Please select the correct statements $($select three$)$ about why we DO NOT use the clock $($running time$)$ counter to evaruate algorithm's efficiency.

Because running time varies between algorithms and this is what we do not want.

Because running time varies between computers, we depend on computers, not the algorithms. Thus, we cannot measure the performance of the algorithms but the performance of the computers running the algorithms.

Because running time is not predictable based on small inputs.

Because we cannot express the relationships between input and the time $($even though time varies for different inputs$).$