What Is Knot Theory?
Knot theory is the study of closed loops in three-dimensional space. A mathematical knot is like a loop of string whose ends have been joined together, so it cannot be untied by pulling on loose ends. Two knots are considered the same if one can be smoothly deformed into the other without cutting the loop, gluing new parts together, or passing one strand through another.
Because three-dimensional knots are hard to compare directly, mathematicians often draw them as flat diagrams. A knot diagram shows a projection of the loop onto a plane. At each crossing, the diagram records which strand passes over and which strand passes under. That over-under information is essential: without it, a drawing is only a shadow of the knot.
The main challenge in knot theory is that the same knot can look very different in different diagrams. A simple knot may be drawn with extra twists, loops, or crossings. Knot invariants help solve this problem by assigning values to a knot that stay the same even when the diagram is redrawn in an equivalent way.
Why Knot Theory Matters
Knot theory is useful because it turns visual questions into precise mathematical ones. Instead of asking only whether two drawings look alike, knot theorists ask whether the loops can be transformed into one another by allowed deformations.
This makes knot theory a helpful introduction to topology, the branch of mathematics that studies shapes through deformation rather than measurement. It also connects visual reasoning with algebra, because many knot invariants are numbers, polynomials, or coloring rules derived from a diagram.
For students and educators, knot diagrams are especially valuable because they make abstract ideas visible. A drawing can show crossings, over-under structure, transformations, and mistakes that would be difficult to explain only with symbols.
Key Terms to Know
-
Knot: A closed loop in three-dimensional space, considered up to smooth deformation.
-
Unknot: The simplest knot, equivalent to a plain circle with no essential knotting.
-
Knot diagram: A flat drawing of a knot that shows over-under information at crossings.
-
Crossing: A point in a diagram where one strand passes over or under another strand.
-
Visible crossing count: The number of crossings seen in a particular drawing.
-
Crossing number: The smallest possible number of crossings among all diagrams of the same knot.
-
Reidemeister moves: Three local diagram moves that change the drawing without changing the knot type.
-
Invariant: A quantity or property that remains the same for equivalent knots.
-
Alexander polynomial: A polynomial invariant associated with a knot.
-
Determinant: A positive integer invariant often computed from the Alexander polynomial.
-
Unknotting number: The smallest number of crossing changes needed to turn a knot into the unknot.
-
Tricolorability: A rule-based coloring test that can distinguish some knots from others.
-
Torus knot: A knot that can be drawn on the surface of a torus, controlled by two winding numbers.
How Knot Diagrams and Invariants Work
A knot diagram is not just a picture. It is a compact way to record a three-dimensional loop. The curve is drawn in the plane, and every crossing tells you which strand is above the other.
The same knot can have many different diagrams. Reidemeister moves explain why. These moves let you:
-
add or remove a small twist,
-
add or remove a pair of nearby crossings,
-
slide a strand across a crossing.
If two diagrams are related by a sequence of Reidemeister moves, they represent the same knot. This is why a diagram with more visible crossings is not automatically a more complicated knot.
For a particular diagram \(D\), let \(c(D)\) be the number of crossings visible in that drawing. The crossing number of a knot \(K\) is the minimum crossing count over all diagrams of that knot:
$$
c(K)=\min_D c(D)
$$
The difference between \(c(D)\) and \(c(K)\) is important. A messy drawing of the unknot may have several visible crossings, but the unknot has crossing number \(0\) because it can be redrawn as a plain circle.
Knot invariants help compare knots without checking every possible diagram. If two knots have different invariant values, they are definitely different knots. If they have the same invariant values, they may still be different, because many invariants are useful but not complete.
One important invariant is the Alexander polynomial, usually written as \(\Delta_K(t)\). Different sources may use slightly different but equivalent polynomial conventions. A closely related integer invariant is the determinant:
$$
\det(K)=\left|\Delta_K(-1)\right|
$$
The unknotting number is another way to measure how far a knot is from the unknot. It asks for the smallest number of crossing changes needed to turn the knot into a plain circle:
$$
u(K)=\min\{\text{crossing changes needed to transform }K\text{ into the unknot}\}
$$
Tricolorability gives a more visual invariant. A diagram is tricolorable if its strands can be colored using three colors so that at least two colors are used and, at every crossing, the incident strands are either all the same color or all different colors. This test is simple enough for beginners but still powerful enough to distinguish the trefoil knot from the unknot.
A 3D visual projection can also be drawn using a torus-style parametric curve. A common form is:
$$
x=(R+r\cos(qt))\cos(pt)
$$
$$
y=(R+r\cos(qt))\sin(pt)
$$
$$
z=r\sin(qt)
$$
Here, \(R\) and \(r\) set the torus shape, \(t\) is the curve parameter, and \(p\) and \(q\) control winding behavior. This kind of formula is useful for visualization, but it is not the same thing as proving knot equivalence or recomputing knot invariants from an edited drawing.
Examples of Knot Theory in Practice
Example 1: A visible crossing count is not always minimal
Suppose a diagram of the unknot has two visible crossings because the loop was drawn with a twist. That diagram has:
$$
c(D)=2
$$
But the unknot can be redrawn as a simple circle, so its crossing number is:
$$
c(\text{unknot})=0
$$
The diagram looks more complicated than the knot really is. This is why knot theory separates the crossing count of one drawing from the minimal crossing number of the knot itself.
Example 2: Finding a determinant from an Alexander polynomial
For the trefoil knot, one common Alexander polynomial convention is:
$$
\Delta(t)=t^{-1}-1+t
$$
To find the determinant, evaluate the polynomial at \(t=-1\) and take the absolute value:
$$
\Delta(-1)=(-1)^{-1}-1+(-1)
$$
$$
\Delta(-1)=-1-1-1=-3
$$
So the determinant is:
$$
\det(\text{trefoil})=|-3|=3
$$
This integer does not describe the whole knot, but it gives useful invariant information.
Example 3: Comparing a figure-eight knot diagram with its crossing number
The figure-eight knot has crossing number \(4\). If a particular non-minimal projection shows six visible intersections, then:
$$
c(D)=6
$$
while the stored crossing number for the figure-eight knot remains:
$$
c(\text{figure-eight})=4
$$
That difference means the current drawing is not a minimal diagram, or that a visual crossing detector is counting the sampled drawing rather than proving a topological fact. It does not mean the underlying preset invariant has changed.
Example 4: Tricolorability as a quick test
The trefoil knot is a classic example of a tricolorable knot. The unknot is not tricolorable under the usual nontrivial coloring rule, because it cannot use at least two colors while satisfying the crossing rule.
This gives a simple way to see that the trefoil is not equivalent to the unknot. However, tricolorability is not a complete classification method. Many different knots can share the same tricolorability status.
How to Interpret the Result
The result should be read as a combination of visual information and preset knot data.
-
Library crossing number is the proven minimal crossing number for the selected knot.
-
Current projection crossings is the number of transverse self-intersections detected after projecting the current camera view. It can change as the same knot is rotated.
-
Result summary identifies the selected library knot and reports its proven crossing number.
-
Determinant is a dimensionless integer invariant for the selected preset knot.
-
Unknotting number is the stored number of crossing changes needed to unknot the selected preset.
-
Alexander polynomial is a polynomial invariant in the variable \(t\).
-
Tricolorability reports whether the selected preset has the stored tricolorability property.
-
Library note gives a short description of the selected knot.
-
3D projection and the crossing diagram are two views derived from the same preset embedding.
-
Downloaded graph is a PNG image of the current canvas, not a symbolic knot file.
A low crossing number usually means the knot has a simple minimal diagram. A higher crossing number means every regular diagram requires at least that many crossings. A non-minimal camera angle may show more crossings than the library crossing number.
The most important interpretation rule is this: the invariant cards describe the selected preset knot, while the visible crossing count describes the current sampled drawing.
Common Mistakes and Misconceptions
-
Confusing visible crossings with crossing number: A diagram can have extra visible crossings even when the knot has a smaller minimal crossing number.
-
Assuming rotation changes the knot type: Rotating the camera changes the planar projection, not the embedded knot.
-
Assuming every invariant is derived from the current pixels: Determinant, Alexander polynomial, tricolorability, chirality, and unknotting number are verified library data; only the projection crossing count is detected geometrically.
-
Treating the Alexander polynomial as a complete fingerprint: Different knots can share the same Alexander polynomial, so matching polynomials do not always prove two knots are the same.
-
Treating one projection as a proof of minimality: Matching the library crossing number is a useful visual check, but the detector does not prove that the projection is minimal.
-
Expecting animation without the 3D projection: Animation depends on the 3D projection being enabled.
-
Assuming a PNG export contains knot data: A downloaded image preserves the drawing visually, but it does not store a Gauss code, braid word, polynomial, or other symbolic knot description.
When to Use Knot Theory Concepts
Use knot theory concepts when you want to:
-
compare simple knots such as the unknot, trefoil, figure-eight, and cinquefoil,
-
understand why one knot diagram can look different from another diagram of the same knot,
-
distinguish a visible crossing count from a minimal crossing number,
-
introduce invariants such as determinant, Alexander polynomial, unknotting number, and tricolorability,
-
explore how camera rotation affects a knot projection,
-
teach or learn topology through visual examples.
