I mostly bet proportionally around a set flat stake. The amount I bet depends on the overlay value of the bet. The lower the value, the lower the bet and vis-a-vis. This type of proportional betting particular suits punters who handicap their races [do the form].

I work out the overlay value by using my own priceline for what I believe are the winning chances in a race. Pricing is dependent on the number of winning chances and the field size of the race. I'll generally ignore races with more than 14 runners or races with too many winning chances [more than 4], except for the odd big Group or Cup race.

A punter needn't price every horse in a race, just the possible winning chances and after a bit of practice it's easy enough to to do in your head. It's more art than science.

Three good winning chances in race of 12 starters? OK, to work out an initial basic SP price, that's a 25% winning chance for each horse with a 25% chance for the rest of the field and the unexpected.
So my winning contenders have an initial 1 chance in 4 each or an SP of $5 each. Fine tuning kicks in now and from my [further] handicapping efforts I'll work out a final SP price for each winning chance. Which horse I back will depend on the prices offered [value], not just backing which of the three I think will win, because we are so often wrong [I'm chasing value, not what I think are 'winners'].

There is no rocket science involved here [just practice], because regardless of method, everyone ends up with an educated guess regarding a fair SP.
Among my winning chances I will often back a runner because of a large overlay above the SP price I've set for it, even though I think one of my other winning chances has a slightly better chance of winning.

EG: I backed Jokers Wild who came 3rd in the Cadbury Guineas because I picked it up at 27/1 early on Betfair and only just missed on a very large collect. It lost but it was still a very good bet because it almost pinched the race. Luckily I also backed JW for the place, which payed more than the winner Miss Finland [great win!] anyway.

Yesterday [Mond. 12th] I picked up Pinsemont (NZ) at the Valley, a good back-marker in an outside alley at 14/1. A very good chance against Fearless Waters, who at 4/1 was poorish value. Pinsemont ran over the top of them and I'd put twice as much on it as I would have on Fearless Waters [ 2nd]. Pinsemont was way over it's true odds as Back-markers in outside alleys often are, BECAUSE they are in outside alleys and most punters wouldn't know a back-marker from a on-pacer.

I used the same betting approach on Jillimarg in R4 at Morphetville. At 12/1 it was way overs, but it was just beaten by the favorite Maloose [3/1]. Win some, lose some.

Cheers.