Unmathing the Gradebook: Why I Left Percentages Behind
My Origin Story : Part I of III
Jul 10, 2026
For my first six years in my current district, I played the traditional grading game. I didn’t assign or grade homework, but I did meticulously grade and apply points to every single lab and classwork worksheet. My logic was simple: since I was right there in the room watching them do the work, it felt fair.
But if I’m being honest, those scores were more about compliance than actual learning.
Science is an inherently collaborative subject. Because of that, most worksheets and lab documents were completely shared by the end of the period. All the students had to do was copy down the answers right in front of them. I used to comfort myself with a classic teacher justification: At least they are getting exposure to the vocabulary and the concepts. Going through the motions is worth something, right?
I was also constantly wrestling with the math of the traditional gradebook. When setting up my syllabus, I realized that because of how the calendar fell, the first quarter of the semester had almost double the total raw points available compared to the second quarter. Yet, at the end of the term, the software automatically weighted both quarters as exactly 40% of their final grade, with the final exam making up the remaining 20%. I was forcing messy human learning into rigid, arbitrary formulas.
This wasn’t my first time dealing with the illusion of gradebook math, either. Before this, I worked in a past district that required all science classes to designate the exact same weighted categorical system for consistency. We judged student performance by homework, quizzes and labs, and assessments. But here’s the kicker: it was never discussed a given percentage was used for each category. It was just assumed that as long as we were all consistent, that was what mattered.
If anyone has played around with a digital gradebook in this way-or made a mistake when entering weights-you quickly realize how easily a student’s grade can swing higher or lower with a slight change in the settings. It sparked a question that would take me years to fully understand: Does a student actually understand more or less content just because I decide to weigh homework as 20% or 10%? Was curving a really hard test because it was AP, or giving bonus points because scores were so low to match the legacy framework of the 0–100 grade scale, really telling the true story of my students?
The Catalyst: Fall 2023
The turning point came in the Fall of 2023. Our school did a building-wide book study on A Repair Kit for Grading: 15 Fixes for Broken Grades by Dr. Ken B. O’Connor, with roughly 20 staff members.
The text challenged my entire framework. What exactly is the purpose of homework grades when students are just beginning to learn the content, yet we expect perfection 24 hours later? Students end up copying each other or cheating in other ways just to get the work -but the content isn’t actually .
A Repair Kit for Grading provided the hard evidence for why the traditional grading system we were all indoctrinated to use-not taught, by the way-is completely broken. It offered practical fixes. So, the very next semester, I decided to make a definitive change. If I wanted to push how far an alternative grading approach could go, I was going to remove as much of the traditional system from my grading as possible.
The First Experiment : Spring 2024
For my first attempt, I chose my General/Inclusion Biology course rather than my AP Physics 1 course. My high school runs general science on a double-block semester timeframe, while AP courses run as a full-year double-block. This setup was a massive benefit; it allowed me to implement, reflect on, and tweak two full semesters of a course in a single calendar year.
My first structural move was dismantling the traditional, overwhelming unit test. Instead, I started using smaller summative assessments distributed throughout the semester, striking whenever natural breaking points in the content presented themselves. I wanted a quicker diagnostic response loop as a teacher to see if the essential knowledge was actually being retained.
But designing how to grade these mini-assessments was a massive work in progress. While I knew from my readings that Standards-Based Grading (SBG) is supposed to be built around clear rubrics of achievement, I was having a really hard time digesting the language of traditional matrices. Phrases like “general understanding with little to no errors,” “partial understanding,” or “completed the standard independently” felt far too subjective. Unable to find clarity in vague rubrics, I retreated to what felt safe: a mathematical workaround to measure proficiency.
To keep the math clean-and to have a fallback plan if this entire experiment crashed-each weekly assessment consisted of exactly 5 questions targeting essential knowledge . If a student answered 4 out of 5 correctly, they earned a 4; 3 out of 5 was a 3, and so on. A score of 1 was given just for attempting the quiz, completely eliminating the devastating weight of a traditional zero.
