Summary Every core FitCraft feature maps to a specific mechanism tested in randomized controlled trials published in JAMA, JAHA, and other top medical journals. Streak systems are grounded in loss aversion research (+759 to +981 steps/day across 3 RCTs). Competition mechanics come from the STEP UP trial (+920 steps/day, n=602). Self-chosen goals are based on the ENGAGE trial (+1,384 steps/day, n=500). Progressive difficulty draws from the GAMEPAD trial, where effects grew post-intervention. This page shows the mapping between each feature and the trial that validates it.

"Backed by science" is the most abused phrase in fitness marketing. Every app claims it. Almost none cite a specific study.

FitCraft is different. Every game mechanic in the app corresponds to a specific finding from a peer-reviewed clinical trial. Not a blog post. Not a pop-science book. A randomized controlled trial published in a medical journal with a PubMed Central ID you can look up yourself.

Below is the complete mapping — feature by feature, trial by trial, with exact effect sizes.


The Complete Feature-to-Evidence Map

FitCraft Feature Psychological Mechanism Supporting Trial(s) Effect Size Publication
Streak System Loss aversion + commitment consistency ALLSTAR (n=150), Stroke RCT (n=34), Veterans (n=180) +759 to +981 steps/day JACC CardioOncology 2025, JAMA Neurology 2022, JAMA Netw Open 2021
Competitive Challenges Social comparison + competition STEP UP (n=602), iDiabetes (n=361) +920 steps/day (sustained) JAMA Internal Medicine 2019, JAMA Netw Open 2021
Self-Chosen Goals Goal autonomy + immediacy ENGAGE (n=500) +1,384 steps/day (sustained) JAMA Cardiology 2021
AI-Adaptive Difficulty Flow state + automated coaching GAMEPAD (n=103) +1,074 steps (grew post-intervention) JAHA 2025
Collectible Cards Variable ratio reinforcement Skinner reinforcement schedule research + gamification meta-analysis (16 RCTs, n=2,407) g = 0.42 (pooled effect) J Med Internet Res 2022
Quest Progression Competence + autonomy (SDT) BE FIT (n=200) — points + levels + progression 53% vs 32% goal achievement JAMA Internal Medicine 2017
Social Accountability Relatedness (SDT) + partner accountability BE FIT families (n=200), Postpartum HDP (n=127) +647 to +953 steps/day JAMA Intern Med 2017, JAMA Cardiol 2022
Tiered Leagues Personalized competition (avoid leaderboard backfire) Fitbit leaderboard econometrics (large observational) +1,300 steps sedentary / -630 active PMC10403254, 2022

Feature Deep Dives

Streak System

FitCraft tracks consecutive workout completions. Breaking a streak feels like losing something you already earned — that's loss aversion, one of the most robust findings in behavioral economics. Daniel Kahneman and Amos Tversky's prospect theory (1979) showed that losses are psychologically about twice as powerful as equivalent gains.

Three clinical trials used loss-framed point systems — where participants were endowed with points at the start of each week and lost them for missed goals — and all three found significant step increases.

Supporting evidence
ALLSTAR Trial (2025) — 150 cancer survivors, 64% Black, 35% Hispanic. Weekly point endowment with daily loss for missed goals. Result: +759 steps/day (P=.007) and MVPA +16 min/week (P=.010). MVPA gains retained at follow-up. PMC12805409
Stroke Gamification RCT (2022) — 34 stroke survivors. Loss-framed points + levels + support partner. Result: +981 steps/day (P=.01), goal-achievement days +0.41 (P<.001). JAMA Neurology
Veterans RCT (2021) — 180 veterans. Loss-framed incentives combined with gamification mechanics. Result: +1,224 steps/day (P=.005) — the combined approach was the most effective arm, demonstrating that loss aversion and game mechanics reinforce each other. PMC8271358

How FitCraft applies this: Your workout streak acts as a soft loss-framed system. You don't lose money — you lose visible progress. Research shows this psychological mechanism is powerful even without financial stakes. The streak counter makes the "cost" of skipping a workout tangible and immediate.

Competitive Challenges

FitCraft includes challenge features where you can compete against friends or matched opponents. This isn't just for fun — it's the single most effective social incentive design tested in gamification research.

Supporting evidence
STEP UP Trial (2019) — 602 overweight adults across 40 US states. 4-arm RCT comparing competition, collaboration, support, and control. Competition produced +920 steps/day (P<.001) — the highest of any arm. At 12-week follow-up, only competition remained significant: +569 steps (P=.009). Collaboration lost significance. PMC6735420
iDiabetes Trial (2021) — 361 adults with uncontrolled type 2 diabetes. Competition produced the strongest results: +606 steps/day (P=.003), outperforming both the support arm (+503, P=.01) and collaboration. Confirmed the STEP UP finding that competition is the most reliable social motivator in a clinical population. PMC8144928

How FitCraft applies this: Competitive challenges are a core engagement mechanic. But we learned from the leaderboard research: one-size-fits-all competition hurts already-active users (-630 steps/day in Fitbit observational data). FitCraft uses tiered matching so you compete against people at a similar level, maximizing the motivational benefit while avoiding discouragement.

