What actually matters in a coach, why the FIDE title says less than it seems, and the five questions worth asking in the first lesson.
Choosing a coach is rather like choosing a doctor: almost everyone looks at the credentials, and what decides the outcome is something else.
A FIDE title — master, international master, grandmaster — certifies playing strength. That is real information and it is not the information that matters most to you.
The reason is simple: taking someone from 1200 to 1800 requires explaining concepts a FIDE master already commands comfortably. The difference between 2300 and 2600 changes nothing in that task; the difference between explaining well and explaining badly changes all of it.
Where the title does weigh: above 2000, when what is needed is repertoire preparation and work on complex games. There the coach's strength starts to be a real ceiling.
That they routinely teach at your level. A coach who works daily with 1300-rated players knows exactly what those players struggle with, in what order it gets fixed and which explanation works. One who only trains titled players will give you excellent, badly calibrated material.
That they ask for your games before the first lesson. It is the most reliable signal there is. A coach who wants to see how you play before talking to you is doing it right; one who starts with their programme without looking at you is not.
That they can explain something simple simply. In the first lesson, ask them to explain why a bad bishop is bad. If the answer is clear and comes with an example, good sign. If it is long and full of variation names, bad one.
That they speak your language comfortably. It sounds obvious and it is not: chess is taught with nuance — "here it is usually better", "here it is almost always" — and that nuance disappears in a language neither of you fully commands.
Before committing to a long plan, one standalone lesson with these questions tells you almost everything:
1. "What do you see in my games?" You want a concrete, ranked answer: "your main problem is this, and this other one is secondary." A generic reply — "you need tactics" — is not useful.
2. "What should I train over the next two months?" A short, specific plan should come out of this. If the answer is "we will see how it goes", something is missing.
3. "How do we work between lessons?" The right answer includes homework. A coach who leaves you no work between lessons is selling hours, not progress.
4. "What level do you usually work with?" There is no bad answer, just answers that fit you or do not.
5. "What do you expect from me?" A good coach has expectations: how many games a week, how much study time, what to write down. Stating them is a sign of method.
Promising a rating. "I will get you to 2000 in six months" is something nobody can know. Progress depends above all on what you do.
Spending the lesson on their own games. Entertaining, and not your study material.
Not adapting the material. If lesson three is identical for you and for the previous student, there was no diagnosis.
Selling you a long package at first contact. A twenty-lesson plan before seeing one of your games is a sale, not a plan.
The sensible route:
If after two lessons you do not have a clear plan and a concrete assignment, the problem is not your level: it is the fit. And changing coach is not a failure.
A badly used first lesson tells you nothing, and it is the one that decides whether you hire somebody for six months. Three things that make it work:
Send three games beforehand. Two losses and a draw or a hard-won win. Pick them from recent weeks, at a long time control. Have them looked at before the lesson, not during: the lesson is for talking, not reading.
Write down what you think you are missing. Two or three sentences. Then compare it with what you are told. If it matches, you have an identified problem and a plan; if it does not match, you have something far more valuable: a blind spot named by somebody seeing it from outside.
Ask what you should do between this lesson and the next. It is the question that separates a coach from a commentator. If the answer is vague — "keep practising" — the lesson was entertainment.
And a warning about how to read that first lesson: liking them is not sufficient information, but disliking them is. You are going to spend hours showing this person your worst games; if you are embarrassed to do it, you will learn nothing however excellent they are.
This is the decision that saves the most money and almost nobody thinks about it: frequency matters more than quantity.
One lesson a week only makes sense if you have five or six hours a week to work between them. Otherwise you arrive without having done the homework and the lesson gets spent repeating.
One every two weeks is what works for almost everybody who works or studies. It leaves time to play four or five games and to work on one subject.
One a month works surprisingly well for a disciplined person. The lesson becomes a review and a course correction, and in between you work alone.
A package of four and then nothing is the best option for the commonest case of all: somebody stuck who needs a diagnosis rather than accompaniment. Four lessons are enough to be told what you are missing; what follows, you do yourself.
The rule: more lessons is not faster. Progress is produced by the work between lessons, and the lesson only steers it.
At three months, three honest questions:
Can you name what you are working on? If the answer is "chess in general", there is no plan. Good training has one subject at a time and you know what it is.
Has anything measurable changed? Not the rating — which moves with a lag — but something concrete: fewer pieces lost, fewer losses from winning positions, better clock management. Something you can count.
Are you doing more work on your own than before? It is the most important sign of all. A good coach makes you more autonomous, not more dependent. If after six months you do not know how to study without them, something went backwards.
And the sign that it is time to stop, which also has to be read: when the lesson stops producing new errors. At that point what is missing is not teaching, it is mileage, and that is bought by playing.
They will not study for you, will not analyse your games for you, and will not make you improve by you showing up. What they will do is make the hours you were already going to spend on chess pay double, because they will be spent on the right thing.
That is a lot. It is, in fact, the only thing that can be bought in this game.
How much stronger than me do they need to be? Between four hundred and eight hundred points, which is the band where they understand your errors because they made them recently and still remember. Somebody fifteen hundred points above often can no longer explain why you find hard something that comes to them automatically.
Does coaching over video call work? Just as well as in person for almost everything. The only loss is work on a physical board, which matters if you are going to compete; it is compensated by replaying the games on a board at home while you talk.
Can I train with a stronger friend instead of paying? Yes, and it works quite well on one condition: that there is structure. A friend who goes through one game a week with you pays off; one you only play against does not. What a coach adds is method, and that can be agreed on.
Is it a bad sign if they do not make me play during the lesson? Yes. A lesson where you only listen returns a fraction of one where you get asked. The healthy ratio is more than half the time with you deciding moves.
How much should a lesson cost? Between fifteen and thirty dollars an hour gets you a good coach for club level across most of Latin America. Above forty you are paying for playing strength that does not return anything at club level.
Can I change coach without a problem? Yes, and it is normal: a coach who suited you at 1300 may not be the best at 1700. No explanation and no drama are needed; the only thing worth doing is taking the record of what you worked on with you.
How long before a lesson shows? Three to six weeks, and not in the rating but in the type of error. If after three months you cannot name what changed, something is wrong in the method or in the work between lessons.
Is a coach useful for a seven-year-old? Less than a club. At that age half the hook is social and there is one thing to teach — look before you move — which does not need a private lesson.
If you want to put this into practice with structured material:
A comprehensive six-stage course to improve decision-making, master endgames and openings, and develop tactical and strategic resources for your games.
ARDFBy GM Andrés Rodríguez and GM Diego Flores
A collection of six courses by GM Andrés Rodríguez to expand your opening repertoire, recognize weaknesses, and make better strategic decisions.
ARBy GM Andrés Rodríguez
What you can do for free, what is worth paying for and when, and three concrete budgets depending on how much you want to invest.
Read the articleHow far free material takes you, where it stops, and how to use it well so you do not spend years knowing a lot and playing the same.
Read the articleThe three ways of studying chess compared by what they cost, what they give, and who each one genuinely suits.
Read the articleARDFBy GM Andrés Rodríguez and GM Diego Flores
