Data is beautiful and I like to understand things. After @splinterlands launched their SPS validators, I was curious how much of an impact this has on the chain. Every transaction on Hive is immutable, and lives forever.
You can read more about SPS Validators here. Basically though, SPS Validators is the replacement for SPS staking rewards as well as offering a witness like functionality to SPS.
This post really isn't about SPS validators, but it got me on track of this discovery and discussion is needed to understand my interest.
I do want to quickly cover the Hive impact of validators, before sharing my reserch.
There are two types of SPS validators, the first are $LICENSE holders who just want the equivalent of the previous SPS staking rewards. These users run node software that need to produce one custom json every hour to be eligible for rewards.
The second are the real "validators" who act like a Hive witness for Splinterlands (to some extent) and are voted in to position and produce zero to three (?) custom jsons a minute.
As you can see these numbers are not insignificant and warrants further discovery.
As rewards didn't go live until recently, I had no idea what sort of impact this would have on the blockchain and our ongoing Hive "bloat" problem. Being the curious cat I am, I decided to do some digging and gather the data.
Below you will see what goes into 1,708 blocks. Pay attention to the custom json breakdown, specifically for sm_sps ids. These represent the new SPS Node broadcasts.
As you can see custom jsons are roughly 65% of all data stored on Hive. Of these, the new SPS Validators take up about 10% of that. I am sure this number will go up as more validators come online and users get their nodes registered.
From what I heard of hwo the validators work, I was expecting a lot more than this. It is certainly something I'd recommend keeping an eye on, but it isn't a cause for alarm. If you are curious like me, there is a lot of other interesting data in this breakdown to understand What we are putting into our blocks.

🥰🥰
Only a portion of the transactions will increase as more validators come online. It works similarly to Hive witnesses in that one SPS validator is chosen to validate the SPS transactions in every Hive block. The chances that a particular validator will be chosen to validate a specific Hive block is proportionate to the amount of SPS stake voting for them (like with the "backup" Hive witnesses).
All that just means that the SPS validator network as a whole will broadcast one custom_json operation for each Hive block as part of the main validation process and that will not change no matter how many validators are running on the network.
So in your data, you showed 2,153
sm_sps
transactions in 1,708 Hive blocks. 1,708 of those transactions were the validation transactions for each block and 445 were other transactions, such as validators that don't get assigned a block "checking in" to prove the software is running so they can earn rewards, and I think there are some other things like price feed updates.The one transaction per block will not change no matter how many validators are running, but the other transactions will increase proportionally as more validators come online.
Thanks! That lines up with what I’m seeing, I was expecting a lot more transactions initially.
Blocks? Memes?
You're welcome.
Even this early ten percent kind of seems like a lot. What happens if we have nine new projects doing something like this?
Just to add to the share % numbers. Its a share from the number of operations made. Its not acctual share of raw data. A vote has tiny amount of data compared to a 10k words post that also count as one.
The splinterlands jsons are somewhat the size of a comment.
I know you are awere of this.... just a added info for others reading the post how to interpret the data.