Race 2: A--4,6 B--3,5
Race 3: A--4,7 B--2,3
Race 4: A--6,8 C--2,9
Race 5: A--1, B--7,8 C--2,6
Using every horse on one ticket (4x4x4x5) would have cost $640 for $2 or $320 for $1. Instead, to have some combos more heavily for around the same total as the $1 investment, I played as follows:
A/A/A/A+B -- 24 combos at $4 each = $96
A+B/A+B/A/A+B -- 96 combos at $2 each = $192
A/A/C/A -- 8 combos at $1 each =$8
B/A/C/A -- 8 combos at $1 each =$8
A/B/C/A -- 8 combos at $1 each = $8
A/A/C/B -- 16 combos at $1 each = $16
A/A/A/C -- 16 combos at $1 each = $16
B/A/A/C -- 16 combos at $1 each = $16
A/B/A/C -- 16 combos at $1 each = $16
Total = $376
(I hope the above adds up. It's late and I've revised it twice already to fix errors.)
The sick thing about how it worked out was that my three winning A's paid $23.40, $28.00 and $11.40, while my B was a $4.80 favorite about whom I clearly had a bad opinion. [Update: As a commenter points out, I actually had two A's at $23.40 and $28.00 and two B's at $4.80 and $11.40. My bad.] and two B's
And when I went into the last leg alive to five of the eight horses -- the 1,7, and 8 for $2 and the 2 and 6 for $1 -- I wasn't too proud to hedge it out, and I bet the three uncovered horses to win at 14-1, 32-1 and 43-1. Whether these savers are profitable in the long term I can not say, but in the short term they pay for themselves by protecting your sanity and equilibrium.