Hacker News Reader

Also the number of transactions per se is not the best measure of bitcoin's importance; it is the total purchasing power being transferred that should be considered. Bitcoin is a medium of savings that allows people to transfer value easily, from any point on the globe to any other point with minimal fees. The total purchasing power being transferred has growth tremendously because BOTH the number of transactions has grown AND the purchasing power of bitcoin has increased manyfold in the last few years.

This, of course, does not even factor in all the transactions that happen off blockchain.

I think looking at transactions is a reasonable metric. In fact, even this metric may be overstating things. Tim Swanson has analyzed transactions excluding popular addresses (apparently this excludes things like SatoshiDice) to identify transactions that are positive-sum in an economic sense and then the numbers look even lower. [1]

Another factor to consider is that if you count purchasing power transferred (which I assume means dollar value times transactions) you'll end up giving a lot of weight to events like when coinbase/mtgox moved their coins from one wallet to another.

[1] http://www.ofnumbers.com/2014/06/23/separating-activity-from...

I've already had this debate with Tim in private (we're friends). Some of his analysis of transactional use is quite misleading. For instance: http://letstalkbitcoin.com/blog/post/a-marginal-economy-vers...

Here he compares bitcoin to m-pesa which is a deeply misleading comparison. The former has a floating price, the latter is a money substitute and its price is fixed against the currency which is substitutes for. This is why I've suggested to Tim that it is total purchasing power transferred that should be measured, and I think I have convinced him of that.

From Jan'13 to now, which is when most of this merchant adoption has occurred, I can see an increase from around 40,000 transactions to about 60,000 transactions, which is rather modest and one can easily cherry pick points which show that the number of transactions has actually declined.

ed: And like I said in my other comment to vijay, it's not even clear that much of this represents positive-sum economic activity.

"40,000 transactions to about 60,000 transactions"

Exactly. You called this +50% growth "more or less constant", which is why I corrected you.

As to judging how much of this +50% growth is due to increased economic activity, it is impossible to prove due to the anonymous nature of Bitcoin. However indicators like the growth of Bitcoin payment processors (BitPay and Coinbase) very clearly show there is extreme growth. Eg. look at this data about BitPay: https://news.ycombinator.com/item?id=7974197

The Bitpay numbers are very impressive.

I do think that 50% growth in a year, especially when it is not that clear-cut that it really is 50%, isn't that impressive by itself. Perhaps there's a lot more growth in off-chain activity.

I remember reading an estimate that Bitcoin has < 500,000 users which seemed quite small to me given the hype surrounding it. So I'd be interested in seeing if there is growth in the number of wallets holding bitcoin. If bitcoin is taking off, we should be seeing that number explode, right?

That doesn't look like a growth story in the last year and half.

How is going from 40k transactions (Jan 2013) to 60k+ (Jul 2014) not growth? That's +50%!

Because almost all of that growth was reached by Mar 2013 and it's been mostly flat since then(one bubble, a few small dips).

No. For noisy data like this, you need to look at averages to make sense of it. Eg. see the quarterly averages, they show a growth from 45k to 65k transaction/day: http://i.imgur.com/XJs4V74.png (or see www.quandl.com/BCHAIN/NTRAN-Bitcoin-Number-of-Transactions and select "quarterly".)

Same reason why corporate financial reports are done on a quarterly basis. Or else there would be plenty of peaks and dips from week to week.

And if you scroll back to weekly you see any quarterly growth exists in one or two peaks pushing the quarterly numbers up then rapidly falling back down.

Realistically though we're both playing with figures to show the result we expect.

I've found plenty of merchants who accept it who have seen little to no revenue from it or shrinking revenues when they original saw some. Do you know of many merchants that are seeing increasing bitcoin customers and revenue?

Who do you think is more guilty of playing with figures: the one who looks at temporary peaks and dips, or the one who looks at quarterly averages to find the overall macro trend? ;)

I once compiled datapoints showing the growth of BitPay (ie. customers spending bitcoins on real goods and services from real merchants). You will find it interesting: https://news.ycombinator.com/item?id=7974197

One who uses short term spikes to make their long term data look better? We can(and have) gone back and forth on this a bunch.

That reply is actually to me in a previous discussion. As I said before Bitpay has known to do the exact same thing(use short term spikes) and attribute it to daily volume. They did it during the last holiday season along with Coinbase then both ended up deleting all the blog posts when called out on their figures not matching up.

So until I see a long term trend from them that is >$1m/day or I see merchants coming forward and saying they are seeing increased sales I'm going to take that figure with a grain of salt.

As for the rest. Half your stats there are increased employee and office count which come from them raising money.

The other half is an increase in merchant adoption which makes sense. Everything about bitcoins in legitimate transactions is beneficial to merchants by putting all the risk and fees on the consumer side.

"One who uses short term spikes"

Exactly. And you are doing it. You look at the weekly average (short term), while I look at the quarterly average (long term). Please understand that taking the average on longer periods of time will hide/smooth out short time spikes in a way that is statistical and purely objective (unlike your subjective interpretation of short term spikes).

In fact, the peak of number of transactions you said happened around March-April 2013 was precisely due to one of these short term spikes: a bubble where BTC went from ~$15 on January 1, 2013 to ~$260 on April 10, 2013: http://bitcoincharts.com/charts/bitstampUSD#rg730zczsg2012-1... (people were all buying/selling/trading BTC like crazy).

You are inferring too much from BitPay's post deletion. For all we know their numbers in the immediate 2-3 months following Bitcoin Black Friday 2013 were down (Black Friday sales were great, hard to sustain). Maybe they felt a bit disappointed by them, so they took them down. So what? Their overall growth is indisputable: "BitPay processed $5k/day in May 2012, $18k/day in Sep 2012, $160k/day in March 2013, $300k/day average through 2013, and now $1.0M/day." It does not matter if they are doing $500k/day or $1M/day or $2M/day. Growth is here, which is why they got bit VC funding.

