Best TruStories of the Week - #6


Had only thought about the idea of selfish-mining in the context of double-spends, so I searched for the occurance of these in practice on the bitcoin network and found:

… the most interesting fact is that these zero-fee tx’s inbetween winning ones were mined by exclusively. Possibly this was a test attack.

This attack appears to have involved selfish mining, might be worth looking into as a case study

In the first scenario where the honest nodes succeed in finding a block on the public branch, nullifying the selfish pool’s lead, the pool immediately publishes its private branch (of length 1).

The quote above is from the linked paper and makes me wonder about whether this could be achieved in stealth, reading through the conversations from the 2013 suspected double-spend, it appears that people will notice a change in the hash rate from a pool as it devotes it’s resources to a private chain.


The block withholding attack (ie selfish mining) is a bit different than the double-spend attack you referenced. Because in selfish mining, there doesn’t have to be a double spend in the blocks that the selfish miners are withholding. The private chain that is being withheld from the public chain is a completely valid chain (unlike the double-spend attack where one chain is valid and other isn’t).


Claim: 50 Cent (Curtis Jackson) earns 700 BTC from sales of his 2014 album Animal Ambition

Category: Use Case


Evidence/Argument: Screenshot of Curtis Jackson’s BitPay account, forwarded by him, in the following TechCrunch article, which estimates that the earnings from the album were 6 - 7 Bitcoin:

Status: Rejected

Though Curtis Jackson had some BTC earnings from his album Animal Ambition, it was not 700 BTC.

Why I found this claim interesting: There was a time when people wouldn’t accept Bitcoin as a legitimate currency in exchange services/products. I’m African-American. Usually, when I speak with a lot of my peers in my ethnic group about Bitcoin, they often gave me a blank stare or claimed it’s a scam. The publicity that 50 Cent generated with the Bitcoin claim created an avenue to discuss Bitcoin with people in a way I was unable to do prior, even my Dad :slight_smile:

Astra Rai


hey @FatTony, yup, the dataset I downloaded from goes from January 3rd 2009 - December 22nd 2018. I think they make data available for every other day to shorten the number of rows. Even then, there is 1821 data points spanning nearly 9 years.

If you select “All Time” then you get every other day; but if you select a shorter period, 6 months, 1 year or 2 years, then data is available for everyday.

In any case, my primary source was from and definitely a similar analysis using Coinmetrics data would be a very welcomed.


Claim: Blocktower Capital to Lose $1 Million After Betting Bitcoin Would Hit $50,000

Category: Crypto Funds & Fundraising

Source: CryptoSlate

Evidence/Argument: This article implies that the options purchased by Blocktower were a one sided strategy and that the full amount of $1 million was lost because, as of the expiration date today, the price of Bitcoin is not above $50,000. In a recent article by Bloomberg, this ‘bet’ was discussed and while the amount was not disclosed, they did take profit off of the original transaction. Paul tried to explain this on Twitter to those claiming he took a 99.73% loss. As the Bloomberg article states, “Ari Paul, a cryptocurrency fund manager at BlockTower Capital, has indicated that he bought the options while simultaneously selling some of his fund’s Bitcoin holdings. In a December 2017 with CNBC, Paul said the trade allowed him to lock-in some profit, reduce exposure to market declines and potentially earn a big payout if Bitcoin soared above $50,000.”

This article leaves out any mention of the profits or benefits to the fund of this trade, instead stating that the fund will take the full $1 million loss, which is false.

Status: Rejected


The critique of the selfish-mining paper illuminates a fundamental flaw, the paper makes economic assumptions which simply don’t hold in reality.


Explanation of the game theoretic model used to analyse bitcoin

Economic critique medium post

Economic critique paper


You can refute my original claim instead of creating a new one :slightly_smiling_face:


Thanks @preethi, edited my post.

Craig White’s arguments demonstrating flaws in the “Majority is not enough” paper can be summarised as follows:

Selfish mining is more costly than purchasing the extra mining power

Highly recommend you read through this yourself (and reply back with your own critique), but this diagram provides an illustration of his cost-benefit model output.


Honest miners can and would easily detect selfish mining and block the selfish miners from contributing to the network

The large miners all know each other on the network and already use a whitelist to determine which blocks to prioritise in the network. Selfish mining could easily be detected by honest miners and blacklist the node. Even in the case of anonymised IP, the attacker will not be on the whitelist and therefore not have the speed advantage required to execute this attack.

Is the bitcoin protocol still not incentive-compatible in face of these arguments? They clearly outline why a rational miner would not be incentivised to execute a selfish-mining attack.



Mahindra to Leverage Blockchain for Grape Export Business

Category: Use Cases


Evidence: Statement by the MD of the Mahindra group of companies as mentioned here -

Status: Confirmed


Claim : "Less than 1% of the world’s population — no more than 40 million people — have ever used Bitcoin. But, according to the Human Rights Foundation, more than 50% of the world’s population lives under an authoritarian regime.”

