where this road is leading you. Physics takes time to understand. You just have to trust me and the process.”
The shift happened as the reassessment window opened. A student who was sitting on a failing grade would walk in, demonstrate proficiency on four past topics they had finally mastered, and watch their report card mark instantly jump to a B or an A in a single afternoon. There was a distinct excitement in their voices as they finally mastered physics.
For the first time in their academic careers, they realized that an early struggle didn’t lock in their final trajectory.
Over the course of the semester, my physics students completed 251 total reassessments. 180 scores improved, 47 stayed the same, and 24 declined.
By the time Fall 2025 wrapped up, Rick Wormeli returned to our school for his third and final collaborative Q&A session. Having logged hundreds of hours listening to The Grading Podcast and deeply reading Rick’s foundational book, Fair Isn’t Always Equal, I found myself helping facilitate the discussion and answering questions for my colleagues.
I wasn’t just a teacher trying an experiment anymore. I was looking at a complete, verified data set that proved student growth was real.
Longitudinal Data Matrix Reference
The Structural Payoff of the Count
As I monitored the data, a clear trend emerged that validated the count system: the volume of student reassessments drastically declined during the second half of the semester.
It wasn’t because the students gave up or grew lazy. It was because they didn’t need to retake assessments anymore.
Because the count system rewards mastery without penalizing the early learning curve, students actually took the time to truly wrestle with and understand the foundational physics models in the first 5 units of the course. Units 6 through 9 were simply extensions of those exact same structural skills. Building models of motion and interpreting complex data graphs — tasks that used to take hours of remediation in past years — now took my students mere seconds to execute and explain.
They weren’t compliance-driven teenagers chasing decimal points anymore. They had organically evolved into physicists.
When my students looked at their final grades on their individual topic sheets at the end of the year, they knew the mark was an accurate reflection of their knowledge. There was never a fight over a grade. There was only a student looking at their dashboard and saying,
“Yeah. That makes perfect sense.”
Part III: Reassessment Isn’t a Loophole: How Retakes Built Actual Physicists.
