No, the difference should be marginal. If it is not, there could be a large randomness in the strategy.

Only for interest: could you please explain a bit what has changed?
It might be that my script has a large randomness but I only use explicitely priceClose() and try to bring it together in mind with 'only high and low are stored'. Does this mean that the close is calculated somehow (e.g. as an average) of high and low?

Sort of. If EXTRADATA is set, the open or close price of a bar is the open or close price of its first or last tick. Otherwise it's the center price of its first or last tick.

Since the uncertainty range of a tick close price is bigger than the difference of center and close, a backtest without EXTRADATA is not less accurate. The remark under "News" is therefore a bit misleading. But you should set that flag when you use low resolution historical data, such as M1 data for one-minute bars, or D1 data for 1-day bars.

Thanks!

It is a very old thread, but I think some guys here seem to be able to use machine learning algos profitably but on the other hand do miss basic things about how the Z strategies work...

The Z strategies clearly state that there will be updates to the trained parameter sets in new zorro versions. This in fact means that the parametrization of the strategies are adapted to the current market behavior. Although WFO will basically help keeping the previous history record, it is not guaranteed to stay the same as WFOs are expanded at each training due to its' fixed segmentation logic. This is much more true for the latest WFO segment which provides parameter for actual trading.

As a logical consequence, the backtests will change naturally due to the parameters which are expanded in test time and optimized for the CURRENT markets and not the markets in 2015 or 2010. You may want to read about WFO, forward testing and market inefficiencies to understand the reasons.