Ask a player who has been stuck at the same game level for a year what they have been studying, and you will usually get a list of topics rather than a plan: a solver output, a preflop chart, three videos on river play, a podcast, half of a book. Every item on that list is reasonable. The list itself is the problem, because nothing in it tells you which item deserves your next twenty hours.
The missing piece is rarely another concept. It is a routing decision — a repeatable way to work out what to study next, and then to check whether studying it changed anything at the table. What follows is that routing system: a four-stage loop of diagnose → prioritize → study → verify, designed to run continuously so that practice time always lands on your biggest current weakness instead of your most interesting one.
If you take a single idea from this page, take this one. The largest available improvement for most players is not a new line or a new chart. It is correctly identifying the most costly repeated decision in their own game — and then attacking it until it stops showing up in their session reviews.
The loop, and what each stage produces
The loop is deliberately mechanical. Each stage has one job, produces one artifact you keep, and takes a rough slice of your study time. Artifacts matter more than insights: an insight you did not write down cannot be verified three weeks later.
| Stage | The question it answers | Rough share of study time | Artifact you keep |
|---|---|---|---|
| 1. Diagnose | Where am I actually losing, in decision terms? | 20% | A leak ledger with measured baselines |
| 2. Prioritize | Which leak is worth working on first? | 10% | A ranked list with priority scores |
| 3. Study | What drill actually fixes this shape of leak? | 50% | A drill prescription and a fixed study block |
| 4. Verify | Did the behavior change, independent of results? | 20% | A verification log with a verdict per leak |
Notice that 30% of the loop is spent on diagnosis and verification — the two stages almost nobody does. That is not an accident. Diagnosis and verification are what separate studying from collecting.
Stage 1: diagnose — build a leak ledger
A ledger is a short, capped list of recurring problems, each with a symptom, an evidence source, and a number you can measure later. Cap it at eight rows. If you have more than eight, you are describing your personality rather than your play.
Use four evidence sources, and use all four, because each one misses something:
- Marked hands. Tag hands during play rather than trying to remember them. Marked decisions are the raw material for everything else.
- Frequency counts. Count how often you take an action in a spot family, not whether the action worked. Frequencies are far more stable than results over short stretches.
- Spot quizzes. Take a decision out of context and answer it cold. If the correct action in a spot is not available to you instantly, you do not own it yet.
- An honest behavior log. One line per session on how you felt and how fast you decided. Emotional leaks hide from statistics.
The table below is the diagnostic core of the system. Each row pairs a leak category with the symptom that signals it, the evidence that confirms it, and the baseline you should record so verification is possible later.
| Leak category | Symptom you would notice | How to confirm it | Baseline to record |
|---|---|---|---|
| Preflop range discipline from early seats | Dominated postflop spots that feel unavoidable | Tag every hand played from the first two seats in five sessions | Hands played from the first two seats per 100 |
| Initiative and aggression | Pots where you never hold the betting lead | Count preflop call-versus-raise decisions | Your raise share preflop |
| Postflop bet sizing | The same size with every hand in every spot | Sort reviewed flops by size used and by texture | Sizes used in single-raised pots, by texture |
| Multiway adjustment | Plans built for two players run into three | Flag every flop seen three-handed or more | Small-bet share in multiway pots |
| Positional play | Checking back turns that deserve a second barrel | Review all button-versus-blind hands | Turn barrel frequency when in position |
| Stack depth awareness | Bluffs that were never going to be folded to | Note effective depth on every reviewed bluff | Bluff frequency at short effective depth |
| Reading opponents | “Reads” you cannot state out loud | Ask yourself what the same opponent did in this spot last session | Written opponent lines logged per session |
| Decisiveness and tilt | Long tanks after a bad pot; rushed calls | Log decision speed and mood once per session | Tilt-flagged sessions per ten |
Stage 2: prioritize — the leak score
Your ledger will contain more problems than you can fix at once, and fixing them in the order you noticed them is a mistake. Score each row on three factors, each from 1 to 5:
- Impact — how much a correct decision here changes the outcome of pots you are already in.
- Frequency — how often the spot appears, per session, in the games you actually play.
- Effort — how much drilling the fix demands before it becomes automatic.
Priority score = (Impact × Frequency) ÷ Effort. Higher is better. The formula is intentionally crude; its job is to stop you from defaulting to whatever you already find easy to think about.
| Leak | Impact | Frequency | Effort | Score | Rank |
|---|---|---|---|---|---|
| Early-seat range too wide | 5 | 5 | 3 | 8.3 | 1 |
| No small-bet plan in multiway pots | 4 | 4 | 2 | 8.0 | 2 |
| Opponent notes too vague to use | 3 | 5 | 2 | 7.5 | 3 |
| Identical bet size on every texture | 4 | 4 | 3 | 5.3 | 4 |
| Bluffing into committed short stacks | 3 | 3 | 2 | 4.5 | 5 |
Two things fall out of a scoring pass like this. First, the “big” conceptual leak just barely beats a narrow technical one, which is why you cannot rank by reputation alone. Second, the note-taking leak climbs to third despite being unglamorous, because it is cheap to fix and it appears constantly. When two scores land close, break the tie in favor of the leak whose verification metric is easiest to count. You want your first loop to succeed, and a countable metric is what makes success visible.
