I haven't looked at the splits or success rate for the two RB on the Bengals last year. I was just speculating that these are contributing factors and hinting at those are things someone could look up and share, as I didn't have time to at that moment.
Here is the success rate. 41% for Mixon 40% for Bernard. Pretty even. I was actually expecting Bernard to be better than Mixon here. He wasn't.
Bernard had 15% of his runs on 3rd or 4th down. 68.6% of his runs were when the Bengals needed 10 or more yards for a 1st down. He averaged 5.1 ypc on first and 10 to go and 60.1% of his total yardage was gained on 1st and 10 situations. 12.5% of his yards were gained on 3rd downs. Bernard struggled on the 12 carries he had when the Bengals needed 3 yards or less to get a 1st down only having 1.8 ypc on these runs.
Mixon had 3% of his runs on 3rd or 4th down. 70.2% of his runs were when the Bengals needed 10 or more yards for a 1st down. He averaged 3.3 ypc on first and 10 to go and 52.4% of his total yardage was gained on 1st and 10 situations. 5.5% of his yards were gained on 3rd downs. Mixon struggled on the 20 carries he had when the Bengals needed 3 yards or less to get a 1st down only having 2.6 ypc on these runs.
Somewhat strange that Mixon had his best ypc on downs where the Bengals needed 4-6 yards for a first down. He was below average otherwise.
All of that adds up to the not very good success rate from FBO. Which may point more towards the offensive line than anything else. Bernard had 49% success rate in 2015 which is pretty good.
The splits actually show that Bernard was a more effective RB on 1st and 10 situations than Mixon or on just 1st downs in general. 5.1 ypc on 55 1st down runs for Bernard 3,3 ypc for Mixon on his 107 attempts. It seems like the Bengals would have been better off having Bernard handle 1st downs and get the volume Mixon had instead of the other way around.
I agree with you that Bernard can run inside or outside. In my view he is a complete RB. However Marvin Lewis does not seem to see it this way as evidenced by the picks of Hill and MIxon despite already having Bernard.
At least Hill had a better success rate than Mixon did, which again may point towards the offensive line being better when Hill was playing for them.