Segwit, and batching, a must in these times

[Wind_FURY]

Because of a surge in demand, Bitcoin's mempool is flooded, and the blocks are full, taking network fees higher, and our transaction confirmations longer.

https://jochen-hoenicke.de/queue/#0,24h

Veriphi wrote a case-study and found that the network, and ALL of its participants together, could have saved $500,000,000 in fees through Segwit, and batching.

https://veriphi.io/en/blog/case-study-how-segwit-and-batching-help-you-save-money

I'm merely spreading some awareness for us to learn about the benefits of batching, and maybe for everyone to ask their favorite Bitcoin exchange/casino/merchant/service to start using it, to help decongest the network, and help ALL its participants save more.

https://veriphi.io/segwitbatchingcasestudy.pdf

The purpose of this case study is to determine the savings benefits that employing SegWit and Batching in Bitcoin transactions would have had, had there been a full adoption. Nearly half a million blocks of transactions have been scrutinized in order to calculate the amounts of bitcoins and mega- bytes saved thanks to these technologies. The savings potential presented is signi cant and those conducting large amounts of transactions should seriously consider employing these tools in order to remain competitive and save money.

Major takeaways:

* From January 2012 to June 2020 ( until the 637,090th block) 211,266.95 bitcoins were paid in fees to miners. This amounts to a total of around $1,954,219,287 USD with a Bitcoin price of $9250.00 USD at the time of latest update in July 2020. 


* Over 21,131.97 bitcoins could have been saved by Bitcoin users if they would have all been using Transaction Batching. 190,134.98 bitcoins could have been paid in fees instead of the 211,941.32 bitcoins, which represents savings of 9.97%. The 21,131.97 bitcoins saved represents a staggering amount of $195,470,722 USD (with a Bitcoin price of $9,250.00 USD).

*  From August 24th 2017 to the 30th of June 2020, 36,685.72 bit- coins could have been saved by Bitcoin users if they would have all been using SegWit Native (Bech32). Fees would have amounted to 59,848.61 bitcoins, down from 96,534.33 bitcoins actually paid in fees, which is 38.00% in savings. The 36,685.72 bitcoins saved represent $339,342,910 USD (with a Bitcoin price of $9,250.00 USD). 


* The advantages brought through optimized fee management techniques such as SegWit and Batching are mostly impressive and apparent during high transactional activity periods. A large percentage of the possible savings would have been achieved in only a few months over the spawn of 8 years and 6 months analysed. 


* Bitcoin market actors, such as exchanges, should strongly consider imple- menting any optimized fee management techniques with clear nancial advantages. In anticipation of future price appreciation and higher trans- actional activity, Bitcoin users have to acknowledge that bitcoin transac- tion fees will continue to rise and prepare consequently. 



What is Batching?

Bitcoin transactions include inputs, the coins being spent, and outputs, where the coins get sent. In contemporary wallets and applications, Bitcoin transactions have at the very least 2 outputs for the most part, one where the money gets sent and another where you get your change back to another address you control. What if you want to send funds to multiple addresses? You could make two transactions with each two outputs including the change output, however doing so makes you pay fees for every separate transaction. You could instead use Transaction Batch- ing. Brie y, this technique simply signi es having more than 3 outputs in a transac- tion, thus making a payment to two different parties at the same time.

A typical P2PKH (Pay to Public Key Hash) Bitcoin transaction (As seen in Figure 2) containing one input, an output and a change output would be 226 bytes in size. Knowing this, every additional independent transaction would then at least have another 226 bytes. For entities making multiple transactions, the fees paid can add up quickly. Imagine having the below image multiplied by every individual transac- tion you make.

Bitcoin used to have a limit of 1MB for every block of transactions. Since this limit was replaced with a block weight limit that can reach 4MWU, we can calculate that if all the transactions in a block were using SegWit, the block size could reach 2.4 MB. Even though we can t more transactions now, the case still is that we can only t a certain amount of transactions in every block depending on their sizes. The larger transactions must then pay a greater amount in fees if they want to be included in a block, since space is limited. The impact of Transaction Batching becomes more apparent the more transactions we make.

