ByeBye clogged mempool, ByeBye huge tx fees

[stompix]

The situation right now (things might change by tomorrow, after all this is bitcoin )

Blockchain.info:
https://blockchain.info/unconfirmed-transactions
5065 unconfirmed transactions

Bitcointicker.co
https://bitcointicker.co/networkstats/
Mempool Transactions:   475

https://dedi.jochen-hoenicke.de/queue/#24h
On average 1000 tx on the mempool with fees over 5sat/b

As for the fees:

Estimatefee.com
https://estimatefee.com/
For a fee to be included in the next 6 blocks...~3.38 USD
If you want to wait 480mins (8h) the fees are 20 cents.

Bitcoinfees
https://bitcoinfees.earn.com/
The fastest and cheapest transaction fee is currently 120 satoshis/byte, shown in green at the top.
For the median transaction size of 226 bytes, this results in a fee of 27,120 satoshis.

27k satoshi are at current price around 2.14$ (slight difference here)


So good news for all the coffee lovers that were unable to buy a cup without spending 30$.
And even better, no more "accelerator services" threads and scammers on this forum.

[whohackedme]

I to want to see all green for bitcoin

But

We are down 1000-1500 less transaction per block vs 6 weeks ago

[pooya87]

that's what happens when the spam attack against bitcoin network stops

Bitcoinfees
https://bitcoinfees.earn.com/
The fastest and cheapest transaction fee is currently 120 satoshis/byte, shown in green at the top.
For the median transaction size of 226 bytes, this results in a fee of 27,120 satoshis.

whenever the fees are dropping this website is shit for suggesting good fees! currently transactions with less than 5 satoshi/byte will be high priority. paying any higher than that is making the fees go up on your own!
this is a much better website for seeing fees and also playing around to learn more about how transaction size is determined: https://coinb.in/#fees

So good news for all the coffee lovers that were unable to buy a cup without spending 30$.
And even better, no more "accelerator services" threads and scammers on this forum.

good news for coffee lovers is this: https://bitcointalk.org/index.php?topic=2674529.0

[Crypington]

Glad to see this.  What do you think caused this change?  I know the theory that miners and BCH backers were spamming the network to keep fees elevated ... did the “crash” make spamming less lucrative?  

[stompix]

I to want to see all green for bitcoin

But

We are down 1000-1500 less transaction per block vs 6 weeks ago

Let's say company A wants to pay four of its customers
A > B
A > C
A > D
A > E
4 transactions. But because of fees, they have chosen to
     | B
 A  | C
     | D
     | E
1 transaction. See the difference?

that's what happens when the spam attack against bitcoin network stops

Bitcoinfees
https://bitcoinfees.earn.com/
The fastest and cheapest transaction fee is currently 120 satoshis/byte, shown in green at the top.
For the median transaction size of 226 bytes, this results in a fee of 27,120 satoshis.

whenever the fees are dropping this website is shit for suggesting good fees! currently transactions with less than 5 satoshi/byte will be high priority. paying any higher than that is making the fees go up on your own!
this is a much better website for seeing fees and also playing around to learn more about how transaction size is determined: https://coinb.in/#fees

Yeah, bitcoinfees is pure garbage sometimes.
But I'm wondering about the culprit behind the change in mempool size.
Could it be that Coinbase is batching transactions?

[stompix]

Some funny transactions are getting confirmed:
https://blockchain.info/tx/4361e7346d3f235be35c8a9278540f8a1258298ba73ffaf4547a090ee0dc2255

This guy waited 3 months for a confirmation ,  2018-02-11 01:22:46 + 138,413 minutes
But it was worth it, he paid just  >     Fee per byte   1.005 sat/B

https://coinb.in/#fees   5 Sat/Byte
https://estimatefee.com/  (6 blocks), 1.5 $, anything else is 0.15$ 
Bitcoinfees is again 1001% wrong , telling you 120s/b.

