Topic: Thinking about buying an April Cointerra unit for ~8 BTC? Take my bet instead. (Read 4861 times)

credit?  purchase orders?

Who would have the ability to buy a miner but not buy XBT

As a professional gambler, let me explain to you what risk is. Risk is when we take a pair of Queens against Ace-King all in preflop (54% favorite)

Gambling is when we have the worst of it, flush draw vs a set, not getting the rights odds to call, but say "ahhh fuck it" and stick in the chips.

What you are doing isnt taking a calculated risk with a positive expectation. What you are doing sir, is purely gambling, hoping to get lucky.

My bet is based on an expected return on mining. The exchange rate only really matters here in that it changes how much you pay me. E.g. say I expect the device to produce 2 BTC if so obviously I'm not going to offer it to you on terms where you only give me 2BTC.   I solved it for the 8 BTC number, I need to go generalize my software to sweep the exchange rates, to come up with the other parameters. Gimme a day.

Exactly. So is there an upper limit for the $/BTC exchange rate at which you would take the bet? Something like "$6000 equivalent in BTC up to $/BTC rate of $X"? Also, if you can codify what the changes in the terms would be if the $/BTC rate did go higher, that would give me a better sense of when I would want to trigger such a bet.

I'm willing to twiddle with the terms, though under alternative ones I'd probably want more than just the 8->5 changed. Keep in mind that changing exchange rates doesn't change the BTC I think it will produce much.

If BTC were to go to $10000 over-night and I thought it would stay there, I'd be buying April machines myself for 0.6BTC.

Interesting Bet. gmaxwell, are you willing to change the bet to be X BTC to be equivalent to the current price of the miner ($5999)? 8 BTC is currently around $6700, so the odds skew as BTC price goes up.

I'm not interested at 8 BTC, but if the $6000 equivalent goes to 5 BTC and the difficulty stays roughly the same, I might be willing to take that bet, for example.

Were your perpetual mining bonds without hardware doing the mining, sporting?

"I thought about doing this myself but making it very clear in the terms. If you you want to place a bet that difficulty will go way up its a great way to make that bet."

Was the lack of hardware made explicit enough that people knew they weren't backed by hardware?

"People who get caught doing bad things like to call me names. I think it helps repair their ego or something. Oh well, I can take the hit for the greater good of BTC land!"

I'm glad there's someone looking out for us.

"A"
(Actual former contents: Mining Bonds Offering)
"B"
"T"

You never forced anyone to buy your hardware-less mining bonds. I agree. But those people who were not informed (or misinformed) about the nature of a MINING bond—is it their own fault for not doing their homework and asking you outright?

gmaxwell makes an actual bet which will force people to actually think about mining returns, and calculate them, and you call it a scam?

Sorry for offtopic but.. Who really looks at these "trust ratings" here at Bitcointalk? It's poor system. A funny fact: A lot people who bought VIP status here - which also often comes with the ability to make peoples' trust show green/red with default trust list settings - have ended up being scammers. Last one was TradeFortress.

I proposed a fixed (and low) energy price denominated in Bitcoin in order to make it simple to handle. Though it's possible to use any rules which can be reliably decided by a human for such an agreement, its just better if they're as simple as possible.

The way this works is that the funds are locked up so that that any two of three people (Me, the other better, an arbitrator) must sign any transaction releasing the coins.  When its time to release we can negotiate among ourselves to decide the payout according to the agreement. If we cannot reach an agreement or one of us vanishes then the transaction is released with the help of the arbitrator who would read the agreement and the evidence and make a decision which at least one of the parties must also agree with.

The escrow is all done by the Bitcoin network itself.  The site is open source, but you don't need the site to spend the transactions— it's just a nice front end. If you have the three involved keys the transactions are deterministic and can be spent with bitcoin-qt alone (in fact, the site will even walk you through using Bitcoin-qt console to do the signing, if you hit the advanced button).

gmaxwell, I'm curious on how such a transaction works. Are there examples that are stored of more complex transactions like this? for example here you need certain inputs. something trivial as calculating energy prices, can be very tricky to do reliably, especially over long periods of time. say you have a script, and rely on a source for prices. now that site goes down or changes the API. then the inputs can be messed up.

It has to mine less than 7.619 for me to make a profit (due to the 5% premium).

Here is another totally unrelated way to think about the future difficulty:

At $750/BTC mining yields $2,700,000/day. At $0.06/KWh this supports a power usage of 1.875 gigawatts. At 1W/GH this suggests a power equilibrium hashrate of 1.875 exahash.