Still mathy, I will admit. After a career spent teaching physics, my brain was deeply hardwired to rely entirely on math and numbers.
Alongside the new scale, I implemented a non-punitive retake policy to replace past performances with a second version of each assessment. I was morally 100% on board with retakes; it just made human sense that students should be able to prove mastery on their own timeline. But logistically, the time commitment required to write version after version of assessments stressed me out. That was when I started listening to with Sharona Krinsky and Robert Bosley. They helped me realize a massive administrative truth: because I was no longer spending hours grading homework and shared labs, I finally had the bandwidth to spend my time on what I truly valued-creating rich, diagnostic assessments. That is the hidden superpower of alternative grading. You stop stressing about whether to deduct two points or three points for a minor mistake on a worksheet. It all becomes pure feedback. If I made an error grading the “other stuff” in class, it didn’t matter, because that stuff wasn’t tied to a permanent grade anymore. So at the end of the quarter and semester exams, I implemented a strategy I learned on The Grading Podcast where students got to “choose their own adventure” by selecting the specific content strands they wished to reattempt and improve.
Get Addison Hilkert’s stories in your inbox
Join Medium for free to get updates from this writer.
Listening to the podcast also opened my eyes to the fact that there are multiple ways to grade without points. It was frustrating at first because there is no cookie-cutter model out there. As teachers, we are always searching for that perfect plug-and-play template to ensure consistency in measuring learning. But it doesn’t exist. I had to realize that the only valid way to navigate this journey was in person, through active trial and error, not hypothetical templates.
I was also transparent with my students from day one. I told them what I was trying to do, that this was meant to empower them, and promised them that my trial-and-error wouldn’t harm their final report cards. We openly discussed the system. For the first time in my career, it felt like we were in this together, rather than an adversarial ‘me vs. them’ dynamic.
To calculate final report card grades, I took the mathematical average of these 1–4 scores and mapped them to “GPA-style” breaking points, which I chose, to convert them back into traditional letter grades. While I wasn’t internally proud of relying on an average, it still felt like a major upgrade. It completely removed the structural trap of a single low test score that so often drags down students in a traditional system and I was averaging their top scores and not all of their attempts.
Looking back at that very first tracking spreadsheet two years later, my documentation on those early reassessments was admittedly messy. My LMS did not support my method of grading and all of my data is recorded on spreadsheets where I simply typed the new scores directly over the old ones without tracking whether the mark was an original or a reattempt, meaning I can’t pull exact statistical data on our growth. What I can tell you is that 12 out of those 18 students went on to pass the state End of Course (EOC) exam. More importantly, as the semester wrapped up, interest caught on; additional teachers in my building began reading A Repair Kit for Grading , and I was able to compile and present my initial findings on how a simple retake loop was already reshaping student learning.
The Physics Paradox : Fall 2024
Going into the Fall of 2024, my grading architecture fractured into two completely different experiments based on the calendar design of my preps. Because I had far more experience managing alternative grading structures in Biology, and because it operated on a fast-turnover single semester block, I felt comfortable escalating its scale. I shifted my 21 Biology students from a 4-point scale to a 5-point scale. My initial motivation for this was simple: both AP exams and our state EOC tests scale from 1 to 5. I also had to translate grades at the end of the semester to a traditional 5 letter system. I was curious if there be a correlation between my report card grades and the achievement levels on high-stakes state tests. It also allowed me a larger range of flexibility with averaging final scores and distributing the GPA scale into wider grade bands.
Simultaneously, I was teaching Dual-Enrollment Physics. Because Physics was a high-stakes, full-year course populated by college-bound students, I did not feel structurally secure enough to upend its entire grade grading system yet. In that room, I held onto the traditional 0–100 scale and standard curves to handle university complexity, but implemented a singular, strict goal: full reassessment with score replacement.
This move was driven by a fundamental flaw in advanced courses like AP: the reliance on grade curving. Because a 70% raw score on a College Board exam often equates to the highest score of a 5, teachers frequently design complex formulas to adjust traditional 90–80–70 grading scales to match AP expectations.
