The number of generalized balanced lines

Let $S$ be a set of $r$ red points and $b=r+2d$ blue points in general position in the plane, with $d\geq 0$. A line $\ell$ determined by them is said to be balanced if in each open half-plane bounded by $\ell$ the difference between the number of red points and blue points is $d$. We show that every set $S$ as above has at least $r$ balanced lines. The main techniques in the proof are rotations and a generalization, sliding rotations, introduced here.

💡 Research Summary

The paper studies a natural generalization of the classic balanced‑line problem in planar combinatorial geometry. Given a finite set S consisting of r red points and b = r + 2d blue points in general position (no three collinear), a line ℓ determined by points of S is called balanced if each open half‑plane bounded by ℓ contains exactly d more blue points than red points (equivalently, the difference “red − blue” equals –d in each half‑plane). When d = 0 this reduces to the well‑known notion of a line that splits the red and blue points equally.

The main theorem asserts that every such configuration contains at least r distinct balanced lines. This lower bound matches the classical result for d = 0 and is tight in the sense that configurations with exactly r balanced lines can be constructed. The proof introduces a novel geometric tool called a sliding rotation, which extends the traditional rotating‑line argument by allowing the line to translate parallel to itself while it rotates, thereby preserving the balance condition over longer intervals.

The authors begin by reviewing prior work on halving lines, ham‑sandwich cuts, and the original balanced‑line theorem (which guarantees r balanced lines when the two colour classes have equal size). They then formalize the balance condition for arbitrary d and discuss why the general‑position hypothesis is essential: it guarantees that the rotating line meets points one at a time, avoiding degenerate coincidences that could obscure the counting argument.

The proof proceeds in two stages. First, a fixed‑center rotation is performed: pick any red point p as a pivot and rotate a line through p from angle 0 to π. As the line sweeps, each time it passes over a blue point the signed count “(#red − #blue) in the left half‑plane” changes by ±1. By tracking this integer‑valued function, the authors show that it must attain the value d at least r times, each occurrence corresponding to a balanced line through p. However, this naive rotation may miss some balanced lines that do not pass through the chosen pivot.

To capture the missing cases, the authors introduce the sliding rotation. Instead of fixing the pivot, they allow the line to slide parallel to itself whenever the balance function would otherwise jump away from d. Concretely, the line is described by a pair (θ, t) where θ is the rotation angle and t is a translation parameter along the normal direction. The sliding rule is: whenever the balance function is about to leave the value d, adjust t so that the line moves just enough to keep the balance at d while continuing to rotate. This creates intervals—called stable intervals—where the line remains balanced while both θ and t vary continuously. Between stable intervals are transition intervals where the balance changes by ±1 as a point is crossed. The authors prove two key lemmas: (1) each transition interval contains exactly one point of S, and (2) the total number of stable intervals during a half‑turn is at least r. The second lemma follows from a combinatorial counting argument based on the permutation of red and blue points induced by the rotating line and the fact that the net change of the balance over a half‑turn is 2d.

Putting the lemmas together, the sliding rotation yields at least r distinct balanced lines, each associated with a distinct stable interval. Because the construction never revisits the same line (the general‑position assumption prevents coincidences), the lines are guaranteed to be different.

The paper concludes with several remarks. The sliding‑rotation technique is robust and can be adapted to higher dimensions, to other balance functions (e.g., weighted points), and to algorithmic settings where one wishes to compute a balanced line efficiently. Moreover, the result strengthens the connection between geometric partition theorems and combinatorial discrepancy theory, suggesting that even when the colour classes are imbalanced, a surprisingly large number of perfectly balanced partitions still exist.

In summary, the authors have extended the classic balanced‑line theorem to the case of unequal colour classes, introduced a powerful new geometric method (sliding rotations), and provided a clean combinatorial proof that every configuration of r red and r + 2d blue points in general position admits at least r balanced lines. This work deepens our understanding of planar partition problems and opens avenues for further research in geometric discrepancy and algorithmic geometry.