I sometimes, but not always, get Error 711, telling me that the target can not be found in the result list. However, see the screenshot where CAR25/DD25 is circled in red, because it is definitely there.
At least twice, simply closing AB and restarting it solved the issue. For example, after closing and re-opening AB, I did the exact same test on the same data, and AB didn't give that error.
However, the walk forward results were corrupted. The variable being optimized, MaxPosValPct, is not properly populated in this screenshot. Some of the values are appearing one column to the left, in the 'Probability of breakeven' column:
Am I missing something here? Any advice on what to look for next?
These columns show just the WINNERS for in-sample period. They don't represent ALL in-sample steps. You have to watch out for ALL parameter combinations for ALL in-sample steps. They MUST produce non-null values for custom metric if such metric is used as optimization target. In other words, in your case CAR25/DD95 MUST NOT be Null no matter what parameter combination is used.