At halving your "profit" doesn't get cut in half. Your Gross income (barring changes in fees) gets cut in half. The difficulty will drop some as negative/marginal players drop out - but ultimately only the leanest players will remain.
Which is why it don't make sense to buy miners before the halving. You are taking a big risk, since there is no solid information what will happen. Price drop, less diff. etc.
But when it comes down to mining in general I don't think it's profitable for the small miner in a year now. Just look at the stats:
https://bitcoinwisdom.com/bitcoin/difficultyToday: 194,254,820,283
1 year ago: 48,807,487,245
That's 4 times more!!!
Well a lot more on the plate at this ½ ing of btc then the one in 2012.
on dec 2012 we were close to 11m diff the entire network was 21or 22 th. almost all gpus some fpga's no asics.
a top card was 35 to 40 percent of a gh so 3 hd7970's were 1 gh to 1.3 gh this would mean 22,000 x 3 or 60,000-70,000
gpus were the whole network.
I now mine 28th which would have been 1.25 x the whole network at the last ½ ing.
we have more hear now mining eth coin I figure about 100,000 r9 390's is the eth network.
we have about 300,000 s-7's mining the btc network now.
A lot can happen different.
including a big retreat to eth coin.
look at the market cap
https://coinmarketcap.com/6.8 bill = bitcoin
1.1 billion = eth coin
I am juggling the two coins at the moment. basically while anything can happen
I fail to see a 3000 usd btc price in the next 7 months.
while I could see a 100 usd eth price in the next 7 months
note
3000/445 = 6.74 x 1 this would be future btc price vs current btc price jan 1 2017
100 /14.4 = 6.94 x 1 this would be future eth price vs current btc price jan 1 2017
I could see Eth coin doing that jump but not btc.