Self-Chosen Goals

FitCraft lets you set your own workout targets — frequency, intensity, and focus areas. This isn't a UX convenience. It's the most effective goal-setting strategy tested in gamification research.

Supporting evidence
ENGAGE Trial (2021) — 500 adults from lower-income neighborhoods (66% Black, 70% women). Held gamification constant and experimentally varied goal-setting: self-chosen vs assigned, immediate vs gradual. The "self-chosen + immediate" combination was the clear winner: +1,384 steps/day (P<.001), sustained at follow-up +1,391 (P<.001). MVPA: +4.1 min/day (P<.001). This combination produced 2-3x the effect of any other goal-setting approach tested — making it the most effective design choice available. PMC8411363

How FitCraft applies this: When you start FitCraft, the AI coach Ty guides you through a 32-step diagnostic — but you choose your goals, difficulty level, and schedule. The ENGAGE trial is unambiguous: letting people choose their own targets and start immediately produces 2-3x the effect of any assigned-goal approach. FitCraft builds the structure; you choose the destination.

AI-Adaptive Difficulty

FitCraft's AI coach Ty adjusts workout difficulty based on your performance, feedback, and progression data. The goal: keep you in the flow channel between boredom and frustration, where exercise feels challenging but achievable.

Supporting evidence
GAMEPAD Trial (2025) — 103 PAD patients, mean age ~70. 16-week automated gamification + educational texting. Fully home-based, fully automated. During intervention: +920 steps (P=0.06). At 8-week follow-up: +1,074 steps (P=0.03). This is the only trial where effects grew after the intervention ended — suggesting participants internalized the coaching and self-regulated their own progression. PMC12826907
Fitbit Feature Deconstruction (2024) — 70 adults, 4-week pilot. Tested isolated features: self-monitoring vs goal setting vs social comparison. Goal-setting and social comparison trended toward higher steps (P≈.051-.06) with significant changes in motivation/readiness measures. PMC11282379

How FitCraft applies this: Ty continuously recalibrates your workout difficulty. Too easy? The next session scales up. Struggling? It pulls back. The GAMEPAD finding is key: when an automated system teaches people to self-regulate their own progression, the behavior sticks — and even improves — after the formal intervention ends. That's the goal: FitCraft trains you to train yourself.

Collectible Cards

After workouts, you earn collectible cards with varying rarity. You never know when a rare card will drop — that's variable ratio reinforcement, the most powerful reinforcement schedule identified by B.F. Skinner's decades of research.

Supporting evidence
Gamification Meta-Analysis (2022) — 16 RCTs, 2,407 participants. Variable reward mechanisms were components of interventions producing a pooled effect of Hedges' g = 0.42 and ~1,421 additional steps/day. Interventions using game mechanics with unpredictable reward elements consistently outperformed feedback-only controls. Mazeas et al., J Med Internet Res 2022

How FitCraft applies this: Cards come in varying rarities — common, uncommon, rare, and ultra-rare. The unpredictability means every workout could be "the one" where something exciting drops. This keeps the reward circuitry active across sessions, long past the point where fixed-schedule rewards (a badge every 10 workouts) lose their motivational power. It's the same mechanism behind trading card games and gacha systems — redirected toward a healthy behavior.

Quest Progression System

FitCraft structures your fitness journey as a series of quests — themed challenges with clear objectives, visible progress, and escalating difficulty. This directly satisfies two of the three core needs from Self-Determination Theory: competence (you see yourself advancing) and autonomy (you choose which quests to pursue).

Supporting evidence
BE FIT Trial (2017) — 200 adults from 94 families. Used points + levels + daily feedback as a progression system. Result: gamified participants achieved step goals on 53% of days vs 32% for controls (adjusted difference 27 percentage points, P<.001). Mean daily steps +953 vs control. Effects partially sustained at 12-week follow-up. PMC5710273
MapTrek Trial (2018) — 146 sedentary office workers. Mobile "walk race" with leagues, competition, and virtual map progress. Result: +2,183 steps/day (95% CI: 992-3,344). Progression through a virtual map provided visible advancement tied to real-world activity. PMC6064890

How FitCraft applies this: Quests give structure to what would otherwise be "just another workout." Each quest has a narrative wrapper, clear milestones, and a completion reward. You're not just doing push-ups — you're progressing through a chapter. The BE FIT and MapTrek data show that visible progression systems significantly increase goal achievement and daily activity.

Social Accountability

FitCraft's social features let you share progress, challenge friends, and build accountability partnerships. This addresses the relatedness need from Self-Determination Theory — the drive to feel connected to others.

