Does an ideal of 0.25 really mean a 25% chance of getting a slot each Epoch

Staking & Delegation Operate a Stake Pool
21

Thanks

Sorry maybe I was getting my terminology wrong but this is exactly what I run to check for slots which confirms I did not get any

I can confirm that this is not the case and I do not have a problem

1 Like
22

There’s nothing wrong with the protocol. It’s about the people. Cardano is a tool. A hammer is also a tool. If I use a hammer to build a house, then I’m using the tool as it is intended to be used, and in a way that is useful to society. If I use a hammer to hit someone over the head, then that is not useful to society and may also be expected to escalate into a legal matter fairly quickly. In any case, there is nothing wrong with the hammer. The problem is behaviour. GNP1 seems to be raising concerns that perhaps Cardano is being used irresponsibly, even though the protocol may be functioning exactly as intended.

For example, using a familiar point of reference, I imagine how a similar conversation related to concerns about investment returns may go with a bank. I can’t imagine such a conversation going the way that the above thread went. The discussion above is more akin to what you may expect if you lost money at a casino. I would wonder about determining returns on sizable investments using what may be conceivably considered a “loot box” reward strategy in some ways.

The funds at stake are significant. GNP1 raises some very important concerns about what constitutes appropriate use of Cardano as a tool for the greater good.

CHG

23

41 Epochs since Epoch 320

1 slot allocated in Epoch 351 with a continuous ideal of 0.25

but of course the protocol is working fine

certainly looks like something changed after Epoch 320

image
image1606×317 54 KB

24

@GrahamsNumberPlus1, I just wanted to say thank you for your post. I am also struggling to mint blocks as a new small pool. I do not have the amount others have when it comes to pledge and amount staked. But as you mentioned, there is a significant monetary amount investment on my part as an individual.

My hope is to mint a block (I don’t think that is to much to ask) before the cost of the infrastructure becomes to much for me to bear. Furthermore it is really difficult to convince others (even friends) to stake to a pool with 0% ROS. Regardless of anyone’s willingness to support the cause and/or network they will want some kind of return greater than 0%.

Like you I am willing to stick around and continue to be part of this amazing community. But just need some rewards to be able to sustain the pool and attract others. I believe in Cardano like the rest and truly hope to stay part of it as an SPO. This thread has been very helpful and informative.

2 Likes
25

Welcome to the community

This is so true and unfortunately the way the Pool Fee is set up it is completely unfair for small pools - please refer to my other post in relation to that matter ( 340 ADA Pool Fee is becoming more of a concern - #10 by GrahamsNumberPlus1 )

I have been doing this for 2 years now and it takes every bit of strength to try to find reasons to keep going right now as it is near impossible to get any returns or rewards. I only have around 300,000 ADA in stake but that was almost 1M USD worth at the height of the bull market but yet it only has produced me one slot allocation in 41 Epochs (should have been around 10) - prior to Epoch 320 I minted 26 blocks so not sure what has happened

Anyway I don’t want to put you off and it is a great community and I wish you the best of luck but I just wish they would make this whole process a fair playing field and stop protecting larger pools with the 340 min Pool Fee.

26

Our pool also has no block for about 20 epochs, with 300K active stake

Could it be the problem that we only have one relay?

1 Like
27

Hi there - now my understanding is that the slot allocation only depends on the pool being successfully registered on the network and any configuration issues and relay issues only come into play when your pool is tasked with minting the block at the specific slot (if I am wrong about this then someone please correct me). If you have been selected as slot leader in the past then things should be ok and going 20 Epochs is extremely unlikely. Have you checked every 5 days to see if you are being allocated a slot?

This is what I have been doing and I simply have not even been allocated one so minting is not relevant at this stage?

