About Proof of Stake


Tom clears up the confusion in Ethereum’s new method of verifying its blockchain and explains the different approaches blockchains utilize.

Featuring Tom Merritt.

Episodes mentioned:

About Blockchain



A special thanks to all our supporters–without you, none of this would be possible.

Thanks to Kevin MacLeod of Incompetech.com for the theme music.

Thanks to Garrett Weinzierl for the logo!

Thanks to our mods, Kylde, Jack_Shid, KAPT_Kipper, and scottierowland on the subreddit

Send us email to feedback@dailytechnewsshow.com

Episode transcript:

I heard one of those big crypto outfits merged something to proof of stake

Does this mean only rich people control cryptocurrency?

Or does it somehow involve beef?

Confused? Dont’ be!

Let’s help you know a little more about Proof of Stake.

In our episode on how a blockchain worked we talked about one of the big advantages of the blockchain is that it’s really hard to fake a record or manipulate how transactions get recorded. It is very difficult to alter the ledger.
See that episode for more details, but one way to preserve this aspect of decentralization is to make it hard to be the one who adds and validates new blocks to the chain. You want to make it hard and hard to predict who will get to do the block validations. In that episode we talk mostly about proof of work. This is the system that requires a node to be the the first to calculate a very hard problem in order to have the right to add the block to the chain. In Bitcoin’s blockchain it’s called “mining” because you’re also rewarded with a bitcoin. The calculation is such that just having the most power won’t assure you do it first, but you will need a lot of computer power, so not just anyone can attempt it. And the downside is, it uses a lot of energy.
But we briefly mentioned another method called Proof of Stake. Instead of requiring work to be done, Proof of stake uses a random factor to distribute who gets to add the block. This still prevents actors from dominating the block validation, but without the energy use caused by proof of work. One proof of stake method is coin age. The actor that has the coin they have held on to the longest without spending OR being used to prove their stake, gets to calculate the next block. The example we give is you have kept a coin for 90 days another actor has kept a coin for 75 days and everybody else has kept their coins for 60 days or less. You get to calculate the next block, but then your coin age is reset to 0. The next block would go to the actor who had a coin for 75 days. In practice other factors are also used to prevent participants with large collections of coins from being able to dominate the network.
Ethereum switched from proof-of work to proof of stake. It wasn’t the first blockchain to implement proof of stake but it certainly is the biggest.
Ethereum uses a system of at least 16,384 validator nodes– it has more than 400,000. To become a validator, a node must “stake” 32 ETH. That’s 32 ETH per node. The stake cannot be spent. Validators are chosen at random to propose validating a block, using Ethereum’s own “good enough” RANDAO system. A committee of 128 validator nodes then attests to the block.
Each validator node on the committee adds its verification to a block of “shards.” When 128 shards have been attested that shard block is done. When 2/3 of the validators on a committee agree that the transaction is valid, it is finalized and closed and replicated throughout the blockchain. Validators receive transaction fees as a reward for both proposing a block and attesting to it.
One side note about the shard aspect of this. Sharding lets multiple blocks be processed at once. So instead of validating a block. Adding it to the chain and validating the next block. Ethereum will run 64 shard chains at once. As each shard block is finished its added but it doesn’t have to wait for all the other shard blocks to finish. It can just be added when it’s done. This means the Proof of Stake Ethereum chain can process transactions at least 64 times as fast as the proof of work chain could.
You could do sharding on a Proof of work system, but because computer power would be lower on each shard chain, it would be less secure. Since Proof of stake doesn’t use computer power as an element of choosing the validator, it is not susceptible to that weakness.
As we mentioned Proof-of-work uses energy. It requires computations be difficult so that it’s hard to win the right to make a block. Proof-of-stake doesn’t use nearly as much energy since there is no intense computation required to win the right to record a block. Before it switched to proof-of-stake on September 15th, Etehreum estimated energy consumption on the Ethereum blockchain would drop by around 99.95%.
In a proof of work system the limiting factor on someone validating blocks is equipment and energy. In a proof of stake system, it’s the amount of tokens they can buy or already own and then just random luck.
Proof of stake doesn’t reward the actor with the most expensive equipment. When balanced properly it’s expensive enough to buy in as a validator that there is a barrier to bad actors from entering, but it is not so high a barrier that a wealthy actor can dominate the system. Random assigning of validators means if the pool of validators is diverse and balanced, then it would be extremely difficult to manipulate the system.
One safeguard is the requirement to use tokens as a stake in order to validate and they can lose that stake. Misbehavior can see a participant lose some or all of their stake. At the very base if they qualify as a validator and then fail to participate they lose their stake. The penalties are small enough on each instance, that getting knocked offline won’t wipe out a stake. Consensys.net estimates that “if a validator is participating correctly more than half the time then her rewards will be net positive.” This is meant to keep bad actors from teaming up to sabotage the network by not validating new blocks.
On the Ethereum chain there are two dishonest behaviors that can result in an actor losing their entire stake. One would be proposing multiple blocks at once, also called equivocating. Another would be proposing contradictory attestations. In other words saying the block is valid but also saying it’s invalid. The more validators that attempt this at once, the higher the penalty. One validator acting alone would lose 1% of their stake, but the more that attempt it the higher the percentage up to possibly losing the whole stake. A violator can also be ejected from the network.
Other kinds of attacks on the blockchain can also be attempted. Ethereum argues that since validators do not need to do energy-intensive work to conduct validation, they have more flexibility in fighting off attacks. For instance, if a bad actor attempted to fork the chain to their advantage the good actors could promote a minority fork of the blockchain without the bad actor as the valid one and cut out the bad actors stake.
Most proof of stake systems have other security features that they intentionally do not advertise in order to reduce attempts to circumvent them.
Proof-of-stake is more complicated. On Ethereum, users have to run three pieces of software to participate. Proof-of-work systems generally only need one.
But in the end Proof-of-stake is more accessible and quite a bit more energy efficient. Its proponents argue that it should also be more efficient. Expect more blockchains to copy Ethereum’s successes, and learn from their mistakes. In other words, I hope you know a little more about Proof of Stake.

%d bloggers like this: