Sport psychology, by the evidence

Think of every finding as carrying three labels: its design, its sample and its outcome. A 2022 review of 30 meta-analyses found factors expected to help averaged d = 0.51, a moderate difference, and factors expected to hurt averaged d = -0.21. That sounds settled until you read closer. The same paper lists two earlier reviews of task cohesion at d = 0.24 and d = 1.00.

A stricter 2024 meta-analysis of competitive athletes shows why design matters. Skills training reached g = 0.83, mindfulness and acceptance approaches 0.67, and imagery 0.75. Set aside non-randomized trials and subjective ratings, and none of the three stayed significant. Only three of the seven skills-training studies were randomized, and none had registered its plan in advance.

The second risk is publication bias: studies with significant results are likelier to be seen, so averages run high. The third is the sample. Few reviews isolate elite athletes, and some pooled samples include college students or novices.

The common beginner mistake runs in both directions. One reader takes a confident headline at face value. Another hears "not significant" and concludes a technique does nothing, when it may mean the trials are too few and too weak to say.

Suppose a coach cites a study showing self-talk raised performance. Ask whether athletes were randomized, whether a stopwatch or a coach's opinion measured the gain, and who the athletes were.

Rule: before you repeat a finding, name its design, its sample and its outcome.

Two questions organize this area. The first is why you do it. Self-determination theory places motivation on a line from intrinsic enjoyment, through extrinsic forms such as valuing the result or avoiding guilt, down to having no reason at all. Autonomous reasons feel chosen, controlled ones feel pressured. It proposes three needs: autonomy, competence and relatedness.

The second question is what counts as success. A task orientation measures it by improving, an ego orientation by comparison. Coaches shape this through the climate, meaning what the team rewards. In 17 youth soccer teams, players rated the climate as more performance-focused and less mastery-focused than their coaches did, so your intent as a coach is weak evidence of what the team feels.

Beginners often assume every reward is bad, yet disliked strength work done because you see the point is extrinsic and still autonomous. They also assume goals must be specific. A 2026 expert statement on physical activity cites a meta-analysis where goal setting had a medium effect (d = 0.55) and specific goals were no better than vague ones. Its authors argue the SMART acronym (often specific, measurable, achievable, realistic, time-bound) has been mistaken for the theory. That evidence comes from exercisers, so apply it to competitors with care.

Goals can cost something too. In a small trial of 24 swimmers, an individual time goal went with a larger improvement, 4.3% against 2.5%, and a much larger rise in mood disturbance, 47.1% against 19.1%.

Suppose you are new to a sport and set a score target. A learning goal, such as mastering one skill, may suit you better.

Rule: ask why the athlete is doing this, then choose the goal that fits that person now.

Start by separating what is often lumped together as nerves. State anxiety is a passing condition, trait anxiety a lasting tendency. Cognitive anxiety is worry, somatic anxiety is the racing heart.

The oldest model is the inverted U, from a 1908 law: performance rises with stimulation, then falls, with the best level lower on hard tasks. Multidimensional theory predicts worry hurts in a straight line while bodily arousal follows the U. The catastrophe model adds an interaction. When worry is high, arousal helps up to a point and then performance falls sharply. The individual zones model drops the single curve and asks each athlete to find a personal mix of helpful and harmful emotions, though it does not explain how zones work.

These remain contested. As a 2021 theory paper reports it, one meta-analysis found neither kind of anxiety related to performance, and another found a weak negative link for worry.

Interpretation may matter as much as amount. A challenge state goes with belief in your resources and a focus on succeeding, a threat state with doubt and avoiding failure.

Beginners treat all arousal as the enemy. Suppose your heart pounds at the blocks and you read it as readiness. You may have nothing to fix.

Trials report lower state-anxiety scores after psychological interventions, with larger differences during regular training (SMD -0.94) than just before competing (-0.46).

Rule: ask how much tension there is and how you read it before trying to change it.

