Opening Gap Finder

How to use Gap Finder #

Tl;dr Find your worst openings, see specifically where they are going wrong, and structure your study in the way that will help your game the most.

What this tool does #

Finds the openings that cost you rating: where your score measurably suffers, measured against your own average and against what players at your level score in the same positions.
Narrows it to the line that is going wrong: an opening is where to start, but a single position is usually where the damage is.
Shows you the alternatives: one position and every move out of it, your games beside the reference’s, so you can see how your choices compare with everyone else’s and what that does to the score.

A worked example, in four steps #

The page opens on the author’s blitz games over the last two years, so you can follow along without typing anything. Figures move as he plays, so read them as shapes rather than exact numbers.

Start with the first opening
Sam’s worst opening is the Bishop’s Opening, where he scores about 42% over roughly 70 games against his own 51% average.
The card calls it problematic because the score is under 45%. That is the whole claim at this stage: this opening is losing him games. It does not yet say where inside it.
Find the line inside it, and read the move list
The cards below the openings name exact positions. His worst is 1.e4 e5 2.Bc4 Nc6, about 24% over roughly 25 games: the same opening, now narrowed to one position. Opening that link puts it on the board above, where the move list is the answer:
Move He plays Peers play He scores They score
Nc3 about 80% about 5% about 24% about 52%
Nf3 never about 24% – about 52%
Read the Nc3 row across. He plays it almost always; his peers rarely. But the last two columns are the ones that matter: players at his level who play Nc3 score about 52% with it, and he scores about 24%.
So the move itself isn’t the problem: it scores perfectly well for everyone else who plays it. The problem is what he does after it, which points at something positional to learn rather than something to stop doing.
Look at what is missing
Step into 3.Nf3, a move about a quarter of his peers play and he has never once chosen from the position before it.
The board shows two games here anyway, reached by a different move order. Two is far too few to judge anything by, and that is the point: this is a well-travelled road he has never deliberately taken.
Go and study it somewhere else
This tool finds the line and stops. It can tell you where you are losing games and what the people beating you play instead; it cannot tell you the right move or why. It provides three options for building on these initial findings:
- Study this opening sits at the end of the third paragraph on a card, wherever the opening has a name. It opens that opening’s page on Lichess, with the main lines and the ideas behind them.
- Open on Lichess sits under the board and carries the exact position you are looking at. That is the one to use when you want an engine, or to play the line out against yourself.
- Copy PGN is beside it and puts the line on your clipboard as a normal PGN, for pasting into whatever you already keep a repertoire in.

Move	He plays	Peers play	He scores	They score
Nc3	about 80%	about 5%	about 24%	about 52%
Nf3	never	about 24%	–	about 52%

One line at a time is enough. The list is ranked so that the top of it is where the rating is.

What to do with your own #

Type your username at the top and the same four steps apply.

Start with the first opening, then the first line. Both rank by what they cost you, which is how far below your average they run times how often you reach them. A mild problem you hit constantly outranks a disaster you meet twice a year.
Compare “You play” with “Peers play”. A move where those two are far apart is the most actionable thing here, in either direction.
Check the game count before believing a percentage. Twenty games and two hundred support very different certainty about the same number.
Try both colours. They are analysed separately, so a clean White page says nothing about your Black.

How it works #

It builds your opening tree out of your own games, then asks a reference database how each of those positions really scores among players around your rating. Openings lead the page and individual lines follow, because an opening is the thing you can go and study.

Every line gets two readings, and both are reported because they answer different questions.

Where do you score badly? Your score against your own average. This decides what is at the top of the page, because it is what tracks your rating. A line you score 32% in is costing you games whether or not anybody else does better there.
Where do you score worse than the line deserves? Your score against what the reference says the position returns at your level. This is the narrower claim: the problem is you rather than the position, so there is something specific to learn. A sharp opening where everyone scores badly is worth leaving; a quiet one where only you score badly is worth studying.

How bad is bad #

Severity is read off your score alone, with no reference involved, so it works on a thin history where the comparisons cannot.

Critical, under 35%. You are losing roughly twice as often as you win.
Problematic, under 45%. Behind by a clear margin.
Borderline, under 50%. Losing ground mildly.

Nothing appears unless it is also below your own average, because a 48% line belonging to a 42% player is one of their better ones. Openings use the same thresholds; the line cards show only the first two, to stay a list of things to do.

Stepping through a line #

The board shows one position at a time with every move out of it, and every line the page names links into it. Two columns say how often a move is chosen, yours and the reference’s, and two say how it then does. That pairing is the diagnosis: a move you play far more than anyone else and score far worse in is a line to drop, and a move the reference plays often and you have never faced is a line to prepare.

Moves you have never played still appear. Every score shown is yours, including on your opponent’s moves, and the address bar follows you so any position can be shared as a link.

Positions, not move orders #

A position is a position, however you got there. 1.d4 d5 2.Nf3 and 1.Nf3 d5 2.d4 reach the same board, so they count as one with their games added together. Counting them separately would halve the evidence for each and could hide a weakness that is really there.

Which positions get looked up #

The first eight moves for each side, past which a position stops being opening theory and starts being a middlegame. Reached at least 10 times, since a position you have played three times tells you nothing. And at most 60 lookups a run, spent on the positions you reach most often, because the reference asks not to be hammered and they go one at a time. What gets cut is the part of your repertoire you meet least. The page says how many positions qualified and how many were looked at, so nothing is hidden.

Lookups are cached for a month and shared, so common openings are near instant and a second run is much faster than the first.

Rating bands are Lichess ratings #

The reference groups its games by rating band, and those bands are Lichess ratings. Your Chess.com rating is on a different scale, and for most club players it reads lower than their Lichess equivalent.

No conversion is applied, because any single number for that would be a guess dressed up as a measurement. Your rating is matched to the nearest band as-is, labelled an estimate, and you can change it. One band is used rather than two, and they are two hundred points wide, so the comparison is against players close to you.

What it reads, and what it keeps #

Your games, from your public Chess.com history, through the same caching service the Tilt Clock uses. Nothing is stored against you and there is nothing to sign in to. The reference is the Lichess opening explorer, hundreds of millions of games grouped by rating band and time control.

White and Black are analysed separately and never mixed, because the same position means opposite things depending on which side of it you sit.