Greetings All,
I have created a new Cardano protocol simulator that uses MRI and Isosurface processing techniques to Compare Multipool Busting CIPs 23, 74, 75, 50, and 187?
Any one can experiment with the simulator which is found at the link below:
Rather than tracking all the pools and trying to figure out which ones are multipools I simply track the parameter space where it is possible for multipools to exist. These are any parameter settings where splitting pledge and delegation between multiple pools is more profitable than running a single pool.
The screen shot below shows that parameter space which has a Pledge axis, a Delegation axis, and a Margin axis.
The Fixed Fee is set with a slider and has a default value of 340 ADA.
CIPs 23, 74, and 75 call for removing the fixed fee so the Fixed Fee slider enables that simulation.
- The Render Mode is set to Isosurface
- The Isosurface level slider is set to 1.2. That means show me all the points in the parameter space where the multipool operator earns 20 percent more income by splitting his pledge and delegation over multiple pools
- The Multipool Count is set to 10 so we are doing the math assuming pledge and delegation are split over 10 pools.
- The Above Saturation Hidden button is selected. That means any combinations of pledge and delegation that exceed 75M ADA are hidden. No SPOs want to operate above saturation so we don’t need to see these points.
- The Color Metric pull down menu is set to Splitting advantage — ratio (N-pool ÷ 1-pool). That’s the second option on the menu.
- We selected Base Line No CIP for the Scenario Preset. That means no CIPs are being evaluated. Rather we are looking at the current behavior of the protocol.
The shape you are looking at is the parameter space where a multpool operator can earn 20 percent more ADA each epoch by splitting pledge and delegation across multiple pools. The purpose of the simulation see what CIPs make that shape disappear.
Now we press the Fee Reform Only button and the shape disappears as seen in the screen shot below.
Spoiler alert. CIPs 23, 74, and 75 win. These are the CIPs that eliminate the Fixed Fee
Some will argue that fixed fees help operators cover the cost of running the pool. But a small honest stake pool operator can make up the fixed fee by raising margin
Fixed fee is flat (a fixed ADA amount, regardless of pool size). 170 ADA is a big ROA drag on a small pool’s delegators and a rounding error for a large pool’s. It actually makes small pools less attractive to delegators. Margin is proportional to net rewards, so it’s size‑neutral.
So shifting cost recovery from a flat fee to margin is arguably better for decentralization, because it removes a charge that was disproportionately punishing small pools’ delegator ROA.
Next the CIP-50 Only button is pressed.
As we see in the image below, only a small part of the isosurface was removed.
In MRI mode (shown below in the next two screen shots) it can be seen that the section of the isosurface that was removed is the zero pledge regime.
That bright yellow area in the screen shot below shows a masive advantage for multipool operators at when pledge is at zero.
As seen below, enabling CIP-50 shuts down that incentive.
But CIPs 23, 74, and 75 also shutdown incentive for multipool operation at zero pledge so you may be wondering what CIP-50 offers that CIPs 23, 74, and 75 do not. CIP-50’s main goal is to increase Sybil resistance by forcing stake pool operators to put up some pledge in order to benefit from the delegation they control. Disincentivizing zero pledge multipools is an extra benefit but it is not the main value of CIP-50. Sybil resistance (making sure reward‑earning stake scales with pledge) is the main value of CIP-50
Next the CIP-187 Only button is pressed. As you can see below, only a small portion of the multipool parameter space has been eliminated. This is a different section of the regime than CIP-50 removed. That’s interesting, but remember, CIPs 23, 74, and 75 knock out the entire regime. So now you may be wondering what does CIP-187 do that CIPs 23,74, 75 and 50 do not.
The problem CIP-187 fixes is that delegators are punished for doing the right thing
Under today’s formula, the pledge bonus a pool earns is shared across all the stake in the pool. So when you delegate to a good, high-pledge pool, you dilute that bonus — and every delegator already in the pool sees their return go down. A quality pool gets less attractive the moment anyone joins it. The protocol’s equilibrium assumes delegators will rationally fill good pools, and then the formula punishes them for doing exactly that.
CIP-187 changes one thing: it scales the pledge bonus by utilization — how full the pool is relative to saturation. An empty pledged pool earns no bonus; a half-full pool earns half; a saturated pool earns today’s full bonus. The pledge bonus per delegated ADA becomes independent of pool size. Adding delegation no longer dilutes anyone.
A second brake on splitting — one that survives fee reform
There’s a bonus the cube does hint at, faintly. Utilization scaling quietly punishes the multipool operator too: split a real pledge across N self-funded pools and each pool is now under-utilized, so each one’s pledge bonus is scaled down — the total pledge bonus across all N pools shrinks by a factor of N versus running one full pool.
That matters because it’s a brake on splitting that does not depend on the fixed fee existing. Fee reform closes the fee-farm — but the day the fee is gone, it has nothing further to say about splitting. CIP-187 keeps penalizing fragmentation at the formula level, permanently. It’s defense in depth: even in a post-fee-reform world, the math still rewards concentration over splitting.
So, to the question the screenshots provoke — if CIP-187 only erases a sliver of the multipool surface, why bother? — the answer is that the surface is the wrong scoreboard. CIP-187 is the only proposal in the family that:
- removes the dilution that punishes delegators for filling good pools,
- makes per-delegator ROA rise with delegation instead of fall,
- adds a structural penalty on splitting that outlives fee reform,
- lowers the cold-start barrier, so a small honest pool can attract delegation as it fills instead of needing whale-tier pledge on day one,
- and gives a large, rational delegator — a cooperative like Pool Ranger — real protocol-level value to bring to the table, because the utilization it delivers is now something the formula actually rewards.
The real conclusion of the simulator
Set the three families side by side and the picture is not “fee reform wins.” It’s that each CIP guards a different door:
- CIPs 23 / 74 / 75 (fee reform) — close the fee-farm. The biggest, most visible region on the cube, because the cube is measuring fees. But removing the flat fee alone leaves the other two doors wide open.
- CIP-50 (leverage cap) — close the Sybil door: reward-earning stake must scale with pledge, so controlling stake costs capital in proportion to the stake controlled.
- CIP-187 (utilization-scaled pledge bonus) — close the dilution door: align the delegator’s incentive with the protocol’s, so good pools fill instead of being punished for filling — and splitting stays penalized whether or not a fee exists.
The simulator isn’t crowning a winner. It’s showing — in three dimensions — that these reforms are not rivals. They are complementary, and the protocol is healthiest with all three. That is exactly why All three families is the preset where the multipool parameter space finally, and completely, goes dark.