[xdrpx]

This week and the previous few weeks have been a good time for people who wanted to perform transactions with lots of inputs / outputs and to send all of their dust and micro transaction inputs from multiple addresses into one SegWit address. If anyone hasn't done this yet, THIS is the right time. Move all your funds from legacy addresses into a SegWit address. Also, since you've already linked to coinb.in's fee estimator, I'd like to say I've been a fan of that page and it easily helps determine total transaction costs based on the number of inputs and outputs. I hope the miners start accepting and mining 1 Sat/byte transactions as well to eat through the remaining transactions.

So good news for all the coffee lovers that were unable to buy a cup without spending 30$.

I always loved to see micro-transactions happen with Bitcoin and it's been my long term desire. I was excited when Alex Bosworth posted this video https://twitter.com/alexbosworth/status/955870434230132736 of a Lightning-mainnet transaction to buy a cup of coffee. Lightning payments should bring down the fees to very tiny amounts, probably "pennies".

[TheQuin]

Bitcoinfees is again 1001% wrong , telling you 120s/b.

It nearly always will estimate many times over the required fee.

The predictions are based on blockchain data of the last 3 hours, as well as the current pool of unconfirmed transactions (mempool).
First, a likely future mempool and miner behavior is predicted using Monte Carlo simulation. From the simulations, it can be seen how fast transactions with different fees are likely to be included in the upcoming blocks.
The predicted delay shown here is chosen to represent a 90% confidence interval.

Although blocks are found on average every 10 minutes because there is variance in luck involved in mining there will frequently be gaps of 25 minutes or even 50 minutes to an hour. Taking a sample of the last 3 hours almost always involves one or more of these gaps. That means the 90% probability of getting in the first block will always require a very high fee to get in the first block immediately after one of these gaps. Then the problem is compounded by so many people using the overestimated fee. If you look at the bar chart the confirmed transactions in the last 24hrs always has a large spike wherever Bitcoinfees recommendation has been. Right now it shows 39,782 transactions at 120 sats/byte and then almost nothing until under 10 sats/byte.

[stompix]

Bitcoinfees is again 1001% wrong , telling you 120s/b.

It nearly always will estimate many times over the required fee.

The predictions are based on blockchain data of the last 3 hours, as well as the current pool of unconfirmed transactions (mempool).
First, a likely future mempool and miner behavior is predicted using Monte Carlo simulation. From the simulations, it can be seen how fast transactions with different fees are likely to be included in the upcoming blocks.
The predicted delay shown here is chosen to represent a 90% confidence interval.

Although blocks are found on average every 10 minutes because there is variance in luck involved in mining there will frequently be gaps of 25 minutes or even 50 minutes to an hour. Taking a sample of the last 3 hours almost always involves one or more of these gaps. That means the 90% probability of getting in the first block will always require a very high fee to get in the first block immediately after one of these gaps. Then the problem is compounded by so many people using the overestimated fee. If you look at the bar chart the confirmed transactions in the last 24hrs always has a large spike wherever Bitcoinfees recommendation has been. Right now it shows 39,782 transactions at 120 sats/byte and then almost nothing until under 10 sats/byte.

I really start think estimatefee is doing this on purpose
If there would be a problem with their 72h/3h model, that would have persisted only for a maximum of two days, but now they still predict the same tx as before.
It's either the model of their prediction which is flawed, the configuration of their mempool, or ...as I said , on purpose.
And I've watched that site quite a few times the last days, it's no difference even if blocks are mined faster or there is a gap, it's flat on 120.

As long as w have 5 days with almost every block having a few 5sat/b or at least 10sat/b it's clear that 120sat/b is insane.

I wonder if localbitcoins has dropped their receiving bitcoin fee....

I hope the miners start accepting and mining 1 Sat/byte transactions as well to eat through the remaining transactions.

Before that, I hope that we will see exchanges not spamming the network and implementing segwit
When I checked the mempool two days ago half of the largest txs where made by hitbtc:
https://blockchain.info/address/1EEqRvnS7XqMoXDcaGL7bLS3hzZi1qUZm1

[TheQuin]

I think it's the model. They've set too high a threshold with 90% confidence.

Localbitcoins deposit fee is 0.0001607 BTC which is a lot lower than last time I looked but that was a few weeks ago now. I'm pretty convinced they don't use bitcoinfees for their estimation as it moves around a lot more.

At least hitbtc use Segwit deposit addresses and sweep them with low fees.

https://btc.com/2718fd52b085af55a69099d0e97b70f403c109b933b78a7a3589e7f49eb6cf97

Size (rawtx) 154,400 Bytes
Virtual Size 81,835 Bytes
Fees Rate (BTC / kVB) 0.00001242

[joebrook]

Some funny transactions are getting confirmed:
https://blockchain.info/tx/4361e7346d3f235be35c8a9278540f8a1258298ba73ffaf4547a090ee0dc2255

This guy waited 3 months for a confirmation ,  2018-02-11 01:22:46 + 138,413 minutes
But it was worth it, he paid just  >     Fee per byte   1.005 sat/B

https://coinb.in/#fees   5 Sat/Byte
https://estimatefee.com/  (6 blocks), 1.5 $, anything else is 0.15$ 
Bitcoinfees is again 1001% wrong , telling you 120s/b.

Considering he paid that small fee to send such a small amount, I certainly hoped that it would have taken a year for that transaction to have been confirmed. Anyways we were all yearning for low transaction fees and we are getting it.

[stompix]

And slowly we're close on seeing the mempool empty , 2700 at the moment.

Been a while since I haven't seen this many half or even worse empty blocks


Also block 510199:
https://blockchain.info/block/000000000000000000558f7b8cd65ea586b42ce17050d144da641f144e98250f

Had just:
Transaction Fees   0.04183592 BTC
More than half of it was made out of trasnactions with 1sat/b.


Meanwhile bitcoinfees (not going to link to them) are still in a parallel world, telling people to spend 50sat/b.

[TheQuin]

Have a look at the last block 510,204 mined by F2Pool with a size of 51,871b. Not because there weren't enough transactions in the mempool but because the pool (and several others) are refusing to include any transaction less than 5 sat/byte.

https://btc.com/000000000000000000237fb82eb6f9b6455cafa2036f339dd9fc55da46d46bdd?page=1&asc=1&order_by=fee

   

[stompix]

Not because there weren't enough transactions in the mempool but because the pool (and several others) are refusing to include any transaction less than 5 sat/byte.

Yeah, seems like this is certainly the case and my prediction on seeing finally the mempool empty will never happen.
In 24 hours the situation has changed quite a lot.



8 full blocks and 15 lower than 1M.
Of course it is size and not weight in kWU but still.

And that garbage site still says "The fastest and cheapest transaction fee is currently 110 satoshis/byte,"  

[TheQuin]

Yeah, seems like this is certainly the case and my prediction on seeing finally the mempool empty will never happen.
In 24 hours the situation has changed quite a lot.

I think we actually did have an empty mempool situation a short while ago. I've been trying to work out what is wrong with that garbage site and posted another thread. https://bitcointalk.org/index.php?topic=2990283.msg30817058#msg30817058

I think it is to do with the mempool settings as each node has its own version of the mempool. It looks like they have many transactions that are not in the miners mempools. So https://jochen-hoenicke.de/queue/#1,24h was showing it empty and the miners were making small blocks consistent with what that showed but fees.extortionate was still showing 6k unconfirmed.

[alyssa85]

So good news for all the coffee lovers that were unable to buy a cup without spending 30$.
And even better, no more "accelerator services" threads and scammers on this forum.

Sadly, none of the shops that disabled bitcoin payment have re-enabled it. Which means you can't get a cup of coffee with bitcoin...

[Memenya]

Glad to see this.  What do you think caused this change?  I know the theory that miners and BCH backers were spamming the network to keep fees elevated ... did the “crash” make spamming less lucrative?  

Yes we need to know the reason but at least the price is less now rather than before.
I am so happy about this news although i am not coffee drinker with bitcoin but at least it can lift the price and demand increase.