Bitcoin mining - calculate difficulty rising

Let's try to calculate how much Bitcoin mining difficulty will rise over next future adjustments. To do it we need determine who the miners are and what is their behavior. All Bitcoin miners can be divided into three categories:

1. enthusiasts;
2. casual miners with gaming/office PC;
3. business rig owners;

Let's look on the behavior of members of this groups. Enthusiasts mine coins to support system and community or to trying to quickly get rich in Bitcoin gold rush. They do not invest much money into upgrade of their rigs. To return initial investments is quite enough for them. They may stop mining if electricity cost would be higher than mining earnings and return after difficulty drop. They mining hardware probably run in 24/7 mode.

Casual miners just experimenting with Bitcoin and do not target some goals in earnings from mining. They probably mine in working hours in the office or at home. Existence of such type of miners gives us intraday volatility of mining difficulty which can be approximated at 10-15% level.

Business rig owners trying to make Bitcoin mining the main source of their earnings. They trend to invest into upgrade and expansion of mining hardware to get more profit. This group gives us main piece of rising difficulty value. Let's calculate how much they can add to computational power of Bitcoin network each round.

The part with always-on mining hardware we may take as 80% of all Bitcoin network performance which for now estimated at 13.73 THash/s. As we saw all of them eventually would invest into hardware upgrade to increase hashing power. All Bitcoin network earn each round 50BTC*6per hour*24hours*14days between difficulty switch, which gives 100800BTC or 1310400 USD. So 1M USD will go to our serious mining group. All consumed power estimates now around 10MWatt for the network, so for this miners it costs 10MWh*24hours*14days*80%*0.1USD/KWh = 268800 UDS. Here we got left 731000USD to spend or invest for them. There is really different behavior in investing strategies among miners from 0% to 100% of profit. Lets say that they invest 50% all the time into upgrade, this is 365000USD. What you can buy for this money: most efficient cards with 1.5MHash/USD gives us 550Gh as evaluation for increase of performance each round. This is 5% of total Bitcoin network performance. So we can postulate that some rounds ahead we won't see increasing difficulty less that 5%, more likely around 10% because of newbie miners joined to Bitcoin.

This process may be interrupted by drop of BTC price. If price move down to 5USD/BTC the net profit for our 80% miners will be able to invest only 134400USD or 100Gh/s or less. This value will be less than 1% of network performance. Newbies at this price also wouldn't have reasons to join the network and difficulty will stabilize.