Breaking rules on first deal

[Jeff Holmberg]

After years of reading books on real estate (most recently by David Greene and Brandon Turner), I'm finally getting started!  Unfortunately, I think I need to break some cardinal rules in order to get into the game. But before I do, I'm hoping to explain the situation and see if more seasoned investors here think I'm being wise or foolish.

I live in the upper Midwest, and I feel like there's a lot of potential to buy properties that fit my criteria in Minneapolis and Des Moines. The most relevant criteria for this post include: bad homes in good neighborhoods, high comps compared to my property, and "all in" cost at <80% of ARV. However, Minneapolis and Des Moines are highly priced compared to surrounding small towns (as much as 10x more expensive). Since I'm sure I'll make a lot of mistakes in my first couple properties, I've decided I'll feel more comfortable starting out my real estate journey in smaller towns (around 25,000 people) with low-cost property, in the $25k to $75k range. After a few years of growing my experience and my capital, I'll feel more comfortable entering larger markets and taking on bigger risks.

In the meantime: I'm finding that these smaller towns don't have rental property that meet my criteria.  Instead of buying a bad house in a good neighborhood, I'd be buying (or rehabbing) a good house in a bad neighborhood.  Not bad like crime-ridden; just bad like: low prices, dilapidated homes, near business districts, etc.  Given the high demand / low supply, sellers of rental property are asking for a lot more than surrounding homes sell for. 

Quick example: I could buy a duplex for $70k with enough cash flow to make me feel good about that price. But the appraisal might come in around $30k-$40k based on surrounding homes, so I'd seek seller financing for the difference. I would likely treat the duplex as a long-term investment, focus on the cash flow, and disregard the fact that it's supposedly worth less than what I paid for it. If my models hold true, my long-term ROI could still be great at around 12%/year. But even if my models are wrong and I end up with lower returns (or even a loss), then at least I'm getting experience with low risk like I wanted.

Am I being penny wise and pound foolish?  Or am I being wise by starting with very little weight on the bar, and only adding weight as I get stronger?