If you want to get an objective value shouldn't you remove the outliers from the data?

>So what?

So what is not that they took them down because sales slumped they took them down because they were using them to justify the types of numbers that you quote immediately following that.

>BitPay processed $5k/day in May 2012, $18k/day in Sep 2012, $160k/day in March 2013, $300k/day average through 2013, and now $1.0M/day.

These numbers. It's easy for them to pick and choose spikes in the data and report that.

> It does not matter if they are doing $500k/day or $1M/day or $2M/day

Then why do you keep going on about them doing $1m/day if you're not convinced yourself?

>Growth is here, which is why they got bit VC funding.

Definitely. In merchant adoption. VC funding could just as easily be a speculative bet that that merchant adoption will eventually convert to sales not that it already has. Beenz and Flooz got VC as well remember(the first one almost as much as all bitcoin vc so far).

Anyhow. We're circling around around to the same points and using the data to suit our own needs and arguing over a figure that doesn't really mean anything(number of transactions) since it can easily be inflated and because we can both use it to show our own views. At least we're not discussing my wallet accounts or coinbase accounts.

I'll ask again. Do you know many merchants who are seeing strong and consistent sales in bitcoins? What are they? Do they specialize in bitcoin related products?

The growth can be seen even if you remove outlier months (especially the bubble months Mar-May 2013.)

Even if you assume BitPay is lying by reporting peak days, it would still likely imply growth: higher peaks likely correlate with higher average volume.

When I say it does not matter whether today they are at 500k or 1M or 2M, I mean it does not matter because they were at 160k in March 2013, so any number above that demonstrates a growth.

CheapAir.com and Overstock recently said they see consistent sales: http://www.coindesk.com/cheapair-tops-1-5-million-in-total-b... Also Amagi Metals has always reported very consistent and growing Bitcoin sales. These are a few that come to mind right now.

CheapAir.com is releasing their first figures so there is no way to say if they are consistent growing or shrinking other than him saying expedia hasn't lead to shrinking sales.

Overstock has definitely not said they have seen consistent sales. Their sales have been shrinking since they started.

Jan 10 - 130K[1]

Jan 11-29 - $26K/day average [1] (600,000-130,000)/18

Next 36 days to March 4th they do $400K [2] - $11K/day average

Next 83 days to May 27th they do $600K [3] - $7200/day average

-----

[1] http://onbitcoin.com/2014/01/31/bitcoin-executives-testify-i...

[2] http://www.coindesk.com/overstock-million-bitcoin-sales-futu...

[3] http://www.coindesk.com/overstock-ceo-patrick-byrne-1-6m-bit...

Amagi Metals reported very strong growth up last year. Literally reporting every milestone they hit($50K, $175k, $220k, $750k) then in April of last year they just stopped. Then a year later after someone else reports selling $10m worth of stuff they suddenly appear again saying "Oh ya we passed $10m already".

Please keep coming up with them though. It would be good to see some winners. The only one I've seen so far is a bar in Florida that hosts weekly bitcoin meetups.

For Overstock you are looking at a very short time period (5 months). Similarly, Google's revenue declined over a 6-month period from $16.9B in 2013Q4 to $15.4B in 2014Q1, so by your logic Google's revenue is shrinking too(!)

In reality sales are highly cyclical: you need to look at least at the year level to measure a trend. This is why I keep pointing you to payment processors BitPay and Coinbase, as they have existed for at least 1.5 years (and they publish numbers despite being private companies). Now that the first public companies are starting to accept Bitcoin (Expedia, Dell, Dish Network, Overstock), it is going to give us more data on Bitcoin's growth as they release public financial reports.

You say Overstock says they see consistent sales and I show you that their sales have been shrinking from day 1 and your response is "Well duh there isn't enough data!". Seriously?

Besides that you're comparing 2 quarters out of many with 2 quarters that are all the data we have. If those 2 quarters were their most recent then yes we could say their revenue is shrinking.

You keep pointing me to self reporting figures randomly released in press releases by companies that have been dishonest with them in the past. You keep doing that not because it makes sense but because it supports your beliefs.

Heres one for you. 5 years in there are only ~600,000 wallets with at least $50 worth of coin. Given how many people split up their coins between multiple wallets how many actual users do you think that leaves?

Anyhow I'm done with this discussion. It is clear nothing will convince you and you'd rather drag goal posts around a field than face the fact that maybe just maybe customer adoption isn't happening in Bitcoin. Good luck with your investment I hope it makes you rich because the alternative is going to be a lot of people in serious financial troubles.

Correct, I should not have given you the Overstock example, since there is not enough data for them.

"yes we could say [Google's] revenue is shrinking"

Here lies your lack of knowledge. No analyst ever said Google's revenues were shrinking when their 2014Q1 results were announced. Instead analysts said they "grew revenues 19 percent year over year." See http://www.theverge.com/2014/4/16/5621450/google-q1-2014-ear... . This is because, as I explained, it makes no sense to compare one quarter with the previous one as this is only 6 months of data and as sales are cyclical (typically a yearly cycle). So analysts compare a quarter with the same quarter from the previous year to cover a span of 15 months (a full yearly cycle).

"how many actual users do you think that leaves"

You make a common mistake: assuming that each user has at least 1 address. In fact, a LOT of users leave their coins on exchanges who tend to consolidate coins in a small number of Bitcoin addresses. Therefore you cannot estimate the number of users from the number of addresses with a non-zero balance.

According to that we're at the same level as we were in March of 2013 with a few dips and a bubble in the middle. Thats 17 months of more or less constant.