Challenge 3: A Multiplication Table with Mixed References — Solution Walkthrough

Setup: B1:F1 = 1, 2, 3, 4, 5 (column headers). A2:A6 = 1, 2, 3, 4, 5
(row headers).

Formula typed once into B2:
  =$A2*B$1

Fill this one formula right across to F2, then select B2:F2 and fill
down to row 6. The completed grid correctly shows every product from
1x1 up to 5x5.

Explanation of each dollar sign:

$A2 (column locked on A, row left relative):
  As the formula is filled RIGHT (across columns), a plain "A2"
  would incorrectly shift to B2, C2, and so on — pulling from the
  wrong row-header column entirely. Locking the column with $A keeps
  every cell in that row of the table reading from column A's own
  row-header, no matter how far right the formula is copied.
  As the formula is filled DOWN, the row number 2 is NOT locked, so
  it correctly advances to 3, 4, 5, 6 — always picking up THAT row's
  own header value from column A.

B$1 (row locked on 1, column left relative):
  As the formula is filled DOWN (across rows), a plain "B1" would
  incorrectly shift to B2, B3, and so on — pulling from inside the
  data grid instead of the column-header row. Locking the row with
  $1 keeps every cell in that column of the table reading from row
  1's own column-header, no matter how far down the formula is
  copied.
  As the formula is filled RIGHT, the column letter B is NOT locked,
  so it correctly advances to C, D, E, F — always picking up THAT
  column's own header value from row 1.

Notes:
- Each reference is locked in exactly the ONE dimension the formula
  must never drift in for that reference, and left free in the OTHER
  dimension, where it genuinely needs to track the cell the formula
  currently occupies. Getting this table to fill correctly in BOTH
  directions from a single typed formula is the clearest possible
  demonstration of why mixed references exist at all.