These concepts are especially useful when a drawing alone is misleading. A diagram may look complex because of extra twists, or it may look simple while still representing a nontrivial knot. Invariants provide a more reliable way to reason about the underlying knot type.
Limitations and Things to Keep in Mind
This calculator is best understood as an introductory visual explorer. It supports a small set of preset knots: the unknot, trefoil knot, figure-eight knot, and cinquefoil knot.
The invariant values are verified library values for the selected knot. The tool does not accept or classify arbitrary diagrams, Gauss codes, braid words, or equations. Camera rotation can change the projection crossing count while the knot and its invariants remain unchanged.
The visible crossing count is based on a sampled drawing. It can be affected by the shape of the curve, pixel-level geometry, sampling density, and how close crossings are to one another. It should be read as a visual count for the current diagram, not as a formal proof of the knot's crossing number.
At a detected crossing, depth in the rotated 3D embedding determines which strand passes over. Nearly tangent or nearly equal-depth intersections are reported as ambiguous rather than presented as reliable crossings. Formal work with arbitrary knots needs an exact representation such as a planar diagram code, Gauss code, Dowker-Thistlethwaite notation, or braid word.
The curves are sampled numerically, so extreme viewing angles or near-tangencies can challenge crossing detection. The reset view provides a regular projection whose detected crossing count matches the library value for every included knot.
For classroom learning, these limitations are usually acceptable. For research, publication, or proof-based work, confirm results using formal knot notation, a reliable knot table, or specialist mathematical software.
How to Use This Calculator
-
Choose a preset knot from the knot preset selector.
-
Use the crossing and 3D projection options to control what appears in the diagram.
-
Turn on animation only after enabling the 3D projection.
-
Drag the canvas, use the arrow keys, or choose Rotate left/right to change the camera view.
-
Use Reset view to return to the verified regular projection.
-
Read the result summary, metric cards, and invariants table to compare library data with the current projection.
-
Use the download graph control to save the current canvas as a PNG image.
Frequently Asked Questions
What is the difference between a knot and a knot diagram?
A knot is the three-dimensional loop itself. A knot diagram is a flat drawing of that loop with over-under crossing information. The same knot can have many different diagrams.
Why can the projection crossing count differ from the library crossing number?
The current projection count belongs to one camera view. The library crossing number is the smallest possible crossing count across all regular diagrams of that knot. A non-minimal viewing angle can make the projected count larger.
Does the Alexander polynomial identify every knot?
No. The Alexander polynomial is a useful knot invariant, but it is not complete. If two knots have different Alexander polynomials, they are different; if they have the same Alexander polynomial, they may still be different.
What does the determinant tell me?
The determinant is a positive integer invariant that can be computed from the Alexander polynomial using \(\det(K)=|\Delta_K(-1)|\). It is easier to compare than a full polynomial, but it contains less information.
What does tricolorability mean?
Tricolorability means the strands of a knot diagram can be colored according to a specific three-color rule. It is a beginner-friendly invariant that can distinguish the trefoil from the unknot, but it cannot distinguish all knots.
Can I enter my own knot notation?
No. This calculator is a four-preset explorer with rotatable views. It does not accept custom Gauss codes, braid words, Dowker-Thistlethwaite notation, planar diagram notation, or user-entered knot formulas.
Sources and References
Books
-
Colin C. Adams. The Knot Book: An Elementary Introduction to the Mathematical Theory of Knots. American Mathematical Society, 2004. Relevant sections: tricolorability, knot tabulation, unknotting number, crossing number, and torus knots. ISBN 978-0-8218-3678-1. American Mathematical Society
-
Richard H. Crowell and Ralph H. Fox. Introduction to Knot Theory. Graduate Texts in Mathematics, Vol. 57, Springer, 1977. Relevant sections: introductory knot theory, diagrams, and algebraic invariants. Springer
-
Dale Rolfsen. Knots and Links. AMS Chelsea Publishing reprint, 2003; originally Publish or Perish, 1976. Relevant sections: knot diagrams, knot groups, Alexander polynomials, and knot invariants. Google Books preview
Online and Reference Sources
-
Eric W. Weisstein. “Alexander Polynomial,” “Tricolorable,” and “Torus Knot.” Wolfram MathWorld, accessed June 29, 2026. Alexander Polynomial, Tricolorable, Torus Knot
-
Dror Bar-Natan and contributors. “3_1,” “4_1,” and “5_1.” The Knot Atlas, accessed June 29, 2026. 3_1, 4_1, 5_1