If you believe that Bitcoin supported 20.5TH/s in 09/2012 primarily on 2MH/j GPUs then you end up with a number like 320PH/s after compensating for the subsidy, exchange rate, and power efficiency differences.

In the past I would have said that we'd never actually reach the equilibrium hashrate— only approach some fraction of it, hashrate was falling when my GPUs were still 2:1 profitable over power— but the practice of enormous opaque pre-order queues and secretive mining operations suggests a possibility that we'll not only meet the equilibrium but actually overshoot it, at least temporarily.

And then there is the question of more advanced hardware then 1W/GH.

So, no I don't think assumptions about 400PH/s are unreasonable assuming Bitcoin survives and holds its value, I think such hashrates and higher are inevitable. The only question is when they'll be reached.

Thats certainly a factor which my bet reflects— I make the other party eat up to a full month delay (potentially two, relative to the earliest possible start date). But I also suggest running the numbers on other recently sold hardware. The recent rapid appreciation in Bitcoin value is concealing large losses for a lot of miners who would have been better off just sitting on coins.

For a 2 TH/s machine not to be able to mine 8 bitcoins from 4/2014 to 4/2015 would suggest you believe the network will average well over 400 PH/s during that period?   Is my math correct?    And the growth to that average would have to be in large steps because even if the network was 100 PH/s in the delivery month, the miner would produce 2 / 100,000 x 3600  every day (or actually every 144 blocks because with that growth you would be having adjustments every 9 days or so and more than 144 blocks a day) so you would have around 2.5 btc in the first month.
Use 50 PH/s for april as a starting point and the bet is very generous.   

I believe the way people actually lose the most money in mining is from assuming delivery dates to be honest.   Even KNC, when selling miners in the Nov 5-12 period for NOVEMBER delivery were telling people the machines would ship in late November.   Having them actually ship around December 10 was a huge difference in coins mined (still a good deal for purchasers but different from expectations).   And that is the BEST actor in the industry.

Well I would take you up on it except I'm all tapped out from buying said machines in fiat.  But anyways my orders are spread through multiple batches though I do believe my April units will ROI as well.  That said the risk is certainly very high that you may be correct but that's the risk we take.  I do think that miners nonstop plowing profits into more mining machines ends up being a colossal waste of money in the end.  Possibly after this generation when there are no longer huge leaps in ASIC performance to be had we will see the insane rush into mining slow down.

So, how much do you believe the lifetime income from the device will be (including disposal)?

I think buying an April Cointerra unit at 8 BTC would be unwise. If you think it would be wise to make such a purchase, that suggests that I believe the lifetime income will be substantially lower than you do. If so, we should be able to structure a contract which pays you better than buying the unit if you are right (and in the event that you would have lost money pays me). I'm willing to consider alternative terms, including increased upside liability on my part, though I may require some mixture of increase power offset, a reduced premium, higher initial price, and/or a initial offset (e.g. the returns start out slightly negative).

I think the point of the bet is noble but in vain.
I do not think rationality will come to the mining market until people stop pre-ordering (which is a ways off).   We almost saw it in November before btc started its run.   
Think about it.  When it started heading towards 200, KNC still had around 300 Jupiters (as it turned out, parts for Jupiters) that were not sold from their Nov. pre-order batch.   This likely means that all the HW vendors were not seeing robust orders because the size of the network was making people really think about whether a miner was worth it (50 btc for a 550 GH/s miner was getting hard to model).  Then second market / china hype saved every HW producer from actually having to act like a normal company and sell in-hand product.
I think when pre-order deliveries get us to a network of 40 PH/s (april?), that it will become difficult to get people to pre-order (but that assumes that price does not rise above 1000.

Goat negative rated me after his response. It's perhaps worth noting that he's currently trying to sell some January Cointerra units. Even at his prices I think they're probably a better deal than April units (and thus my bet).

What's up with gregorys trust rating?

How is this a better bet at all?  The hardware will continue to mine even after the first year and at 28nm it will remain above electricity costs for a very long time.  Even if you mine for one year the machine will not be worthless so this is a silly bet. 

Yes, my bet is far from worthless to the other side: I honestly believe it is a better deal than buying the hardware. It's just that I believe that buying the hardware is itself a poor deal at these prices. I've been telling people as much and I thought it would be appropriate for me to put my money where my mouth is.

Keep in mind Cointerra itself cannot deliver April without also delivering December, January, February, and March first! And while I provide some protection against Cointerra not delivering at all, in that case the calculation is relative as of the end of May.