I don't get it. Why would you split? You choose the most profitable one and stick to it, until you find a more profitable coin, then you switch your entire mining power to it. Now, I remember during the time I was mining that the profitability of certain coins changed all the time, so I had to do a lot of switching but I never split my mining power. Maybe there are legitimate reasons to split but I never did so.
Depends on your situation, strategy, and cash flow I'd imagine. I can give you a few reasons I split my farm up.
1) I have several different rig configurations (RX rigs, Vega, GTX 60/70/80, etc), and they all perform a bit differently on different algos. So I try to keep them pointed where they perform the best/better, for the most part.
2) We dont take profits short term at all. Other than my time keeping everything running, we have zero overhead...and the time I spend on it I dont put a whole lot of value in. It's not that I dont value my time, I do...but it's mostly just time I would otherwise be spending playing video games or a round of golf or something, so it's a value add instead of a value negative. As such, our time horizon is medium to long term hold, so I'm not overly concerned about short term profitibility. Mainly we want our BTC/ETH bags to grow, and have zero time/desire to day/swing trade coins on exchanges. Essentially we are using hash power to dollar cost average in directly to a specific portfolio allocation. For the most part, we only use the coins to buy parts (build/upgrade rigs, replace dead fans/risers/etc).
3) Taking a few swings at newer algos when they first release, and the difficulty is low. Like I said in a previous post, we dont do a whole lot of this...mainly just let software auto switch on it's own. I'd guess, on average, about 20% of total hashing power spends about a week or so a month on something speculative. Now that I type this out and read what I'm saying, I probably need to do a better job of tracking this...
Anyway, I doubt what I'm doing is "right". I'm quite sure there are better ways to go about doing it....but considering the resources we have available, the professional & practical limitations we have as a group, and our longer term goal this is what is working. Currently. I reserve the right to change my mind.