The loop: baseline, isolate, reconnect, prove.

The GuitarHub method is a four-stage loop you run against one goal at a time: baseline, isolate, reconnect, prove. You record where you actually are. You cut the problem down to the smallest piece that still fails. You put that repaired piece back under the pressure of a full performance. Then you record the same thing again and compare the two takes. Each stage has an exit test, so you can tell when it is finished instead of guessing, and the last stage hands you the first stage of the next loop.

It is a loop rather than a curriculum because the useful information arrives late. You cannot know which transition is costing you the song until you have played the song badly onto a recording. Run the stages out of order and you get the most common failure in adult practice: four weeks of honest work on the part that was never the problem.

Stage one: Baseline

Pick one finish line specific enough that a recording can settle it. Play the song from the count-in to the last chord without stopping is a finish line. Get better at rhythm is not. Then record one take: start to finish, no restarts, at a tempo you would actually attempt. Listen once, all the way through, and write down where it breaks.

• Done badly: warming up first. A baseline taken after twenty minutes on the hard part measures your best conditions, which you will not have in week four.
• Done badly: stopping when it goes wrong. The restart is the information. A take that survives to the end shows which failures cascade and which ones you play through.
• Done badly: keeping the best of eight takes. Best of eight is a ceiling. What the month has to move is the floor.

The question this stage answers: where does this fall apart, and when? Not how good you are, which is a rating you can argue with. A timestamp is not.

Move on when you can name the failure in one sentence containing a location. The chorus-to-bridge change drops a beat. If your sentence has no location in it, listen again. If you can hear that something is wrong but cannot place it, the practice plateau diagnostic ends with a named blocker, which is the sentence this stage exists to produce.

Stage two: Isolate

Shrink the failure until it stops failing, along two axes at once. Cut the length, down to the smallest unit that still contains the problem, which for a transition is often one beat either side of the join. Cut the tempo, down to where you play it correctly on purpose rather than correctly by luck. Then repeat it until correct is boring, not until it is occasionally right. Occasionally right is what you already had.

• Done badly: playing the whole song slowly. That is the same problem at a lower speed, plus many repetitions of everything that was already fine.
• Done badly: snapping the tempo back after one clean repetition. One clean repetition is a coincidence until it repeats.
• Done badly: editing the problem away. Simplifying the voicing or switching to an easier fingering is a legitimate arrangement decision and a dishonest fix. Decide out loud which one you are making.

The question this stage answers: what is the smallest change that fixes this?

Move on when you can play the fragment correctly several times in a row, cold: first thing in a session, no run-up. Cold is the test, because in performance the passage arrives unannounced, with your hands wherever the previous bar left them.

Write the tempo down, or the next session starts from a guess. The tempo ladder builder turns one passage into a starting speed, a target, and the steps between. For a join between chords, Strumly's guide to smoother chord transitions covers what to change while you are down there.

Stage three: Reconnect

Put the repaired fragment back and add the pressure you removed, one layer at a time: the bars either side, then the full section, then the full piece, then a click, then a backing track, then one take with no restarts, then a camera.

This is the stage most players skip, and it is why a part works in practice and falls apart in the song. A fragment drilled in isolation comes with its own on-ramp: you know it is coming and your hand is already in position. In the piece it arrives with no warning, with your attention already moving to the next thing.

• Done badly: adding every layer at once. Straight from the isolated loop to a full take with a click and a camera. It falls apart, you decide the fix failed, and you go back to drilling the fragment, which was already fine.
• Done badly: never leaving the loop. The isolated fragment is now the part that feels good, so the week quietly becomes more of it.

The question this stage answers: does the fix survive contact with the music around it?

Move on when you can play the full section at a tempo where the whole thing survives and the repaired spot is no longer what breaks. It does not have to be the best part. It has to stop being the worst one.

Where those layers sit in a week is a scheduling problem too. Strumly's guide to designing a practice routine covers building the week around them. For the session in front of you, the practice session builder turns the minutes you have and the failure you named into blocks that add up exactly, then points its largest block at the repair.