28

42 Now - say what you want something changed after Epoch 320

29

44 Epochs now - say what you want but something changed after Epoch 320

1 slot allocated in Epoch 351 with a continuous ideal of 0.25 - 0.26

but of course the protocol is working fine

surely this is at the stage where it warrants some investigation - it is just not right!

image
image2406×427 105 KB

image
image1008×177 7.38 KB

1 Like
30

Ok so I contact IOHK again and ask them to look into this and the response is just the same. IOHK have confirmed that they have looked into my pool and their are no technical issues with the operation of my Stakepool. They then go on to say that my pledge of 25K is small and my active stake it small and if I increase this then I should have more chance to mint a block.

Am I missing something here? I have just gone 44 Epochs with one slot allocation and I am being told that nothing is wrong. The ideal is 0.26 and has been throughout this time which means roughly a 1 in 4 chance each Epoch. This means approx 10 Blocks during this period and I have been allocated 1.

I understand this works like a lottery system but at what point are we going to start to actually question the protocol or any issues with possible vrf bias? - at 100 Epochs or 200 Epochs - which will never happen as why would any delegator stay with my pool at this point in time.

I am suggesting that this is not gamblers fallacy or extremely long runs of bad luck but maybe actually a potential issue that discriminates against small pools. Come on IOHK I deserve something more than this - every aspect of this has been examined and I don’t care what you say rolling a 4 side dice will not continue to fail to roll one number over and over and over again - it does average out over time - look at the Maths - please!

1 Like
31

There is nothing worse than feeling like you have to blindly trust a “block box”. Unfortunately crypto is like magic to most of us.

In order to satisfy yourself, maybe you could look at the actual formula for how to use your key to check if your are the leader for any particular slot.

I know it is something along the lines of the following:

  • Query the VRF seed for the epoch (from blockchain)
  • Query the total amount of Ada staked for the epoch (from blockchain)
  • Query your pool’s stake for the epoch (from blockchain)

Then for each slot number use your pool’s vrf key to ?encrypt? a combination of the epoch VRF seed + slot number in some format and check if the output of this function produces a number that is lower than your pool’s staked Ada divided by the total staked Ada.

The formula is something like that. Maybe @HeptaSean can properly inform us, he is super knowledgeable about this sort of difficult technicality.

If you can read Rust code, the cncli tool does the calculation. It might be easier to find the calculation in Andrew Westberg’s code rather than the IOG Haskell code base.

Maybe you could manually do a few calculations using the crypto operations for just a few slot numbers so that you gained familiarity with the formula? The crypto library functions will be available as c code and python libraries.

Maybe that would build confidence in how it works. Then you could check the result for some of the slots you were actually awarded in the past. Maybe you could write a simple function that used command line parameters for the seed, total stake, and pool stake figures, as well as the vrf key.

At the very least, going through this process would be educational for yourself and might help others feel more confident by knowing how the calculation is done if you posted the details.

2 Likes
32

Sorry, haven’t looked at the VRF formula in detail, just know in principle how it works. I nominate @AndrewWestberg. :stuck_out_tongue:

What I would do: Change the VRF key of the pool. How to Check if your VRF skey is properly registered to the Ledger might help in this. Especially, you have to wait three epochs for the new one to take over. (Others might comment on if this is a bad idea.)

My reason is that it is then not critical anymore if the old VRF key with which you have the problem is known. Makes it easier to get help in what @Terminada suggested, which would also roughly be what I would suggest. (And in the unlikely case that we really find a problem with the formula or code it is easier to say: “Here! Look! Here is the problem!”)

Ideally, we exhaustively check all slots back to the time, where you were still producing blocks.

And we can also plot/check the distribution of the random numbers themselves (the ones that have to go below the “limbo bar” given by your stake proportion).

2 Likes
33

Yes. And that will provide a good visual appreciation when you see several results that just missed by a small amount.

1 Like
34

I just realised from your screenshot jpg that you used the cardano-cli to check your leaderlog prior to the Vasil upgrade.

You should recheck your leaderlog now that we are post the Vasil change with the “–current” switch.

See my post here: Cardano-cli leaderlog incorrect if run before Vasil

2 Likes