We pulled a random sample of one million logged workouts from the Boostcamp community to find out how people actually train, not how they say they do.
Every lifter thinks they know how people train. They've read the Reddit threads, watched the YouTube breakdowns, argued about volume and frequency in comment sections at 1 AM. But opinions aren't data.
This report is what actually happened. One million workouts, randomly sampled from the Boostcamp community. Every set, every skip, every PR. Some of what we found confirmed what the lifting internet has always believed. Some of it didn't. All of it is real.
1,000,000 workout records were randomly sampled from the Boostcamp platform. Each record includes exercises performed, sets logged, weights, reps, and whether each set was completed or skipped. All records are from real user sessions on the Boostcamp app.
Data analysis was performed with the assistance of AI, with human review and validation of all findings. Aggregate statistics were computed across the full sample. No individual user data was examined or presented at any stage.
All personally identifiable information was stripped from the dataset before analysis. User IDs were replaced with anonymous identifiers. No findings can be attributed to any individual. Boostcamp does not share or sell user data.
Boostcamp users represent a self-selected, program-following population and likely skew toward more structured lifters compared to casual gym-goers. Exercises with identical movements but different names may be counted separately. Findings describe patterns, not recommendations.
The 1M workout sample was drawn randomly and is representative of the broader Boostcamp user base. The full dataset from which the sample was drawn contains tens of millions of workout sessions logged since the app launched.
Combine dumbbell and cable variations and lateral raises appear in 238,000 workouts, making shoulder isolation the single most common movement pattern in the dataset. That's 38% more than bench press (172K) and more than squat and deadlift combined (249K). The dumbbell version leads at 153K appearances versus 85K for cable. Whatever the internet is doing to people's self-image, it's working. Everyone wants wider shoulders.
Bench press appears in 172,000 workout logs, 15% more than squat (150K) and 74% more than deadlift (99K). Lat pulldown (140K) nearly ties squat for second place, which says a lot about how the average gym-goer prioritizes their pulling. Leg extension rounds out the top three at 147K, confirming that machine accessories are a staple, not an afterthought.
Nearly half of all users who train legs never log a single calf exercise. Not once. Meanwhile, the overall skip rate for leg exercises is only 24%, compared to 18% for upper body. The meme is dead: people show up for squats. They just treat their calves like they don't exist. Seated calf raise has the worst skip rate of any exercise in the dataset at 38%.
Among users who logged both lifts, 1 in 6 has a higher bench max than squat max. The median bench-to-squat ratio is 0.80, meaning the typical lifter benches about 80% of what they squat. A bench-to-deadlift ratio of 0.67 suggests most people's deadlift is comfortably their strongest lift. If your bench is creeping up on your squat, you might want to examine your priorities.
The classic International Chest Day pattern holds up. Monday sees 171,000 workouts, 40% more than Sunday's 103,000. Volume drops steadily through the week, with a slight recovery on Saturday before cratering on Sunday. Saturday and Sunday combined still don't match Monday and Tuesday.
Of the 16.7 million sets logged across the platform, 3.5 million were skipped. That's more than one in five. The exercises people bail on most aren't the heavy compounds, they're the accessories they know they should be doing. Calves, abs, and hamstring curls lead the skip charts. Bench press, unsurprisingly, has the lowest skip rate at just 10%. Nobody skips bench.
Most skipped: Seated Calf Raise (38%), Cable Crunch (36%), Hanging Leg Raise (34%).
Least skipped: Bench Press (10%).
Upper/Lower is the dominant training split with 139,000 tagged workouts, nearly double PPL's 74,000. The classic bro split (individual muscle group days) accounts for just 26,000, making it less than a fifth as popular as Upper/Lower. Full body comes in last at 15,000. The Reddit consensus that PPL is king doesn't match what people are actually running.
The most common exercise pairing in the entire dataset is leg curl and leg extension, appearing together in 59,000 workouts. Bench press + incline dumbbell bench comes in at 35,000, while lat pulldown + seated cable row pairs up in 30,000. The data reveals a clear pattern: people superset antagonist muscle groups on leg day, and pair a heavy compound with an accessory variation on push and pull days.
The median consecutive-week streak is exactly 4 weeks. Nearly 60% of users can sustain at least a month of consistent training, but the 8-week mark is where most consistency collapses. Only 17% maintain an 8-week streak. If you've been showing up every week for two months straight, you're in the top fifth of all tracked lifters.
The median session includes 6 unique exercises. Nearly 44% of workouts land in the 5–6 exercise range, and the 75th percentile is 7. Only 7% of lifters regularly program 9 or more exercises per session. The 2-exercise minimalist workout accounts for fewer than 9% of sessions. Most people are running focused, moderate-volume workouts, not marathon junk-volume sessions or ultra-minimalist routines.
The median session clocks in at 58 minutes. The 75th percentile is 79 minutes, and the 90th percentile hits nearly 109 minutes, close to two hours. The mean is skewed to 87 minutes by a long tail of marathon sessions. Most lifters are in and out in about an hour, which is exactly what most evidence-based programs are designed for.
Despite the internet insisting on 5- or 6-day splits, the median Boostcamp user works out fewer than 3 times per week. Only 16% average 4+ sessions, and just 4% hit 5+. The 75th percentile is 3.6 days, meaning three-quarters of lifters are training less than 4 days a week. Your program might be written for 6 days. Your life disagrees.
The data is clear: the lifters who track everything are the ones who progress. Whether you train 3 days or 6, the best program is the one you actually follow.
Start Training with Boostcamp