Internal focus means attention on your body. External focus means attention on what the movement does. "Open your hips explosively" and "jump toward the ceiling" describe the same jump. The constrained-action hypothesis says body-focused attention interferes with automatic control.

Meta-analyses favor an external focus for short-term results: d = 0.58 on muscular endurance, 0.34 on strength, g = 0.279 on sprinting across only six studies. Long-term strength gains were not reliably different overall. The bigger dispute concerns the literature's average. As a 2026 paper recounts it, one review reported g = 0.264 for performance, and a later, larger one correcting for publication bias shrank that to about 0.01. Treat the size as unresolved.

The trap is turning this into a rule that internal cues are always wrong. A 2023 review of reviews found internal instructions seem better for directly changing movement form, and external ones for whole-movement coordination.

The quiet eye is the last steady gaze on a target before the final movement. Experts show longer ones, which is odd since skill usually means economy. Preregistered work with 127 novice golfers suggests the stillness may help on its own, apart from gathering visual information.

Mental fatigue is contested as well: a review of 11 studies found skills declined after demanding mental work, while a bias-sensitive meta-analysis suggested an exercise estimate of 0.50 might really be about 0.08.

Suppose you are teaching a swing change. An internal cue may be the right tool now, and an external one better near competition.

Rule: choose the cue for the job, and treat any single effect size as provisional.

Choking means performing markedly below your normal level under stress. Anxiety is a feeling that may help or hurt, so the two should not be merged. A bad day alone is not a choke either, since ordinary variation produces those.

Two main accounts compete, and they may each be partly right. Distraction theories say pressure fills working memory with irrelevant thoughts, which fits tasks like mental arithmetic. Explicit-monitoring theories say pressure makes you attend to skills that normally run automatically, which fits a practiced movement that falls apart once you think about its parts. A review adds over-arousal and calls the three complementary.

That difference has a practical edge. Conscious attention is normal while a novice builds a skill, so "stop thinking about it" is advice meant for experts.

Instructions can backfire too. In a study of 57 experienced air-pistol shooters, being told to steer clear of a target zone backfired under anxiety, with shots landing in that zone. So "don't miss left" is a risky cue to give yourself.

Claims about protection deserve care: in a review of 25 trials, most comparisons showed no automaticity advantage for implicit learning. And choking is not universal: 32,274 professional darts throws showed no evidence of choking or of excelling under pressure.

Suppose you tense on the last hole. Ask whether you are distracted or over-controlling, because the two call for different help.

Rule: say what to do, not what to avoid, and diagnose the kind of drop before fixing it.

These skills share a logic: rehearse or cue what you want before you need it. They also share an evidence problem. Most reviews are positive, and the strictest are less sure.

Imagery means creating an experience mentally, and ability, content and purpose can each vary. Mental practice beat no practice and fell short of physical practice in an older synthesis, so it supplements repetitions. A 2025 meta-analysis of 86 studies and 3,593 athletes reported benefits, strongest at about ten minutes, three times a week, for 100 days. That dose comes from comparing studies, not randomizing athletes to doses. The stricter 2024 review found imagery's g = 0.75 lost significance when weaker studies were removed, though its three randomized trials still gave 0.81 with an interval crossing zero, which points to imprecision.

Self-talk may be instructional ("elbow high") or motivational ("come on"). A later trial summarizes a meta-analysis of 32 experiments at d = 0.48, larger on newly learned tasks, and says reviews disagree on which type is better. It can also misfire: with 39 basketball players on an attention task, motivational self-talk raised errors under noise.

A pre-performance routine is a planned sequence before a skill, often a breath, a cue word and set actions. Evidence is mixed. A trial of 54 beach volleyball players found no difference from controls, though the pressure manipulation may have failed.

Beginners expect a quick fix. Suppose you adopt a routine the week of a final. In one self-talk trial eight weeks beat one week. One routine team suspected its 10-minute build-up was too short, though longer ones have also produced null results.