Category : Use cases

Source :

Evidence/Argument : Got curious about how much of the world’s population lives under an authoritarian regime and how is that measured. I wasn’t able to find anything from Human Rights Foundation website but found that Democracy Index by the Economist Intelligence Unit aims to measure that.

According to that source in 2017, of world’s population, 34% was living in authoritarian regimes, 16,7% in hybrid regimes, 44,8% in flawed democracies and 4,5% in full democracies. It looks like more than 50% might be living in authoritarian regimes if hybrid regimes can also be considered at least partly as authoritarian regimes, which might be the case as they are not democracies. Source and the white paper with more information:

In general, an interesting data set and a reminder of how things are in the world.

Status : Unconfirmed


Claim: India looking to legalize crypto

Category: Regulation

Source: Pomp’s “this week in crypto” tweet which has drawn 1k RTs at the time of this writing.


While the phrase “looking to” is vague enough to include any possible explorations, one obvious way to interpret this is that overall government sentiment is positive and/or they’re moving faster than before on this topic. Neither is accurate as several India-based responses immediately point out:

: 53%20AM



Let’s dig in ourselves. I’m assuming Pomp’s source for this tweet was a 12/26 Coindesk article which only references this Indian publication. That piece has a single positive comment from an anonymous official :

We have already had two meetings. There is a general consensus that cryptocurrency cannot be dismissed as completely illegal. It needs to be legalised with strong riders. Deliberations are on. We will have more clarity soon. We have also taken inputs from cryptocurrency exchanges and experts and will be examining legal issues with the law ministry. It’s a complicated issue. Once all aspects are decided, then we will have more clarity.

The New Indian Express journalist interprets the comment in this manner: “Holding cryptocurrency like Bitcoin is banned in India, but that could change, as the second interdisciplinary committee is in favour of legalising it, with harsh riders.”

I can’t lend credibility to that interpretation when the first part of the sentence is already false. Holding cryptocurrency is not and was never banned in India. It’s just not recognized as legal tender (and neither is gold…). Here are the only definitive comments the Indian government has made on this topic:

The government does not recognise cryptocurrencies as legal tender or coin and will take all measures to eliminate the use of these crypto-assets in financing illegitimate activities or as part of the payments system - Arjun Jaitley, Finance Minister (source)

There’s a lot of false information on whether crypto is banned in India (it’s not!) and its implications. This tweet and article does a good job of articulating what’s true and blatantly false. I’d recommend reading that if you’re unfamiliar about this topic.

So what’s actually happening now?

The IMC committee is going to have another meeting and draft a report to the Finance Ministry early next year with recommendations on “digital currencies and the use of the blockchain technology in the financial systems in India”. (source)

What is the IMC committee?

It’s led by Subhash Chandra Garg, the Department of Economic Affairs secretary. Members include Ajay Tyagi, chairman of India’s market regulator, the Securities and Exchange Board of India, and BP Kanungo, deputy governor of the Reserve Bank of India, and representatives from the ministry of Electronics and Information Technology). Source

What is the next step / timeline?

They plan to submit a report to the Finance Ministry by February next month. It’s important to note that this is just a soft goal. Progress on regulating crypto in India has been historically delayed multiple times.

Status: Flagged for misleading.

The more I dig into this, the more it feels like the “New Indian Express” journalist spun a somewhat positive quote into a misleading headline (“Bitcoin on highway to being legit” - really?) that ultimately spiraled into tweets like Pomp’s. I personally don’t think any meaningful progress has been made until we actually see this oft-mentioned “draft bill” and hear the reactions to it from those in decision-making power. Meetings don’t mean anything until we either have public records of the meeting or have on-the-record comments from officials.

Best TruStories of the Week - #8

Claim: Augur just resolved a market incorrectly

Category: Project (Augur)


This Reddit thread is gaining traction on Twitter, example:


Here is the author’s entire argument:

The market: How many Hurricanes will Strike the U.S. in the 2018 season?

The resolution: 2

Additional Details: "How many Hurricanes shall make landfall (with center of eye crossing shoreline) as Category 1 or higher, anywhere in the 50 U.S. states or DC, excluding other U.S. lands. Shall the same storm make landfall more than once, that still only counts as one. "

The problem: A hurricane hit Hawaii. The market details clearly specifies all 50 states and not just the Atlantic states. The correct resolution should be (at least) 3.

Hurricane Florence: Landfall in North Carolina

Hurricane Michael: Landfall in the Florida pan-handle

Hurricane Olivia: Landfall on Maui, Hawaii

Augur link

But, no hurricanes hit landfall as this Hawaii News article notes:

Hawaii’s 2018 hurricane season was among the most destructive in years. That’s even though a hurricane never actually made landfall in the islands.

Olivia was a hurricane at some point but by the time it hit landfall in Maui, it had weakened into a tropical storm per Wikipedia:

Olivia weakened into a tropical storm on September 11, before making brief landfalls in northwest Maui and Lanai on the next day, becoming the first tropical cyclone to impact the islands in recorded history.

