March 29, 2020, 02:07:49 PM

Author Topic: Hashrate network 3 times higher than on Coinchoose  (Read 2412 times)

0 Members and 1 Guest are viewing this topic.

Offline AZIZ1977

  • Gold Miner
  • ****
  • Posts: 575
  • Karma: +25/-0
    • View Profile
Hashrate network 3 times higher than on Coinchoose
« on: August 04, 2013, 06:16:07 PM »
I noticed that the Network Hash Rate is 3 times higher than on coinchoose. Is Coinchoose showing wrong Network Hash Rate or is there something going on we dont know?
GLD: DvCsNBJowwzManPa6YSMsNFanFXMxqgFuk
BTC: 16F9CkTbDAAa76EzpqhEhQsRP93Gn49GpW

Offline MicroGuy

  • @realMicroGuy
  • Forum Administrator
  • *****
  • Posts: 3464
  • Karma: +47/-0
  • "Shoot for the Stars!"
    • View Profile
    • MicroGuy.com
Re: Hashrate network 3 times higher than on Coinchoose
« Reply #1 on: August 04, 2013, 06:48:00 PM »
I noticed that the Network Hash Rate is 3 times higher than on coinchoose. Is Coinchoose showing wrong Network Hash Rate or is there something going on we dont know?

I think maybe the hashrate on coinchoose is not updated in real time. Perhaps there is a lag in their reporting?
"The heart of any cryptocurrency can be found in the spirit of its community."

Offline Stouse49

  • Gold Engineer
  • DevTeam
  • *****
  • Posts: 711
  • Karma: +66/-0
    • View Profile
Re: Hashrate network 3 times higher than on Coinchoose
« Reply #2 on: August 05, 2013, 12:46:44 AM »
I checked coinchoose.com vs the client.  They matched on all parameters except for the hashrate.  In this case, the hashrate was 1/2 the client reported hashrate.  They may use a difference calculation. 

The hashrate is calculated by the client using the difficulty and the block solving times.  The network and nodes don't actually measure the hashing power of every pool and solo miner.

The calculated hashrate for each block will differ based on how fast a block is solved even while the actual hashrate of all the computers on the network may not change.  A lucky block may be solved in 10 seconds, resulting in a very high hashrate calcuation.  A 5 minute block will result in a low hashrate calculation.

What the client does to calculate hashrate (with the getnetworkhashps debug command) is take the time it took to solve the last 120 blocks and the difficulty used during that time.  This will give a number that smooths out all fast times of really lucky blocks and the slow times of the other blocks.  Since it is a hashrate of 120 blocks, sudden changes in network hashing (such as a pool going offline suddenly, will not show up in the hash calculation right away).  Goldcoin adjusts it's difficulty every 60 blocks using a similar calculation to ensure that the average block time is 2 minutes.  Coinchoose may be using a different number of blocks to calculate its hashrates.  It has a weekly hashrate as well, which definitely is a different calculation than what we see in our client at any one time.
« Last Edit: August 05, 2013, 12:50:21 AM by Stouse49 »
GLD: E1A6WogjpqU3KjjoRUD63f7kofrFozmyfg

Offline tadspoles

  • Legendary
  • *****
  • Posts: 629
  • Karma: +46/-0
    • View Profile
Re: Hashrate network 3 times higher than on Coinchoose
« Reply #3 on: August 05, 2013, 02:18:39 AM »
Stouse...this community is very lucky to have you.  Thank you so much for all your hard work.

Offline AZIZ1977

  • Gold Miner
  • ****
  • Posts: 575
  • Karma: +25/-0
    • View Profile
Re: Hashrate network 3 times higher than on Coinchoose
« Reply #4 on: August 05, 2013, 05:35:09 AM »
thnx stouse49 for your answere.
GLD: DvCsNBJowwzManPa6YSMsNFanFXMxqgFuk
BTC: 16F9CkTbDAAa76EzpqhEhQsRP93Gn49GpW

Offline dreamwatcher

  • Gold Miner
  • ****
  • Posts: 138
  • Karma: +11/-0
    • View Profile
    • CryptoCoin Explorer
Re: Hashrate network 3 times higher than on Coinchoose
« Reply #5 on: August 20, 2013, 02:35:24 PM »
Coinchoose uses my site (CryptoCoin explorer) for its information. The Goldcoin explorer still has a network hashrate API I developed a few months ago. I believe Sal is still using that API.

This was a very powerful tool that allowed for much flexibility in how the hash rate was calculated in terms of how many blocks and interval of blocks reported. It calculated average difficulty itself based on the parameters used.

I dropped the project because it would require a lengthy explanation on how to use it properly and I was starting to get some emails about inaccuracy that ended up being caused by the user using parameters that caused the inaccurate readings.

As of now the hash rate you see on a coins homepage (with a couple of exceptions) is taken straight from the daemon, but even then on some coins is not accurate as many coins have not changed the original Litecoin parameters to match theirs.