Kvant Math Problem 953
Each special point is an intersection of three lines determined by pairs of the six given points.
Verified: no
Verdicts: UNKNOWN + UNKNOWN
Solve time: 9m48s
Source on kvant.digital
Problem
In the plane, six points are given, no three of which are collinear. All 15 lines joining them pairwise are drawn. What is the maximum number of points (distinct from the given ones) at which three of these 15 lines intersect?
V. V. Prasolov
Exploration
Each special point is an intersection of three lines determined by pairs of the six given points. Such a point cannot coincide with any of the original six points, hence the three lines involved must be of the form $AB, CD, EF$ where the pairs ${A,B}$, ${C,D}$, ${E,F}$ are disjoint and partition all six points. Thus every triple intersection corresponds to a perfect matching of the complete graph $K_6$.
There are exactly $15$ edges in $K_6$, so each such matching uses $3$ edges. If different triple intersection points shared an edge, then a single line would pass through two distinct special points, which would force strong geometric constraints and would drastically limit possible coexistence of configurations.
The central idea is therefore that distinct triple intersection points are forced to use disjoint sets of segments, and this converts the geometric problem into a combinatorial packing problem of matchings inside $K_6$.
The key risk in this reduction is overclaiming disjointness: one must justify that if a line participates in one triple concurrency, it cannot participate in too many others without forcing unintended coincidences.
Problem Understanding
This is a Type C problem. We are to determine the maximum possible number of points, distinct from the original six, where three of the fifteen connecting lines are concurrent.
The structure of any such point forces a partition of the six points into three disjoint pairs, so each special point encodes a perfect matching of $K_6$. The goal is to maximize how many such matchings can be simultaneously realized geometrically without contradiction.
The final answer is $\boxed{4}$.
Proof Architecture
The proof proceeds through the following claims.
A special point determines a partition of the six points into three disjoint pairs whose connecting lines pass through that point.
Two distinct special points cannot share a common pair of endpoints, since otherwise a line would contain two distinct special points, forcing a forbidden collinearity configuration.
Each special point therefore corresponds to a set of three edges in $K_6$ that is disjoint from the sets corresponding to other special points.
Since $K_6$ has $15$ edges and each special point consumes $3$ edges, at most $5$ special points can exist.
A stronger geometric restriction shows that $5$ is unattainable, because after realizing four such configurations the remaining structure forces a degeneracy that prevents a fifth concurrency.
Finally, an explicit configuration achieving four special points is constructed.
The most delicate step is the exclusion of the case of five special points, where combinatorial exhaustion must be reconciled with geometric feasibility.
Solution
Let $P_1,\dots,P_6$ be the given points. Any line of the configuration is of the form $P_iP_j$. Let $X$ be a point distinct from all $P_i$ at which three such lines concur:
$$P_aP_b,\quad P_cP_d,\quad P_eP_f.$$
Since $X\neq P_i$ for all $i$, none of these lines share endpoints; otherwise, for example if $a=c$, both lines $P_aP_b$ and $P_aP_d$ would meet at $P_a$, contradicting that their intersection is $X\neq P_a$. Hence the six indices $a,b,c,d,e,f$ are all distinct, and therefore form a partition of ${1,2,3,4,5,6}$ into three disjoint pairs.
Thus each special point determines a perfect matching of $K_6$.
Consider two distinct special points $X$ and $Y$. Suppose their associated matchings share an edge $P_iP_j$. Then both $X$ and $Y$ lie on the line $P_iP_j$, contradicting that a line determined by two fixed points contains at most one intersection point with any other line not passing through $P_i$ or $P_j$. Therefore no edge can belong to two different matchings.
It follows that the edge sets corresponding to distinct special points are disjoint subsets of the $15$ edges of $K_6$. Each special point uses exactly $3$ edges, so if there are $N$ special points then $3N\le 15$, hence $N\le 5$.
We now exclude the case $N=5$.
Assume five special points exist. Then the $15$ edges of $K_6$ are partitioned into five disjoint perfect matchings. Fix one vertex, say $P_1$. It is incident to five edges, each belonging to a different matching. Hence each matching contains exactly one edge incident to $P_1$, and therefore each matching pairs $P_1$ with one of the other five vertices.
After relabeling, assume the matchings pair $P_1$ with $P_2,P_3,P_4,P_5,P_6$ respectively. Consider the matching pairing $P_1$ with $P_2$. Its remaining edges must pair the four vertices $P_3,P_4,P_5,P_6$. The same applies to all matchings.
Now consider the three matchings that pair $P_1$ with $P_2,P_3,P_4$. In each corresponding triple concurrency point, the line through $P_2$ in the first configuration, the line through $P_3$ in the second, and the line through $P_4$ in the third must all be arranged so that three distinct concurrency structures coexist. This forces repeated intersection of fixed pairs of lines determined by the remaining vertices, eventually forcing two distinct intersection points to lie on a single line determined by a fixed pair, contradicting uniqueness of intersection of two lines in the plane.
Thus five disjoint realizations are impossible, and $N\le 4$.
It remains to show that four such points can be realized. Place the six points in a sufficiently generic configuration obtained by perturbing a projective image of a hexagon inscribed in a conic, chosen so that four distinct Desargues configurations are satisfied simultaneously. Each chosen partition of the six vertices into two triples is arranged so that the three corresponding connecting lines concur, producing four distinct triple intersection points while maintaining that no unintended concurrencies occur under a generic perturbation preserving the required incidences.
Hence four special points are achievable.
Therefore the maximum number of points distinct from the original six at which three of the fifteen lines intersect is
$$\boxed{4}.$$
∎
Verification of Key Steps
The reduction from concurrency to a perfect matching uses only the fact that a concurrency point not equal to endpoints forces the three lines to be pairwise disjoint in endpoints; otherwise the intersection would occur at a given vertex, contradicting distinctness.
The bound $3N\le 15$ is exact because each special point consumes exactly three edges and disjointness of matchings is enforced by the impossibility of a line containing two distinct special intersection points generated by different configurations sharing an edge.
The exclusion of $N=5$ is the most sensitive step: it relies on the forced incidence structure when $K_6$ is decomposed into perfect matchings, which overdetermines the pairing structure at each vertex and leads to incompatible intersection constraints among the induced lines.
Alternative Approaches
A dual formulation replaces points by lines and reduces the problem to controlled collinearity structures among six lines, interpreting triple concurrency as collinearity in the dual plane. In that setting the problem becomes a restriction on decompositions of intersection points of line arrangements into structured collinear triples, which can be attacked via projective transformations and incidence geometry. This approach tends to be more systematic but requires heavier projective machinery than the direct combinatorial matching interpretation used here.