A colored square is the last step in reading a preflop chart, not the first. The important information is the game, seats, stacks and actions that produced it.
Read the header before the hand.
Identify the format, number of players, positions, starting stacks, blinds and antes. In cash games, check the rake model. In tournaments, distinguish a chip-EV reference from one modeling payouts or bounties. “20 BB” does not tell you whether the table is heads-up or nine-handed, or whether the hand begins before or after forced bets.
Next, reproduce the action sequence. An unopened button chart does not answer a button decision after a raise and a caller. A push/fold solution answers a restricted action problem; it does not establish that a small raise or limp is inferior in a game where those actions are available.
GTO Wizard's Study Mode documentation shows selection by positions and actions. Use a provider's setup information to identify the spot rather than treating a screenshot as a complete model.
A 169-cell matrix is not 169 equally likely deals.
In standard hold'em, the 13 by 13 matrix groups the 1,326 unordered two-card combinations into 169 hand classes. Before any known-card removal, a pocket pair has 6 suit combinations, a suited nonpair has 4, and an offsuit nonpair has 12.
| Class type | Classes | Combos each | Total |
|---|---|---|---|
| Pocket pairs | 13 | 6 | 78 |
| Suited nonpairs | 78 | 4 | 312 |
| Offsuit nonpairs | 78 | 12 | 936 |
Thirteen plus 78 plus 78 gives 169 classes, while 78 plus 312 plus 936 gives 1,326 combinations. Selecting one offsuit class can therefore represent three times as many combinations as selecting one suited class. Read the matrix's actual labels and legend rather than assuming every product places suits or colors identically. SplitSuit's range-reading guide connects matrices, written ranges and percentages.
A mixed color describes a conditional frequency.
When a range is conditioned on earlier actions or known cards, both weights and available combinations can change. Do not average cell percentages without considering their weights. Use the combinations workshop for blocker and weighted-range examples.
Check what “effective” left out.
Fictional mismatch: you recorded 20,000 chips before posting a 1,000 big blind and a 1,000 big blind ante. The record is 20 BB before posting, but 18,000 chips remain behind. A source asking for a pre-posting stack and a source asking for chips behind need different inputs. Do not subtract the forced bets twice.
With several opponents, a single “effective stack” can conceal different caps on different pots. Preserve the relevant players' stacks, positions and contributions. A two-player chart does not become a multiway answer by entering the smallest number at the table. The short-stack guide supplies the input checklist.
Use the output to ask a better study question.
Once the spot matches, compare the available actions and their expected values. A colored preference without EVs does not reveal whether alternatives are nearly tied or substantially different. A small displayed difference may also be smaller than the uncertainty created by your assumed ranges or approximate model.
Make one controlled comparison: same positions and objective, but a different opening size or clearly documented range assumption. Label it as a model comparison, not a measurement of how opponents actually play. An exploitative adjustment needs a defensible read; a result alone is not that read.
Save the source, version or access date, full inputs, and the question answered. Do not redistribute a provider's paid charts. This page teaches chart literacy; it does not publish validated opening or push/fold ranges.
Three useful checks.
Does one AKs square carry the same number of combinations as one AKo square?
No. Before blockers, AKs has four combinations and AKo has twelve. Equal-looking cells do not have equal probability mass.
A chart says “button, 20 BB.” Can it settle a decision after someone has already raised?
Not without the action history and other settings. An unopened-pot reference answers a different question. Mark the mismatch rather than borrowing the chart's precision.
Can you use this guide or its linked charts during a live hand?
Use this site's study material away from play. Some online policies also restrict tools while the playing client is open. Check the rules and fair-play guide and the operator's own policy.
Go to the source
These selections retain the directory's access labels, individual review dates and limitations. Public-page reading is not a product accuracy test.
James Sweeney / SplitSuit
Poker ranges: matrices, combinations and percentages
An author-written introduction to expressing ranges as matrices, combinations and percentages.
Before you use it: Public written guide read, not every embedded video or promoted product tested. Suit placement and color legends should still be checked in the actual chart.
Public page read · · Full directory listing
GTO Wizard
GTO Wizard Study Mode documentation
Learn how the study interface specifies a spot and displays its solution.
Before you use it: Documentation is free to read; software access is separate.
Public page read · · Full directory listing
GTO Wizard
A beginner's guide to poker combinatorics
Learn how counting specific card combinations supports range analysis.
Public page read · · Full directory listing
HoldemResources
Heads-up push/fold reference
A heads-up push/fold reference with explanation of the model and how to read the tables.
Before you use it: Heads-up and restricted-action scope. Do not apply it as a six-player tournament solution.
Public page read · · Full directory listing
Prepared with AI assistance. Published . Worked examples are fictional, not Nolan's hands or results. This is educational material for before or after play, not live assistance.