Rule: practice a mental skill in training long before you rely on it, and promise yourself nothing about its size.

Confidence and results move together, and the arrow runs both ways. A meta-analysis of 41 articles and 3,711 athletes put the correlation at r = 0.25, small and stable since older reviews found 0.24. The same paper cites Bandura's point that where feedback is plentiful, confidence ratings track recent performance. Self-efficacy, a performer's belief about executing a specific behavior, is more situation-bound.

Three popular ideas need a warning label.

Mental toughness is an umbrella term with several competing models. A 2017 review linked it to many good outcomes. A published commentary called those conclusions possibly premature, citing unassessed study quality and doubts about what its questionnaires measure. A later meta-analysis found it correlated with narcissism at about r = 0.33.

Ego depletion claims that exerting self-control weakens later self-control. An early estimate was d = 0.62. Bias correction pointed toward zero, and a later paper reports large preregistered replications at d = 0.04, 0.06 and 0.10. Defenders argue the correction was misapplied, and a small meta-analysis in athletes reported SMD -0.38. It remains contested.

The hot hand was long taught as a fallacy. A 2022 basketball paper credits later work with finding a streak selection bias in the original analysis. Its in-game data show some individuals with real streaks, while the average player regresses after consecutive makes.

The beginner's error is treating any of these as settled fact. Suppose a coach benches a player for lacking toughness. Ask what was observed, because the label adds no information the behavior did not already give you.

Rule: when a trait explains everything, ask how it was measured and what would count against it.

A team is defined by interdependence, which varies greatly between a relay and a golf squad. Cohesion is usually split into task and social forms, seen at group and individual levels, so a team can bond over its work without being friends. Task cohesion's effects on performance range from d = 0.24 to 1.00 across reviews, and winning may itself build unity. Team building showed larger effects over more than two weeks, so a single retreat fits this evidence poorly.

Leadership is not only the coach or captain. Athlete leaders can be informal, teammates others follow without a title.

Coaching research centers on autonomy support against control. Controlling behavior, as one scale measures it, means using rewards to control, making regard conditional, intimidating, and exerting excessive personal control. Surveys link controlling coaching with more burnout and with young athletes' anxiety. A struggling athlete may also rate a coach harshly, so the direction is unsettled.

The beginner's mistake is treating harshness as simply a style. In a survey of 995 Canadian national team athletes, neglect and psychological harm were most often reported, and coaches topped the list for every harm except sexual harm. If you suspect abuse or a safety concern, tell someone with the authority to respond, and call emergency services for immediate danger.

Suppose a captain says the team is close, yet nobody passes under pressure. Social ease and task commitment are separate, so ask about each.

Rule: name which kind of cohesion or leadership you mean, and never file harm under coaching style.

For most children, the question is less about performance than whether they stay and what they get from it. A 2025 meta-analysis of 38 longitudinal studies linked participation with better later physical activity, health and wellbeing, at small to medium sizes. Families choose sport, so selection may explain part of that.

Sport is not one exposure. In about 11,000 US children aged 9 to 13, team sport went with lower parent-reported problem scores and individual sport with higher ones. These are cross-sectional data, so individual sport may not be the cause.

Dropping out is often not about motivation. In 4,042 Australian children, lower household income and parent education predicted leaving by age 10.

Adults set much of the tone. A review of 29 studies favored moderate involvement, autonomy support and a focus on improvement.

Early specialization is concentrated training in one sport, more than eight months each year, before puberty. A review of 93 studies linked it to more injury, poorer psychological outcomes and no clear advantage for later success. The mechanisms are largely unknown, so keep that caveat when you advise a family.

Suppose a parent of a 9-year-old asks you about year-round training. You can share the risks honestly and say what is unknown.

Rule: judge youth sport first by whether the child keeps wanting to come back.

