What 1,811 benchmark workouts say about scoring

PRzilla's catalog holds 1,811 non-custom workouts. Most carry a benchmark: a score type, a direction, and nine thresholds that divide performances into ten levels — 1,790 have the full ladder, and eight have no score type recorded at all. Building that out meant deciding, workout by workout, what "better" means. The answers are more varied than I expected.

Time is the default, and it is not close

Across all 1,811, with eight carrying no score type at all:

| Scored by | Count | Share | | --------- | ----- | ----- | | Time | 1,338 | 73.9% | | Rounds | 206 | 11.4% | | Reps | 205 | 11.3% | | Load | 34 | 1.9% | | Distance | 7 | 0.4% | | Unscored | 13 | 0.7% |

Three quarters of the canon is "how fast". Load — the thing most people outside CrossFit assume strength training measures — accounts for 34 workouts, under two percent.

That imbalance shapes what progress feels like. If most of your benchmark history is timed, most of your visible improvement will be measured in seconds saved, and the strength underneath it stays comparatively invisible. It is a large part of why we track the work inside a workout separately from the score on it.

Only 644 of the 1,811 carry a real time cap. The other 1,167 have none, which means "did not finish" is not a state most of the canon has an opinion about.

The five workouts where slower is better

Every timed workout scores lower as better. Five do not.

Each ladder below runs from its level-2 cutoff to its level-10 cutoff. Level 1 is anything under the first number.

• Games: Handstand Hold — 5 seconds up to 240 seconds
• L-Sit Hold: Max Time — 8 seconds to 120 seconds
• Macho Man — 9 minutes to 40 minutes
• Bergeron Beep Test — 6 minutes to 22 minutes
• Stimulus Travel WOD 18 — 3 minutes to 8 minutes

These are the workouts where the clock measures endurance rather than speed: hold a position, or keep repeating a round, until you cannot. In our schema the field that marks them is called higherScoreIsBetter, and it is an exception flag rather than a general setting — direction follows from score type, and these five override it.

Five out of the 1,338 timed workouts. That ratio is the interesting part. A scoring system that assumed "timed means faster is better" would be right 99.6% of the time and would silently invert the grade on a handstand hold, turning a strong result into a terrible one. Rare cases are where scoring quietly breaks, and the only defence is encoding them rather than inferring them.

The spread on those five is also worth noticing. Handstand Hold's graded range runs 5 seconds to 240 seconds — a 48-fold span from the bottom of the ladder to the top. Almost nothing else in the catalog separates athletes that hard.

Men's and women's ladders are not one ladder times a constant

Of the 1,790 workouts carrying a full ladder, 1,690 — 94% — have a separate women's ladder rather than reusing the men's.

The obvious way to build those would be a fixed handicap: take the men's times, add fifteen percent, done. That is not what the catalog contains. Across the 1,271 timed workouts that have both ladders, the ratio of the women's level-10 threshold to the men's has a median of 1.13 and a range from 0.79 to 1.50.

So the typical women's elite standard is about 13% slower — but the actual figure runs from 21% faster to 50% slower depending on the workout. Eighteen workouts have identical ladders for both.

Three familiar workouts show the shape of it, comparing level-10 thresholds:

| Workout | Men | Women | Gap | | ------------------- | ---- | ----- | ---- | | Grace | 3:00 | 4:00 | +33% | | Fran | 3:20 | 4:00 | +20% | | Helen | 8:00 | 8:30 | +6% |

Grace is thirty clean and jerks and nothing else. Helen is mostly running, with kettlebell swings and pull-ups. Fran sits between them. The gap tracks how much of the workout is barbell at a prescribed absolute load — which makes sense, because an absolute load is a different fraction of bodyweight for different athletes, while a 400 metre run is a 400 metre run.

Run the numbers on that fifteen percent and you can see what it would cost. Helen's men's level 10 is 8:00, so a flat handicap gives women 9:12 — a full 42 seconds softer than the 8:30 the catalog actually sets. Grace's 3:00 becomes 3:27, against a real standard of 4:00, making it 33 seconds harsher. One formula, wrong in both directions at once. Getting this right required looking at each workout, which is why there are 1,690 of these rather than one rule.

Why this matters if you are just logging scores

You do not need to know any of this to use a tracker. But it explains why two workouts that both say "for time" can mean different things, why your level can move differently on Grace than on Helen, and why a handstand hold needs its own rule.

Scoring looks like the simple part of tracking a workout. It is the part with 1,811 individual decisions behind it.