Splits the job into one test (or one whole optimization) per symbol, as if every other symbol were absent from the data. Can be combined with OptByStrategy.
When parameters are being optimized (in Combinatorial, Random or Genetic mode), each parameter combination is generated once and run for every strategy and symbol before the next combination is generated. In Random and Genetic modes every strategy and symbol therefore evaluates the same set of combinations rather than independently drawn ones.