Benefits of Batching

The input of the above transaction represents about 65% of the transactions (148 bytes) and the outputs around 30% if we assume it has a change output. In order to save on fees paid when making multiple transactions, you can either use as few inputs as possible by consolidating your UTXOs, or you can use Transaction Batching, where the output part scales linearly but the input part stays the same, no matter the amount of outputs.

By aggregating multiple outputs into a single transaction, you are effectively saving up on space required for your payments, thus reducing the amount of fees paid to miners. Instead of making a second separate payment of 226 bytes, effectively doubling your fees, you are instead paying 15% (34 bytes) more than the single transaction for each output (as seen in Figure 3).

The savings further increase the more outputs you add onto the transaction. A batched transaction that contains 10 outputs represents 25% of the total byte size of 10 individual payments, saving you up to 75% in fees paid.

Batching also has the bene t of minimizing your UTXO set since you’d only have one change address per batched transaction instead of one for each in- dividual payment made. By reducing your number of change outputs you are further saving on fees spent since a change output typically takes up at least 69 vbytes when spent later on.

There are however potential privacy concerns when it comes to batching. A user can very easily see every other address that received a payment in the same transaction, or someone else analysing these transactions can identify users whose public address is known and tie them to an exchange with an amount. On the other hand, even if the transactions are done separately, those links can still be inferred by analyzing the change outputs.

Batching is a simple yet effective method to implement in order to save on fees paid by making your payments more ef cient. Large scale organizations that have multiple transactions per day can greatly bene t from batching, especially during times when fees spike like in the December 2017 to January 2018 periods where fees rose signi cantly in USD terms. As your number of payments increases, there are potentially huge savings to be made which in turn can either increase your revenues or be passed along to your customers.

[1715567181]

1715567181

[1715567181]

1715567181

[davis196]

I agree that batching can provide some benefits,but as the article author is stating,only big organizations that are making a significant number of transactions every day will get the benefits.
Such big organizations can be only the major crypto exchange platforms and the big online crypto casinos.
If you are a small crypto trader or your cryptocurrency business is below medium-sized,implementing batching won't have a significant impact over the fees that you are paying.

[DooMAD]

I agree that batching can provide some benefits,but as the article author is stating,only big organizations that are making a significant number of transactions every day will get the benefits.
Such big organizations can be only the major crypto exchange platforms and the big online crypto casinos.
If you are a small crypto trader or your cryptocurrency business is below medium-sized,implementing batching won't have a significant impact over the fees that you are paying.

Arguably, the larger companies batching their transactions does benefit everyone, because it potentially frees up space for other transactions.  If an exchange had 100 separate transactions in a single block, it would take up more space than one singular batched transaction to 100 different recipients.  The more efficiently the available space is utilised, the lower the fees will theoretically be to get your transaction included in that block.

[franky1]

saving 20k of 200k
so your saying only a 10% improvement if you do a batching campaign
however
if people campaigned devs to implement a better fee calculator. and even a new fee formulae. that would have saved more than 10%

also instead of allowing 4mb of weight for bloated scripts. allowing 4mb weight for lean transactions. would have saved more than 10%
oops
did people forget that part

sounds to me someone making this topic was too quick at grabbing a segwit positive campaign. and yet doesnt actually understand the true history. nor done any true math of best efforts to get best results

seems to me all they want is more exchanges to just use segwit and then batch transactions
the only way to batch transactions is to not do payment on demand.. but implement delayed payments where people only see a transaction appear once every 30-60minutes so that the exchange can batch up 3-6 amounts of transactions together
exchanges already do multispend transactions per block. but this study is based on the return change being reused again in the next block which the article writer calculates that a saving can be made if all done at once,, missing the point that not everyone asks for withdrawals at the same time or wants to pay all their bills in one go