Supporting evidence
BE FIT Trial (2017) — 200 adults in 94 families. Family-based collaboration and accountability drove +953 steps/day (P<.001). The social unit was families with existing ties — highlighting that accountability works best among people who know each other. PMC5710273
Postpartum HDP Trial (2022) — 127 postpartum women. Team-based gamification with social incentives and text messaging. Result: +647 steps/day (P=.009). Demonstrated remote social accountability works even in challenging life circumstances. DOI: 10.1001/jamacardio.2022.0553

How FitCraft applies this: Social features are opt-in — you choose your accountability partners. The research is clear that accountability among people with existing relationships (family, friends) works far better than assigned stranger partnerships. FitCraft makes it easy to invite real friends and family into your fitness journey, not random app users.

Tiered Leagues (Personalized Competition)

One-size-fits-all leaderboards are dangerous. FitCraft uses tiered competition instead — grouping users by activity level so competition motivates rather than discourages.

Supporting evidence
Fitbit Leaderboard Econometrics (2022) — Large observational dataset of Fitbit users. Average effect of leaderboard adoption: +370 steps/day (+3.5%). But the heterogeneity was extreme: sedentary users gained +1,300 steps/day (+15%) while highly active users lost -630 steps/day (-5%). Unmatched leaderboards help beginners but actively harm experienced exercisers. PMC10403254

How FitCraft applies this: Instead of a single global leaderboard, FitCraft groups competitors by activity tier. You compete against people at your level — so a beginner doing 3 workouts/week competes with other beginners, not with someone doing 7. The research is unambiguous: personalized competition works. Generic leaderboards backfire.


Why We Chose This Approach (And What We Do Differently)

Being research-driven also means choosing the approaches that produce the most durable results.


The Bottom Line

Most fitness apps are designed by intuition. FitCraft is designed by evidence.

Every streak, every competitive challenge, every quest, and every adaptive workout maps to a specific finding from a randomized controlled trial. Not a blog post. Not a marketing claim. A clinical trial with a sample size, a control group, a P-value, and a PubMed Central ID.

That matters because the fitness app market is full of apps that "feel" right but don't actually change behavior long-term. The research gives us a blueprint for what actually works — and what doesn't. FitCraft follows that blueprint.

The studies are all linked on our Gamification & Fitness Statistics page if you want to check the data yourself.

See the research in action

Take the free 2-minute diagnostic and experience every evidence-based mechanic firsthand.

Take the Free Assessment Free · 2 minutes · No credit card

Frequently Asked Questions

Is FitCraft backed by scientific research?

Yes. Every core game mechanic in FitCraft maps to a specific finding from peer-reviewed clinical trials published in JAMA Internal Medicine, JAMA Cardiology, JAMA Network Open, and the Journal of the American Heart Association. These include the BE FIT trial (2017, n=200), the STEP UP trial (2019, n=602), the ENGAGE trial (2021, n=500), and the ALLSTAR trial (2025, n=150), among others. FitCraft's streak system, competitive features, self-chosen goals, and progressive difficulty are all grounded in mechanisms that produced statistically significant increases in physical activity in randomized controlled trials.

What fitness app is backed by the most clinical evidence?

FitCraft is one of the few fitness apps that maps its features directly to specific randomized controlled trials. Most fitness apps claim to be "science-backed" without citing specific studies. FitCraft's game mechanics are based on findings from 15 RCTs involving over 2,500 participants, published in top-tier medical journals. Each feature — from streaks to competition to adaptive difficulty — corresponds to a specific mechanism that produced measurable physical activity increases in clinical settings.

How does FitCraft use behavioral science?

FitCraft applies behavioral science through six evidence-based mechanisms: (1) loss-framed streaks based on loss aversion research from 3 RCTs showing +759 to +981 extra steps/day; (2) competitive challenges based on the STEP UP trial showing competition produces +920 steps/day; (3) self-chosen goals based on the ENGAGE trial showing +1,384 steps/day when users choose their own targets; (4) progressive difficulty based on the GAMEPAD trial where automated coaching produced effects that grew post-intervention; (5) collectible card rewards using variable ratio reinforcement schedules from Skinner's research; (6) social accountability based on family/partner mechanisms from the BE FIT trial showing +953 steps/day.

Why does FitCraft use competition instead of collaboration?

Because the evidence is clear: competition outperforms collaboration for fitness. The STEP UP trial (2019, n=602) directly compared competition, collaboration, and support arms. Competition produced +920 extra steps/day versus +637 for collaboration. More importantly, only competition maintained significant effects 12 weeks after the intervention ended (+569 steps, P=.009). The iDiabetes trial (2021, n=361) confirmed the same pattern — competition produced the strongest and most consistent results across both trials. FitCraft uses tiered competition to get the benefits while avoiding the leaderboard backfire effect documented in observational data.