A tropical storm has winds below 74mph (source), a category 1 hurricane is 74 - 95mph (source). Hurricane Olivia had winds at 70mph on September 11th (source), two days before landfall, where it was still a tropical storm (source).

Status: Rejected
Only two Hurricanes hit landfall in the United States in 2018. Olivia was a tropical storm when it hit Maui. The Augur definition holds.



BTC Lightning Network Company OpenNode Rejects $1.25 Million from Roger Ver to Build on Bitcoin Cash vs. Bitcoin

Category : Projects
Source :

Evidence :

Proof of the actual offer:

  • Roger Ver announces it at 20:43 in the following video.

Status : Confirmed


Similar to the issues with selfish mining, I came across some issues with “Ghosting”, a potentially new mining attack for Proof-of-Work blockchains that I hypothesized during a TruStory [discussion] ( I am bringing this attack vector hypothesis to the public’s attention to spark further discussion, as I do not know of its viability in practice.

“Ghosting” works as follows:

A large nefarious mining pool decides to turn off and on their hash power every 2016 blocks in order to take advantage of the changing difficulty setting. For the 2016 blocks that have a lower difficulty, the nefarious miner adds back their hash power, thus achieving a higher efficiency (block reward/hour) than the mining pool would achieve had they been honest miners and mined every 2016 blocks. The mining pool then removes their hash power from the network at the end of the 2016th block, right before the difficulty is increased, only to add back their hash power right after difficulty decreases again 2016 blocks later.

The most significant reason why this attack might not be viable is that the gained income overtime from this strategy might not offset the foregone income had the mining pool just been honest. That being said, many other mining attacks have shown to be not cost effective as well (51% attack on large Proof-of-Work chains, selfish mining, etc.)

Two professors, Cyril Grunspan and Ricardo Perez, pointed out the lacking cost effectiveness of “ghosting”. I cannot completely follow all of their math, so if someone can help translate that would be much appreciated.

Here are their comments:

"You can follow the procedure that we present in our paper to evaluate the ‘revenue ratio’ of your strategy. The attack cycle consists in 2 consecutive difficulty periods. If q is the relative hashrate, you have

E[R]=0+qb * 2016=qb * 2016 (b is the block reward)


E[tau]= 2016 * (tau_0 * (1-q) + tau_0 /(1-q)) > 2* 2016 * tau_0 (where tau_0 = 10 min and we use x+1/x >2 for x>0)

So you get (if I am not mistaken)

Gamma (Strategy) = E[R]/E[tau] < (qb/tau_0) /2 < qb/tau_0 = Gamma (Honest)

Thus the revenue ratio is not more than half of the honest revenue ratio."

Furthermore, there are other costs/savings that aren’t accounted for in the abstract model, such as electricity savings and depreciation savings from not using the ASICs every other 2016 blocks. Those savings, depending on the time horizon in which this attack would be executed could dramatically affect the profitability (or lack thereof) of this attack.


This particular article on medium caught my interest:

Written by self-described “bitcoin maximalist” StopAndDecrypt, it criticizes some fundamental parts of Ethereum and future scaling solutions. It got some traction and even a response from Vitalik, which lead to follow-up articles from the author as well as other people.

I’m going to try and dive deeper into the arguments he makes and also look into the counter-arguments. My first TruStory contributions will therefore be based on these series of articles.

I will start with this one:

Ethereum’s blocksize is increasing exponentially.

Projects - Ethereum

See article above. Direct quote:

“Because of Ethereum’s exponentially growing blocksize, the bottleneck is not regulated below these external factors and as such results in a shrinking and more centralized network due to network demands that increasingly exceed the average users hardware and bandwidth.”

It is important to note that the claim is about the current trend. While we cannot know if the trend will change in the future, we can look at historical data and see if it is reasonably exponential at this time and in the past.

Here is a plot from Etherscan:

This chart displays avarage block sizes for each day. While it has increased a lot in the past, it has gone down and been on avarage 18k bytes since October 2018. This does not resemble exponential growth.



As a sidenote on my post above, the author also states that Ethereum has an unlimited blocksize. This is not entirely true, as there is a gas limit in each block. However, the blocksize can change over time because miners have the option to increase or lower the blocksize by a certain amount compared to the previous block. So, in theory the blocksize could go to infinity. It is however difficult to say what will actually happen.

One thought that occurred to me is that, miners with above-avarage hardware might be incentivized to increase the block size because they want to increase the network transaction throughput. Since the miners with stronger hardware (i.e more hashing power) get to mine blocks more frequently, could that imply a tendency to increase the block size over time?


I keep seeing this headline pop up and every time it references this Bloomberg article.

I did find one medium post that dug a little deeper:
The author even read through the recent job posting for Facebook to find the blockchain related positions. They felt that the job postings implied this was possibly going to happen sooner rather than later, but was unable to find any evidence to confirm the claim made by Bloomberg.

This should stay Unconfirmed imo. I wanted to post the Medium article though because I think it is a good evaluation of the claim and fits with our community.