funny part is its only a 10% difference
as for the segwit.. only 39% saving..
well if people campaigned to devs to not only
implement better fee calculators
implement a fee priority formulae
but also
instead of just letting 4mb space for bloaty scripts. but instead 4mb for lean transaction data..
it would allow more transactions. meaning less fees
far far far cheaper than the 10%-39%

kinda funny how these guys are pumping hard these schemes where as other options that have been declined by these certain segwit lovers would have helped save more then 40%

remember the segwit promise of 75% discount.. atleast they are admitting its more like 40%
all thats remaining is them to admit that if they done the other bip.. it would have been more than 75% without having to adopt new transaction methods
now i wait for the standard insults about how i am ruining their promotion/campaign
but just a reminder. devs have already said for years now 4mb wont destabilise the network. so dont even try using that false narrative about 4mb would hurt. as it got debunked in 2017

[ranochigo]

Perhaps as well as a higher usage of LN? Admittedly, most people wouldn't use it so it's better to just concentrate on optimizing the onchain transactions.

You should also shift some attention to companies like Bitmex. They use multisig (with uncompressed keys) P2SH on their addresses and broadcast the transaction from it everyday, resulting in a huge unnecessary bloat on the mempool at certain points of the day.

[tbct_mt2]

I remembered in 2017 when BTC price rose up to the Moon and mempool was congested, even big exchanges had to delay their customers ' withdrawals for a while because they can not waive the fees for customers for every personal withdrawal. During that period, they batched customer withdrawals to save their fees on the network.

It says another story that if people put their funds on third-party services, they have to wait to get withdrawal acceptance from those services. There is no instant withdrawal as they usually claim to provide to customers.

[squatter]

You should also shift some attention to companies like Bitmex. They use multisig (with uncompressed keys) P2SH on their addresses and broadcast the transaction from it everyday, resulting in a huge unnecessary bloat on the mempool at certain points of the day.

I've been saying it for a long time -- they are a plague on the network, the worst of any service. I'm amazed at how no one seems to care.

Coinbase was a contender for that spot, but they finally started batching transactions a few months ago.

if people campaigned devs to implement a better fee calculator. and even a new fee formulae. that would have saved more than 10%

That is easier said than done. The fee market is dynamic and not predictable. No algorithm can predict when a giant influx of transactions will suddenly occur, or when the network will go an hour or two without finding a block, or what other users are willing to pay in the future. It's impossible to write perfect fee estimation algorithms.

[mk4]

We've already had SegWit for a while now yet freakin Blockchain.com and Binance still doesn't have them, even though Blockchain has a significant control over the wallet marketshare and Binance is a giant exchange. Smh. And Binance has the audacity to have a "beginners guide to segwit": https://academy.binance.com/blockchain/a-beginners-guide-to-segretated-witness-segwit

[DooMAD]

if people campaigned devs to implement a better fee calculator. and even a new fee formulae. that would have saved more than 10%

That is easier said than done. The fee market is dynamic and not predictable. No algorithm can predict when a giant influx of transactions will suddenly occur, or when the network will go an hour or two without finding a block, or what other users are willing to pay in the future. It's impossible to write perfect fee estimation algorithms.

Indeed.  While we're at it, let's write some software that tells you exactly how much you need to bid to win an auction before the auction has taken place, you have no idea who the other bidders are, or what their bids will be.  Easy, right?   

[BrewMaster]

Perhaps as well as a higher usage of LN? Admittedly, most people wouldn't use it so it's better to just concentrate on optimizing the onchain transactions.

IMO LN is not viable option, unless you frequently make transaction.

it can actually be a very viable option. look at why fees spike most of the times, in 90% of the cases the fees spike hard when price has a big rise or a big drop. that means it is the transactions that go to and come out of exchanges that is causing the spike.
now if a good portion of those transactions could go to a second layer then it would solve most of the fee problems. it may not be a complete fix though.

