[Crosby]

I think this is a solid play, though I prefer to have a lot of datamined hands or personal experience with them. The thing is that, yes, if they are a nit (let's say they 3bet 5.2% of hands and call 2.6%) then you lose $80 half the time and win $48 half the time so you lose $32/hand. However, we know he's a 25/20, so although we don't know anything specific about his 3betting range or 4betAI calling range, I think it's safe to assume this is impossible.

If he is loose-aggro but too weak, or to put it another way, he thinks you open the button very light but call very tight, then as others said, this is definitely going to make money. If he 3bets 12% and calls 3%, you make $16/hand. If he starts really trying to exploit you but doesn't open up his calling range, then this could be as good as the worst case scenario is bad.

If he opens up and starts calling our 4bet with even as much as say 6.8% of hands (88+AQo+some suited broadway), then we only lose $52 each time we're called, so although he could make us bleed money by opening a little more than double that and down from there, it's relatively negligible.

I would put huge emphasis on your own opening range from the button in this spot, which is info you didn't happen to include. If you open very wide and we know the button isn't a nit, it seems like plays like this are mandatory because the BB will have to 3bet light enough to make this between marginally unprofitable and very profitable. I believe the result will be on the very profitable side more often because this sort of thing is very counter intuitive to a lot of people and it seems very appropriate for the stakes that the average to somewhat above average player will be somewhat aggressive but not particularly strong.