Stage 3: study — match the drill to the shape of the leak
Most study fails here, because the default fix for every leak is “consume more content.” Content suits mapping leaks and almost nothing else. Match the drill to the shape of the problem:
| Leak shape | The default fix that misses | The drill that works |
|---|---|---|
| Range and mapping | Watching more preflop content | Time-pressured recall: name the action for 50 random hands in five minutes |
| Sizing | Reading a chapter on bet sizing | Stop-and-size: for 20 reviewed flops, write the size and your reason before checking anything |
| Frequency and timing | “Just play more aggressively” | Block-and-count sessions: pick one spot, count every occurrence, act on a fixed frequency |
| Adaptation and reads | More theory, no observation | Written opponent lines: three statements per regular, updated every session |
| Stack depth | Memorizing charts you never recall in time | Depth labelling: state the effective depth out loud before you act |
| Decisiveness and tilt | “Focus harder” | A pre-decision routine: one breath, one sentence naming the reason, then act |
Run drills inside a fixed block, so a study session has a beginning and an end rather than a vague duration:
- 10 minutes — reopen the log. Read the previous block’s notes. No new material yet.
- 60 minutes — drill one leak. One. Not two, not “one and a bit.”
- 20 minutes — cold test. Answer fresh spots with no notes visible, then score yourself.
The cold test is the part players skip, and it is the part that tells you whether recognition has become recall. Knowing an answer when it is on the screen in front of you is a different skill from producing it in four seconds with a decision clock running.
Stage 4: verify — prove the behavior changed, not the results
Short-term chip results are noise, and a verification method that leans on them is a coin flip with extra steps. A pot can be played perfectly and still be lost, and a badly played pot can still be won. So verify at the level of the decision: count how often you take the action you planned, in the spots you planned it for.
Give every leak a verification window — a minimum number of occurrences of the spot, not a number of sessions and not a number of pots won. Until you have enough occurrences, the reading is provisional no matter how it feels.
| Leak | Decision-level metric | Baseline | Target | Occurrences needed |
|---|---|---|---|---|
| Early-seat range | Hands played from the first two seats per 100 | 38 | 22–25 | 400 hands |
| Multiway sizing | Share of three-way flops bet at roughly a third of the pot | 12% | 55–65% | 120 multiway flops |
| Opponent notes | Written lines per regular, per session | 0 | 3 | 6 sessions |
Then apply one of three verdicts, and write it down:
- Retire the leak when two consecutive checks land inside the target range. Move on and free the slot.
- Escalate the drill when the metric has not moved at all after the full window — that means the drill is not practising the actual decision.
- Return to diagnosis when the metric moved in the wrong direction, or when a different leak starts dominating your session reviews. Something adjacent is leaking and you were tracking the wrong row.
A full cycle, week by week
Here is how the loop looks when run end to end. Note that only one leak is ever “active” at a time. Two active fixes blur your verification reading, and three make it meaningless.
| Week | Focus | What you actually do | Reading at the end |
|---|---|---|---|
| 1 | Diagnose only — no play changes | Tag hands, run frequency counts, write the ledger, score it | A ranked list of five to eight leaks |
| 2–3 | Priority 1 | Daily 90-minute block on the top leak; count occurrences in play | Metric measured against baseline |
| 4 | Verify priority 1, start priority 2 | Two checks on the first leak; open the second | Retire, escalate, or return |
| 5 | Priority 2 | Continue the block; keep the first leak in the log as a watch item | Second metric measured |
| 6 | Full-loop review | Re-score the ledger, since fixing one leak often promotes another | A new ranked list, and the loop restarts |
Re-scoring matters more than it sounds. Fixing an early-seat range problem changes how often you reach multiway pots, which can change the frequency score of your sizing leak. The ledger is a living document, not a permanent report card.
Failure modes worth checking yourself against
- Studying what is interesting. If your ledger says sizing and your last ten study hours were on river bluffs, the loop is not running.
- Changing three things at once. Verification requires attribution. One active leak, plus at most one mechanical habit you can track without thinking.
- Verifying with results. “I lost three pots with the new line” is not evidence. “I used the new line in 34 of 120 spots” is.
- No recorded baseline. A target without a starting number is a wish.
- Diagnosing forever. The ledger is not the work. If a week passes with no drill, you have been organising, not practising.
- Retiring on one good session. Two consecutive checks, or the leak goes back in the queue.
- Skipping the cold test. Recognition and recall are different skills, and only one of them shows up at the table.
Why recurring lineups make the loop faster
There is a practical reason this system pays off more in club and online-club poker than in a rotating public pool. When you sit with largely the same group of opponents across sessions, your leak categories stay stable and your frequency counts accumulate quickly, because the same spot types keep arriving against the same tendencies. A read you write down on a regular is still valid next week, which makes the adaptation rows in your ledger cheaper to fix than they would be against a fresh table every night.
The same environment rewards the verification stage. In a recurring lineup you can measure your own behavior against a specific opponent — how often you fold to their turn barrel, how often they fold to yours — and get a usable sample far sooner than you could against strangers. The loop does not change; it simply completes faster.
One caveat worth keeping in view: a system like this improves the quality of your decisions, and it says nothing about whether your session budget, schedule, or game selection are sensible. Those are separate questions, and this loop is designed to sit on top of them rather than replace them.
Laws and rules for poker vary by location, and platforms and clubs set their own terms. This article is general strategy information, not legal or gambling advice, and it is not a substitute for checking what applies where you play.