Stage four: Prove

Record the same thing you recorded in week one, the same way: same start and end point, same one-take rule, same tempo range, same point in a session. Then listen to both, in order, once through.

The comparison only works if the conditions match. A cold single take in week one against take nine after a warm-up in week four proves only that take nine beats take one.

• Done badly: deciding from memory. Your memory of week one is generous, because you have heard the improved version many more times since.
• Done badly: recording and never listening back. The file is not the evidence. The comparison is.
• Done badly: letting the finish line drift. A goal quietly rewritten mid-month to whatever now looks reachable will always be met.

The question this stage answers: what changed, and can I hear it?

Move on when you can state the change in a sentence a stranger could check by listening to both files. The bridge transition no longer drops a beat is a claim the recordings settle. It feels smoother is not. Then write down what is now the worst thing: that sentence is the finish line for the next loop.

The song readiness score pressure-tests whether the finished version survives outside the practice chair. The practice evidence log keeps the tempo or clean-pass result beside the session that produced it, and withholds a direction until the same focus has enough sessions to support one.

Why the loop is closed

Each stage produces exactly what the next one consumes. That is what makes this a loop and not a list of good advice.

• Baseline produces the target. Without it, isolate is a guess, and the guess is reliably the part that is most enjoyable to play.
• Isolate produces a fragment that works. Without it, reconnect is playing the whole piece again and hoping.
• Reconnect produces a version that holds up. Without it, prove records a performance that collapses in the same place it always did.
• Prove produces the next baseline. The worst-thing sentence at the end of stage four is the finish line at the start of the next stage one.

The underlying principle is not original to GuitarHub. Practice works when it targets a specific weakness at the edge of what you can currently do, with feedback on whether the attempt worked. That is usually called deliberate practice. What the loop adds is a shape, so target a specific weakness becomes a step you can tell you have finished. There is more on what that looks like at the instrument.

What the method leaves out

Three familiar measurements are missing on purpose. Each measures something real. None measures whether you can play.

Lesson collecting

Saving a lesson is not acquiring a skill, and new material is the cheapest thing available to acquire. Material enters when the loop asks for it. Once the isolate stage has produced the question how do I get from this shape to that one without a gap, a lesson answers a question you already have. Browsing a library first answers questions you never asked, and the pile grows faster than you can play it down.

Watch time

Hours watched measures attendance at somebody else's playing. The unit of progress here is a recording of yours, and nothing on this site counts how long you spent. Watching is an input, and inputs are not evidence.

Streaks

A streak measures showing up, which is real and weak. You can run a sixty-day streak playing only what you already play well, and that is the more comfortable option, so a streak biases each session toward whatever is easiest to finish. The loop counts weeks that ended with evidence instead. A week without a recording produced no evidence, which is a fact about the week and not a verdict on you.

Common questions

How long is one loop?

Four weeks, one stage per week. That is the cadence the planner builds, and it is a default rather than a rule. The exit tests govern the handoff: a stage that meets its test on Tuesday hands off on Tuesday, and one that has not met its test has not finished.

What do I need in order to record myself?

A phone. A voice memo is enough to hear a dropped beat, a missed change, or a tempo that sags. GuitarHub does not upload, store, or listen to your recordings. The planner keeps only your goal and the actions you have checked, in your own browser.

Does this replace a teacher?

No. A good teacher can do the isolate stage for you in about a minute, which is most of what you are paying for. The loop is a structure for the other six days, and it makes lessons more useful, because you arrive with a recording and a specific question.

What if week four shows no change?

Then you have learned something specific and cheap. Either the finish line was wrong for a month of work, or the isolate stage never had a real target, or the fix was never reconnected under pressure. Each points at a different next month, and a clear negative result beats a feeling.

Where to start

If you already know your finish line, build the four weeks. The 30-day breakthrough planner turns one goal into a week-by-week sequence and names the evidence each week has to produce. It needs no account.

If you do not know what to work on, start at the other end. The practice plateau diagnostic ends with a named blocker, which is the input the baseline stage needs anyway.