Skip to content


In any sequence of races, we will have stronger feelings about some horses than others, and we will consider some outcomes more likely than others based on the odds. It makes no sense, then, to weight all of our selections equally given the unequal sentiment in which we hold them.

For example, we put a Pick 3 perm together as follows (horse numbers, and odds), ignoring the notion of ‘one brave race’ (see Bankers) for the purposes of this scenario:

Leg 1 – 1 (3/1), 3 (8/1), 4 (6/5)

Leg 2 – 2 (2/5), 7 (4/1)

Leg 3 – 4 (2/1), 5 (7/2)

The purest form of permutation would be 3 x 2 x 2 = 12 bets. You would be forgiven for thinking, “Well, it’s only twelve bets, at $5 a throw is $60”. But if you have a winning ticket with 6/5, 2/5 and 2/1 on it, you’ll do well to get your money back. And, according to the market, that’s the most likely outcome.

So how can we better utilize that $60 stake? By using a weighted perm structure, of course. The one I favour, though by no means the only one, is ABCX. Essentially, we separate all of the horses in our sequence of races into:

A – main fancies, strong chance
B – fair chance, quite like
C – dark horses, value at the price
X – not interested

Let’s apply that rationale to our initial bet, noting that all horses not selected have been discarded as ‘X’

Leg 1 – 1 (3/1 B), 3 (8/1 C), 4 (6/5 A)

Leg 2 – 2 (2/5 A), 7 (4/1 B)

Leg 3 – 4 (2/1 A), 5 (7/2 A)

Leg 1 has one each of A, B and C; Leg 2 has an A and a B; and Leg 3 has two A’s.

When using ABCX, we are saying that some combinations of outcomes are more likely than others. Specifically, it is most likely that we will win with all of our A selections; it is next most likely that we will win with two A choices and a B; and it is least likely that we will win with either an A and two B’s, or two A’s and a C.

Importantly, we are saying that if we were sufficiently wrong in our analysis that the result comes back with an A, a B and a C, we have to accept that we lose, despite selecting those horses in our initial wide sweep. This can be a difficult concept to grasp, especially for newcomers, but it is crucial to prevent players from throwing cash at the wall and hoping some of it sticks.

Remember, if we haven’t got at least one strong opinion, we shouldn’t be playing. So, if we snake our way out of a winner by optimizing our staking, we wipe our mouth and move on.

Let’s now look at how the original 12 x $5 bet might look using the above ABCX ticket distribution.

A much better 'weighted' spread of funds

A much better ‘weighted’ spread of funds

As you can see, we’ve spent the same $60 across five tickets.

Ticket #1 has a unit stake of $12 per line, and includes all of the A selections.
Tickets #2 and #3 have a unit stake of $6 each per line, and include two A’s and one B each.
Ticket #4 has a $3 unit stake, and includes the C pick from race 1, with the A’s in races 2 and 3
Ticket #5 has a $3 unit stake, and includes the B picks from both races 1 and 2 with the A in race 3

Now, we have an optimal combination of likelihood of success and multiples of the dividend. In other words, if the most likely horses win, we’ll have more units of the skinny divvy, and if some of the less likely horses win we’ll have less units of a bigger payoff.

The good news is, as you might have guessed from the image above, there’s a tool to help you with this.