I previously utilized two methods to address this gap. First, I allowed students to earn half-credit back on incorrect questions by submitting written corrections explaining their errors. While this successfully raised a 50% raw score to a passing 75% and incentivized reflection, by the end of my first year teaching AP Physics, it was clear that the final grades reflected point-accumulation rather than actual content mastery.
To resolve this, I transitioned to a standard square root curve-calculating the square root of the raw percentage and multiplying by 10-while keeping test corrections mandatory but worth zero points. This mathematically aligned raw scores with AP bands, but it revealed a new systemic flaw: the formula fails at the lowest end of the spectrum, such as a raw score of zero. These mathematical limitations ultimately demonstrated that modifying a traditional curve was insufficient; the situation required a complete overhaul of the grading framework itself. This brings us back to full reassessments without penalty.
That fall we had the privilege of Rick Wormeli visiting our building for his first of three distinct visits. I was selected as one of two department representatives to meet with him in an intimate small-group setting. Our conversation heavily revolved around the mathematical injustice of the traditional zero, the 50% floor policies that districts are scrambling to implement, and how to start introducing these ideas into a classroom. While Rick provided a wealth of philosophical validation, there still wasn’t a direct ‘here you go’ structural model for a secondary science class. I had to build the road while walking it. But our discussion validated my mission: if I wanted my students to move toward actual content learning and away from grade-fixation, I had to completely change how my gradebook handled human struggle.
In total, by the end of that Fall Biology term, my 21 students partook in 36 reattempts. 29 of those assessments resulted in clear improvements, and 7 remained at the same score. In a single semester, I had successfully created a group of students from point-grubbing to truly trying to understand biology at a deeper level. This was true for both my General students and my Intervention students. They developed a passion for asking questions and being curious rather than stressing over exams and compliance.
Reassessments had the greatest impact on my intervention students, who quickly realized they still had a genuine opportunity to pass regardless of their past performance. In a traditional grading system, a common and heartbreaking scenario plays out: students check out after the very first unit test because they have already fallen into a mathematical hole so deep they can’t climb out of it. Offering reassessments completely changes that dynamic, preserving their motivation and keeping them engaged in the learning process.
The Confrontation with the State Data
But my alternative journey hit a major roadblock when the state data returned. My final Biology grades that semester still relied on the mathematical average of all assessments, even though I was replacing past performance with recent achievement and mapping it to a 5.0 GPA scale. When the state End of Course (EOC) data came back, it did not yield the results I was expecting based on what I thought my students understood.
If we consider a traditional “C” to be the proficient passing benchmark (equivalent to a 700 score on the Ohio EOC), 8 out of my 21 students did not meet that threshold-even though my gradebook provided clear evidence of at least “C” level learning in my classroom. Only one student in the entire cohort walked away with a lower class grade than their actual EOC performance tier.
A flood of questions hit me. Was creating my own GPA scale and “mathing” a grade finally catching up with me? Should my grades over an 18-week semester perfectly mirror a standardized assessment taken on a single day? Were my weekly quizzes matching the rigorous expectations of the state? How exactly is a standardized test measuring proficiency versus how I am measuring it? Was I inflating grades?
The discrepancy brought me back to a distinct moment earlier in my teaching journey. Six years prior, when I was still using traditional grading, I had a student who did absolutely no homework, lost their lab reports, and performed average on tests. They managed to scratch out a “D” by the end of the course-mostly due to weighted categories, arbitrary points, and averages. Yet in class, they were completely engaged, a core part of the scientific conversation, and asked incredibly deep questions. When their state EOC results came back, they scored an Advanced 5!
Under that traditional system, what was I supposed to do? Go back and retroactively change their report card to an “A” just because the State of Ohio recognized them as Advanced?
Looking at the data, the truth became inescapable. My system was still relying too heavily on mathematical formulas to force alternative grading into a traditional report card box. I needed a way to drop the mathematical average completely and look at pure performance metrics.
Originally published at https://addisonhilkert.substack.com.