[gentlemand]

We've already had SegWit for a while now yet freakin Blockchain.com and Binance still doesn't have them, even though Blockchain has a significant control over the wallet marketshare and Binance is a giant exchange. Smh. And Binance has the audacity to have a "beginners guide to segwit": https://academy.binance.com/blockchain/a-beginners-guide-to-segretated-witness-segwit

Roger Ver is an investor in Blockchain.com which means he's not going to make 'Corecoin' more usable if he can help it. As for Binance, they care so little about Bitcoin's nature they probably think saying 'Segwit' out loud three times when you send gets the job done.

Are Veriblock still stinking up the blockchain? They've been responsible for a huge number of transactions at various points.

[pooya87]

if people campaigned devs to implement a better fee calculator. and even a new fee formulae. that would have saved more than 10%

That is easier said than done. The fee market is dynamic and not predictable. No algorithm can predict when a giant influx of transactions will suddenly occur, or when the network will go an hour or two without finding a block, or what other users are willing to pay in the future. It's impossible to write perfect fee estimation algorithms.

Indeed.  While we're at it, let's write some software that tells you exactly how much you need to bid to win an auction before the auction has taken place, you have no idea who the other bidders are, or what their bids will be.  Easy, right?   

the auction isn't going to happen in the future, it is ongoing so there is no need to predict it just to look at what others have bid. the problem is that when someone bids $50 for example the next dude bids $100 and the next $200 while they could have won the auction with $51 just as easily.
there are lots of software out there that are doing a very poor job at fee estimation and even lots of shitty sites that are popular too that are intentionally inflating fees (eg. bitcoinfees.earn.com these bastards are currently suggesting 220 when the worst case is barely 30!!!!).
it is not rocket science!

[Wind_FURY]

Snip

I know you disagree with the Core developers design-decisions, and development path of maintaining decentralization through keeping the block size smaller. I was merely pointing out the fact the exchanges/merchants/casinos should use the network the proper way by Segwit, and batching, and help decongest the network.

We've already had SegWit for a while now yet freakin Blockchain.com and Binance still doesn't have them, even though Blockchain has a significant control over the wallet marketshare and Binance is a giant exchange. Smh. And Binance has the audacity to have a "beginners guide to segwit": https://academy.binance.com/blockchain/a-beginners-guide-to-segretated-witness-segwit

Roger Ver is an investor in Blockchain.com which means he's not going to make 'Corecoin' more usable if he can help it. As for Binance, they care so little about Bitcoin's nature they probably think saying 'Segwit' out loud three times when you send gets the job done.

Are Veriblock still stinking up the blockchain? They've been responsible for a huge number of transactions at various points.

That's funny, because if he removes "Corecoin" Blockchain.com WILL LOSE 90% OF USERS.

[DooMAD]

there are lots of software out there that are doing a very poor job at fee estimation and even lots of shitty sites that are popular too that are intentionally inflating fees (eg. bitcoinfees.earn.com these bastards are currently suggesting 220 when the worst case is barely 30!!!!).
it is not rocket science!

Okay, but if the "shitty sites" are popular, whose fault is that?  Surely at some point, users need to take some responsibility for what they're doing.  I don't see how you can create a world where sub-par services like the ones you describe don't exist.  People just need to learn how to find the good ones and use those. 

It's not rocket science, no.  But if it were that easy to solve, would it even be a problem to begin with?  Clearly there are improvements that could be made, but people have an innate talent to flock to what they already know.  You could get the most talented coder in the world to write a new fee estimation algorithm which is superior in every way and yet some people would still be lazy/stupid and use a crap one instead.

[gentlemand]

That's funny, because if he removes "Corecoin" Blockchain.com WILL LOSE 90% OF USERS.

We know that. He knows that. Which is why he will never, ever, ever make it easy for us AxaBorgCoreStream zombies while he still has some influence. The logical conclusion is that everyone deserts the services he's intentionally crippling but there appears to be little sign of that so far.

