Quote from: ToxiClay on January 01, 2014, 04:31:54 AM

get the best ROI?

Would you count the least negative of several negative options as the best?

The "best" ROI may, sadly, simply be not to buy just sit on your coins. Thats always an option.

I enjoy mining and hate to discourage anyone to get involved, I'd even mine with a substantial probability of only breaking even... But the current prices are crazy, and I think the hardware makers are exploiting ignorance about how mining income changes over time in order to extract more from customers than the hardware can ever produce. Substantial losses at their current prices seems likely, and this is assuming that their promises are all kept and deliveries are on time.

This depresses me: Wide distribution of hashing power is an essential element to the Bitcoin security assumptions. The COGS on these mining devices are quite low once produced in quantity there is more than enough profit to go around. But this is what we appear to have. There is a bunch of competition selling hardware, but apparently not enough.

Quote

Looking at the Bitcoin Profit Calculator, even if the difficulty is double what it is currently, the heavy ASICS (500GH/s and up) are still profitable enough. So why the hate for the heavy miners,

"even if the difficulty is double" Please pardon me while I titter and direct your attention to the first graph. Red is difficulty. The other lines are hashrate estimates difficulty scales linearly with hashrate every 2016 blocks.

As you can see in the second chart the hashrate (thus difficulty) has been increasing at 2% per day. This means that it doubles in 35 days, and then doubles again and again. This behavior can't continue forever, but it has been remarkable stable for a while and appears that it will continue for at least the immediate future as more and more huge operations come online.

For an example of how this plays out on 2013-08-15 the network hashrate was 392.05 TH/s. Today it is 10392.91 TH/s. This increase of 26.5x corresponds to a compounded increase of 2.368%/day (so in case you were wondering in other threads why miners were willing to pay 55 BTC for 400-2000 GH/s miners to be delivered in October and are so angry that haven't shown up...).

If you're willing to assume the 2% model which as I said, is clearly wrong in the sufficiently-long-term but is probably a reasonable conservative assumption in the immediate future you can figure out the lifetime income from a device with a formula like:

0.014675 BTC * GH/s * 0.98 ^ D = Device lifetime income

where D is the number of days in 2014 until you receive the hardware. This has some minor accounting for power costs at the 1W/GH level baked in, but really the 2% assumption dominates. Under this formula you expect 2TH/s delivered 60 days into 2014 to have a lifetime income of 8.73 BTC. While the 2% assumption may turn out to be wrong hashrate might grow faster or slower in the next few months it's certainly a helluva lot better than just assuming a constant difficulty.

As far as "heavy miners" really thats irrelevant. $/GH, arrival date, and W/GH are the only factors that matter much, and at the moment mostly the first two until the crazy growth stops. It easily can be more profitable to purchase more smaller miners if they're less expensive and/or more power efficient. When thinking about profit you need to think in relative terms.