all-in is not used so often that it can be said to produce serious statistical results. Again - often all-in turns out to be just a bluff or erroneous expectations. Therefore, the success rate of such an action is 10-15%, no more than that