Okay, but if the "shitty sites" are popular, whose fault is that?  Surely at some point, users need to take some responsibility for what they're doing.  I don't see how you can create a world where sub-par services like the ones you describe don't exist.  People just need to learn how to find the good ones and use those. 

People are lazy and they gravitate towards brand recognition. That's what humans do and it ain't changing soon.

[pooya87]

there are lots of software out there that are doing a very poor job at fee estimation and even lots of shitty sites that are popular too that are intentionally inflating fees (eg. bitcoinfees.earn.com these bastards are currently suggesting 220 when the worst case is barely 30!!!!).
it is not rocket science!

Okay, but if the "shitty sites" are popular, whose fault is that?  Surely at some point, users need to take some responsibility for what they're doing.  I don't see how you can create a world where sub-par services like the ones you describe don't exist.  People just need to learn how to find the good ones and use those. 

It's not rocket science, no.  But if it were that easy to solve, would it even be a problem to begin with?  Clearly there are improvements that could be made, but people have an innate talent to flock to what they already know.  You could get the most talented coder in the world to write a new fee estimation algorithm which is superior in every way and yet some people would still be lazy/stupid and use a crap one instead.

well they don't want to learn and they keep using the same low quality services and even recommend it to other people. the problem is that their choice affects everyone else's experience too.

it is not that easy but it is not as hard as you may think either. for starters a lot of clients lack the option to let user set any fee they want specially in fractions of a satoshi (like setting 5.25) so they have to jump up. they also don't give users a concise information about the memory pool status at the time they want to create a transaction so that user can make a better decision.

[DooMAD]

it is not that easy but it is not as hard as you may think either. for starters a lot of clients lack the option to let user set any fee they want specially in fractions of a satoshi (like setting 5.25) so they have to jump up. they also don't give users a concise information about the memory pool status at the time they want to create a transaction so that user can make a better decision.

That's a fair point.  What more can we do to encourage people to use the clients doing a better job in that regard?  And is there a constructive manner in which we can pressure the clients that currently lack in that area to improve in future?

[pooya87]

~
That's a fair point.  What more can we do to encourage people to use the clients doing a better job in that regard?  And is there a constructive manner in which we can pressure the clients that currently lack in that area to improve in future?

we start by discouraging people from using any tool that is severely flawed. for example i have been telling people to avoid using the site i mentioned above for over a year now. good news is that many listened after they saw the flaw. that goes for any other wallets (desktop, mobile, web,...) that are doing a poor job.
next step is reaching out to developers and asking them for improvement, this could be done by creating issues on their github repository or better if we have any programming skills then we should start contributing by writing code and creating a PR instead. and obviously anything that is not open source should not even be used by anyone to begin with.

[Pffrt]

As far as I know, almost all the services, exchanges use batch payment. Coinbase is one who didn't use batch payment but they too do batch payments. Anyway, it's good to use batch payment for its users too. Coinbase charges less fee prior to start batch payments.
While segwit is still to be accepted by a lot of services I know. They don't care because they need to modify their code to enable segwit. May be that's the reason.Hopefully, soon everyone will realize its benefit and enable that.

[squatter]

Indeed.  While we're at it, let's write some software that tells you exactly how much you need to bid to win an auction before the auction has taken place, you have no idea who the other bidders are, or what their bids will be.  Easy, right?  

the auction isn't going to happen in the future, it is ongoing so there is no need to predict it just to look at what others have bid.

That would only work if a block were found precisely after you submit your transaction to the network. The auction closes when a block is found.

Fee estimation requires more than just placing your transaction at the top of the queue or 1MB from the tip. It requires estimating the likelihood of finding another block within a given time, and also infers from past blocks what future users will pay. Both of those things are unpredictable on a fundamental level. Looking back at previous blocks and the mempool just provides a guesstimate.

the problem is that when someone bids $50 for example the next dude bids $100 and the next $200 while they could have won the auction with $51 just as easily.

I agree, this is a major problem.