So here's something I've been wondering. Say you

You ever see Kabuto fight Dark Kabuto?

>>14632745
That fight made the assumption that Put On would ever be a good idea in a fight. Masked was a completely useless form.

>>14632742
The robot with overall better abilities would win if all else was equal.

Otherwise, it's possible for one robot to simply out-think the other if its processor is better, or its victory could be due to small circumstances you wouldn't normally think of.

Power levels are as tricky as they are useless.

>>14632742

The problem is that the robots are limited by time, and can't realistically compute EVERY possible outcome of a given action within the limited time required to both complete that analysis and then enact a physical response to counter it. They can certainly make predictions, but they lack the time and data to be 100% accurate.

Also, their combat prediction is going to be limited by the constraints of their programming. Enemies that the programmer didn't know about won't be in the database, and combat prediction against them is going to drop off heavily because you won't know what their attack style is until after you have already seen them successfully kick you ass.

>>14632742
Too many variables, not enough processor cycles to compute every outcome and choose the right one.
>>14632976
Unless robot with less processing powers has algorithms optimized better than his opponents.
>>14633176
Self learning algorithms can remedy that can remedy that but again, time and repetitions are needed for proper responses to be known.


>>14632742
>TLDR:
Yes, but.
For now we lack sufficient technology to even make our robot walk and predict obstacles in reliable way. If allowed for enough time to run simulation of enough iterations robot can move in decent way but especially bipedal designs are still, well, funny to watch.

>>14632745
No