For precise algebraic form, consult the (section “Key lemma” or “Key polynomial”) or the final paper: “Density Hales-Jewett and Moser numbers” (2012).
Prior proofs gave extremely weak bounds (e.g., Ackermann-type or tower-of-exponentials). Polymath 6.1 sought to reduce the tower height.
[ P(\mathbfx) = \sum_i=1^n \omega^x_i \quad \text(where $\omega$ is a primitive 3rd root of unity) ] polymath 6.1 key
Let $x_1, x_2, \dots, x_n$ be variables in $0,1,2$ (or $\mathbbF_3$). Consider:
[ \textKey function: f(x) = \text(# of 0's) - \text(# of 1's) \quad \textmod something? ] For precise algebraic form, consult the (section “Key
Existing approaches involved iterating a “density increment” step, but each step reduced the dimension dramatically. The key polynomial helped track density increments more efficiently. 4. Specifics of the “Key Polynomial” While Polymath 6.1 did not name one single polynomial “the key,” the following polynomial (or its variants) played the central role:
But the actual breakthrough came from (e.g., $\mathbbF_3^n$). A specific “key polynomial” used in the density increment argument was: The key polynomial helped track density increments more
[ Q(x) = \sum_i<j (x_i - x_j)^2 ]
or more combinatorially: