I also applied the same rule (last start the selection ran fifth or worse) for Unitab's top rater - the result was a 2% profit for Win betting.

However, I presume this is for every 100-pointer on Unitab, so there could be multiple selections in some races.

Interestingly, last start position between 1 and 4 has produced a 9% loss on Unitab.

If both the D.Scott and Unitab results are improved by betting their top selections that finished fifth or worse at last start, then one can easily jump to some contraversial assumptions.