Keep two ideas apart. Burnout is a psychological syndrome with three parts: emotional and physical exhaustion, a reduced sense of accomplishment, and devaluation, a detached and negative attitude toward sport. Overtraining describes reduced performance after hard training without enough recovery, and that physiology belongs mainly to exercise science.

The training side runs along a spectrum. Functional overreaching is a short dip, days to a few weeks, that ends in improvement. The non-functional kind drags on for weeks to months. Overtraining syndrome usually runs several months. More than ninety years after it was first recognized, it has no definitive diagnosis or reliable biomarker, and a suspected case belongs with a sport physician.

Longitudinal studies describe burnout as nonlinear, with 26 factors sorted into risks, protections and consequences. Socially prescribed perfectionism, the sense that others demand perfection, correlated with exhaustion and devaluation, at small sizes.

Intervention evidence is thin. A meta-analysis in young athletes found only five eligible studies, and a review of mindfulness and burnout found only two weak controlled trials.

The beginner's mistake is reading every tired athlete as burned out, or every drop in interest as laziness. Suppose a swimmer who loved training now shrugs at results for weeks. The change in attitude is your clue, and it warrants a conversation about load, recovery and involving a professional.

Rule: tiredness that passes with rest is training; detachment that does not is a reason to get help.

An injury is also a psychological event, and physical recovery does not guarantee a return. In one cohort after anterior cruciate ligament (ACL) surgery, 54 of 63 people aimed to return to their earlier activity, and at one year 17 were recorded as back and 25 as not.

Athletic identity cuts both ways. A narrative review in young athletes linked strong identity to reluctance to report symptoms and playing through pain, and also to stronger coping and better return outcomes.

Readiness can be measured. Two years on, 42% of 103 patients were back at their earlier level, and a psychological readiness scale predicted this where hop and strength tests did not. Its cutoff had a specificity of 45%, so it wrongly flags many people. Treat it as a prompt for a conversation.

Fear tends to be specific. Fifty-four athletes in ball sports named moments such as reacting to an opponent, jump-landing, contact and loss of balance. "Are you scared?" is a weak question, and "in which moment?" gets you somewhere.

Concussion deserves separate care. A review of 27 studies linked it to depression symptoms in elite athletes, with several reporting resolution by about a month and cause undetermined. Those symptoms go to a qualified clinician.

The common error is treating a physically cleared athlete as ready. Suppose a player passes every test and still avoids contact in practice. That hesitation is information.

Rule: ask which moment the fear shows up in, and leave clearance to the clinician.

Performance coaching and clinical care are different jobs. If you coach or study this field, your role is to recognize and refer, never to diagnose.

Athletes are not exempt. A 2016 review judged elite athletes' anxiety and depression risk broadly like everyone else's, while noting that only a quarter of its 60 studies were rigorous. A recent meta-analysis of self-reported symptoms in current elite athletes gave figures between 4% (drug misuse) and 33% (distress). Those are questionnaire symptoms, with large differences between studies.

The barrier is often the sport itself. In focus groups, young elite athletes named stigma as the main obstacle, and a coach's positive attitude as a help. Programs raising awareness improved knowledge and stated intentions, but in one review they did not change actual help-seeking, so a talk alone is not enough.

Disordered eating and relative energy deficiency are refer-only. Screening scores reached cutoffs in about 19% of athletes across 177 studies. Eating disorders can be life-threatening, and early treatment matters. Relative energy deficiency is a widely adopted model whose diagnosis is itself contested, since energy availability is hard to measure in the field. The practical response is the same: do not give weight or diet advice, and involve a sport physician.

Involve a licensed clinician when mood, sleep, eating, substance use or thoughts of self-harm worry you. In a life-threatening situation, call 911. In a crisis in the US, call or text 988, which offers 24-hour confidential support.

Suppose a teammate says everyone would be better off without them. That is not a performance issue, so reach out for help now.

Rule: notice, say something, and connect the person with care instead of managing it alone.