I have prayed earnestly, specifically, and repeatedly for things that mattered — healing, a marriage, a child's life — and heard nothing back. If God were there and listening, prayer would work more often than it does.

Why people find it persuasive

This is often not an abstract philosophical worry but a direct, painful personal experience, and Jesus' own words ('ask and it will be given you,' Matthew 7:7; 'whatever you ask in my name, I will do it,' John 14:13) sound like unconditional promises that experience does not bear out.

Important distinctions

• • Prayer as a vending machine (input request, output result) vs. prayer as relationship, in which requests are one part of an ongoing, two-way engagement with a person who has his own judgment.
• • Absence of the specific answer requested vs. absence of God altogether.
• • Conditional promises embedded in their literary and canonical context ('according to his will,' 1 John 5:14) vs. isolated proof-texts read as blank cheques.
• • Statistically testable, mechanistic claims about prayer's efficacy vs. the personal, relational claim scripture actually makes.

Christian response

• • Jesus' promise texts sit inside a canon that consistently qualifies petition with submission to God's will: Jesus himself, in Gethsemane, prays 'not as I will, but as you will' (Matthew 26:39) immediately before his own most urgent request (to avoid the cross) is not granted as asked — the New Testament's central model of prayer is itself an instance of a request refused, held together with continued trust.
• • 1 John 5:14 explicitly frames confident prayer as prayer 'according to his will,' and James 4:3 diagnoses some unanswered prayer as asking 'wrongly, to spend it on your passions' — the canon itself anticipates and addresses the 'why wasn't this granted' question rather than promising unconditional fulfilment of every specific request.
• • Large controlled studies of intercessory prayer's effect on physical healing outcomes (notably the 2006 STEP study, the largest such trial, involving over 1,800 cardiac surgery patients) found no significant benefit from third-party intercessory prayer on medical outcomes — a finding consistent with the relational model above (God is not a mechanism responding predictably to volume of petition) but inconsistent with any account that treats prayer as a reliable causal lever, and Christians should take this data seriously rather than explain it away.
• • The claim 'prayer would work more often' assumes a baseline of how often it 'should' work that is not independently established — without a controlled measure of what outcomes would have obtained absent prayer, personal anecdote about a specific unanswered request, however painful, cannot by itself establish a general failure rate.
• • Many serious Christian thinkers and much devotional literature (the Psalms' laments, C. S. Lewis's own account of unanswered prayer in A Grief Observed after his wife's death) treat the experience of apparent divine silence as itself a recognized, canonically anticipated feature of faith, not a disconfirming anomaly requiring denial.

Christian counterresponse

Christians generally do not claim to have a satisfying account of every specific instance, only a framework (a good, wise, and free personal agent, not a mechanism, whose reasons in any given case may be inscrutable) that makes continued trust intelligible even amid unresolved grief — which is a considerably more modest and honest claim than that prayer straightforwardly 'works.'