I know Eno Sarris has said he's tried, and he wrote an article on it a year or two ago. I think there are really two big issues: 1. Much, if not most, of their work is not in game strategy or tactics, but in leading a group of people day by day. That's hard to measure, especially if you don't have day to day access to the team, and you don't know which processes he controls vs the FO. Also, even if a team had some kind of internal measures of employee satisfaction or something, we'd never know about them. Culture and that stuff is hard to measure, and I'd bet MLB clubs aren't exactly at the fore front of that measurement. 2. Much of the success is really player driven. Even when you make the right decisions, it often doesn't work. Because the sample sizes are small for tactics, it's hard to measure. Take bullpen usage. I know Sarris tried to figure out if there was an issue with having RPs warm up, but not be used, or warm up in 2 innings. But, that data isn't kept anyplace, but it could have a big impact on that RP's performance. Or, take lineup construction. Studies show it might matter 1-2 games a year, maybe 1-2 more at the extremes. Since most of the time a manager will get things right, it won't matter much. That's really the issue, I think. There isn't much difference in how managers really act, so there isn't much data that shows anything other than the normal way to work. And, when a manager does something that isn't normal, it is only done a few times a year probably, so there isn't enough outcome differential for it to matter when compared to other managers. What's interesting about that, is that manager decisions, over the year, probably wash out a lot. But, in the context of the moment, they probably matter more than we know. People also tend to recall when things go horribly wrong when a move is made, but rarely remember when they go right (as expected most of the time). Yes, Molitor's love for the sacrifice bunt is painful. And yes, it might have cost them a run here and there. But, when you average it all out over the year, it comes out in the wash. That's why I actually like measurements that try not to take averages over the year, but look at discrete outcomes in a game. But even those are tiny differences usually. And even those wash out when you add them up (some negative, some positive). those are my theories, anyway.