Yeah, good questions I think he comes closer to getting the offense right than the defense. I don't see how they can run the all same concepts in the running game without Lowery and Anderson (like pulling the left side) and the experience at RB especially is lacking. But I think with Rourke you have one of the best QBs in Ohio history. And the Offensive line has more questions than last year but a strong, experienced starting four of five with dudes like Plesants, Kitrell, Hayes, Meservy. And WR is poised to make a big jump with the experience from Odom, and the new, explosive guys like Buckner and Cox.

I think the defense has the talent to be significantly better than the number Connelly lists. Ohio has the potential to have the best secondary in the MAC as well as the best DE rotation. Javon Hagan will have +6 ints this year and is probably a day 2 NFL selection barring injury.

Ogun-Semore is poised for a breakout year- I cant remember seeing a guy his age with his motor. OS is relentless. And they have a MLB who is really starting to get it. I think that is enough to get things done. There are questions on the outside at linebacker and at DT especially. One SR we lost I would add to your list is Tyler Gullett-he came in during the MAC schedule and really helped solidify the LB corps.

My biggest concern on the team is RB or DT. Cole Baker can be a rotational DT on a MAC Championship team. As can Brian Arp. Either one of those two will have to elevate their game to be the impact guy. If not Arp or Baker, I think there might be an answer on the roster, depending how they approach the offseason. If I am the coaching staff, getting the best out of K Caesar is one of my top priorities. Hopefully he is hitting the offseason conditioning program like he intends to win starter's minutes. He has good quickness for his size and is a competitor but would be more effective by being lighter and more stronger?

Here is another piece of his analysis, the number he calls "returning production". Rather than just using the simpler "returning starters", he calculates how productive each player is, and what percentage return.
https://www.sbnation.com/college-football/2019/1/31/18204...

TO see if I understand his method, I compared his final numbers for 2018:
https://www.footballoutsiders.com/stats/ncaa
To the preseason numbers at the link in the first post on this thread. The following chart shows the Production returning for offense, and the change he projects from end 2018 to start 2019

Offensive production returning:
Kent State: return 86%, change from 2018 to 2019 +4.1
WMU 84% +3.4
CMU 83% +4.7
BG 71% +2.8
Akron 67% +2.5
Ball St 66% +1.0
NIU 64% -.2
Miami (Oh) 57% -1.1
EMU 53% -2.3
Toledo 50% -1.4
Ohio 41% -4.5
Buffalo 40% -4.6

There is clearly correlation, but he is including some other unknown factors as well. For example, why is Toledo's offense not projected to fall as much as EMU, even though EMU has more returning production? Why is CMU projected to improve the most, even though Kent and WMU have more returning production?

Now the same thing for Defense:
WMU 85% +3.7
Ball St 84% +3.8
NIU 75% +1.3
Kent St 73% +2.0
Ohio 63% -.6
Toledo 57% -.6
Miami (Oh) 57% -1.6
EMU 49% -3.6
Buffalo 46% -3.8
Akron 44% -3.3
BG 42% -3.0
CMU 40% -3.5

Again, there is correlation, but some numbers that obviously include other factors. Perhaps NIU didn't improve much because they were already really good? Why did Akron, BG, and CMU not drop as much as EMU and Buffalo? I have no idea.