pull down to refresh

tl;dr "Variance has always been the hidden cost of running a mining pool. The DATUM Layer turns that cost into a shared function of the network rather than a private liability, freeing pools from the constant drag of short-term debt."

I've had a hard time wrapping my head around DATUM and why people are excited about it. This article does a nice job of getting you from zero to having a basic understanding of the concept. It's pretty cool!

By Mark Artymko, President & Co-Founder, OCEAN

Article Published November 2025 in Mining Disrupt Magazine

FPPS and the Cost of PredictabilityFPPS and the Cost of Predictability



The first and most natural way miners reduce variance is by joining a pool. By combining hashpower, miners smooth out luck and turn random rewards into steadier income. Pooling itself is simple math: larger sample sizes mean more predictable results.

For miners who want even less variance, the industry has leaned on FPPS. Full Pay Per Share promises perfectly smooth payouts, with daily income guaranteed regardless of luck. But FPPS doesn’t remove variance; it just hides it. The pool operator absorbs the swings, paying miners from reserves or credit when luck runs dry.

That model quietly concentrates both power and financial exposure. The larger the pool, the smaller the relative variance, but the higher the absolute risk one party must carry. Smaller FPPS pools disappear or become functional proxies of larger companies that can absorb the risk. The hidden cost shows up in quiet fees, block scraping, or reward adjustments that miners rarely notice - the “invisible FPPS tax.”

There is no free lunch. FPPS buys predictability at the steep cost of less revenue. It replaces organic variance with synthetic stability and shifts risk from miners to operators.

Collaboration through DATUM takes a different approach. It reduces variance like a large pool but keeps control, decision-making, and financial exposure decentralized. Instead of one entity carrying the load, variance is shared across many, restoring balance between predictability and decentralization, and letting miners take home more of the work each produces.

Why Variance Creates CentralizationWhy Variance Creates Centralization



In mining, the Coefficient of Variance (CoV) can be estimated as:

1 ÷ √(blocks found per day × number of days in the period)

It’s a quick way to gauge how much a miner or pool’s luck can swing over time. A smaller CoV means steadier revenue and more predictable payouts. Large pools naturally have lower CoV because they find more blocks, which smooths their results faster. Smaller miners experience higher CoV, translating into bigger financial swings and greater working capital requirements.

The only way to reduce CoV is to increase sample size: in mining terms, combine hashrate. That basic math has quietly driven consolidation: small pools absorbed by larger ones until most new Bitcoin now flows through a handful of entities. Not because of better software, but because variance economics reward scale over independence.

DATUM: A Practical Path Back to DecentralizationDATUM: A Practical Path Back to Decentralization



In 2024, OCEAN introduced DATUM, developed by Jason Hughes, OCEAN’s Vice President of Engineering and former operator of the Eligius pool. DATUM was built so miners can create their own block templates locally while still participating in a shared network.

It is light, reliable, and already running in production. Since launch, hundreds of miners have adopted it and collectively found hundreds of blocks using OCEAN’s TIDES reward system, which distributes payouts based on real proof of work rather than synthetic insurance.

DATUM does more than return control to miners. It also points to a solution for variance itself.

Collaboration Without CentralizationCollaboration Without Centralization



Imagine miners and pools building their own blocks independently yet sharing rewards proportionally across everyone who contributes work. Each participant keeps full control of transaction selection and signaling while variance is shared among all.

That is the DATUM Layer: a network of independent subpools, working alongside solo miners, operating on their own terms but connected through a system that reduces variance for the entire group.

It is not another centralized pool. It is a coordination layer where miners stay sovereign and still gain stability.

For the first time, decentralization and predictability can exist in the same sentence. DATUM proves that mining does not need to choose between the two.

A New Era of Collaborative MiningA New Era of Collaborative Mining



The future of mining could look very different once the DATUM Layer reaches scale. Imagine a global fabric of independent pools and miners, each building their own blocks locally but all sharing variance collectively. Decision-making stays local, rewards are distributed by proof of work, and no single operator holds outsized financial power. The end result is a network where decentralization is not a slogan but a measurable outcome. Variance becomes a shared market function rather than a private burden. Pools are free to innovate again because survival no longer depends on the size of their balance sheet.

Putting Numbers to the VisionPutting Numbers to the Vision



So what does collaboration between pools actually look like in numbers?

To establish a baseline, consider Pool 1 in Figure 3 operating independently - a single pool running 10 EH/s in isolation with total network hashrate at 1,000 EH/s. That pool would expect to find about 1.6 blocks per day, or roughly 5.3 BTC in daily rewards. On paper that looks stable, but in practice variance can be wide.

CoV = (1 ÷ √(blocks found per day × number of days in the period)

For 10 EH/s Pool 1: Daily CoV = 83%, Weekly CoV = 31%, Monthly CoV = 15%

This means Pool 1 could experience luck swings of 83 percent or worse over a 24 hour period and the operator would need to cover roughly 4.4 BTC (about $528,000) out of reserves or debt to make FPPS payments to its miners. Over a bad week (31% CoV) that exposure compounds to 11.5 BTC, tying up $1.4 million in working capital. And over a bad month (15% CoV) that translates to 24.2 BTC, tying up $2.9 million in working capital. It is easy to see why only large, well-capitalized pools can afford to play this game.

Now, to show how collaboration between independent pools changes the math, let’s put real numbers to it.

Scenario 1 : DATUM Layer = 12% network hashrate

Assume the DATUM Layer reaches 12% adoption (OCEAN + participating sub-pools).

Running the same scenario inside the DATUM Layer, Consider Pool 2, also contributing 10 EH/s of its own hashrate as is Pool 1 but alongside other miners and pools, all totaling around 120 EH of collective collaborative hash rate.

CoV for the cumulative 120 EH/s DATUM Layer:

Daily CoV: 24%, Weekly CoV: 9%, Monthly CoV: 4%

The reserve requirements for Pool 2 is reduced to the following: Daily: 1.27 BTC ($153k), Weekly: 3.3 BTC ($400k), Monthly: 6.44 BTC ($774k) - a 73% reduction in required monthly working capital

Scenario 2 : DATUM Layer = 50% network hashrate

Taking the thought experiment further, imagine half the Bitcoin network, hundreds of independent pools and miners, collaborating through the DATUM Layer.

CoV for the cumulative 500 EH/s DATUM Layer:

Daily CoV: 11.8%, Weekly CoV: 4.4%, Monthly CoV: 2.1%

The reserve requirements Pool 2 is reduced even further to the following: Daily: 0.63 BTC ($75k), Weekly: 1.63 BTC ($196k), Monthly: 3.39 BTC ($406k) - an 86% reduction in required monthly working capital.

The bottom line: In the DATUM Layer, independent pools collaborate by sharing rewards proportionally while still building their own blocks. Risk drops dramatically as the network grows, making it viable to run a small pool or even solo mine.

The benefits of collaboration go beyond smoother payouts, they fundamentally change how capital and liquidity are managed:

The Cost of Money: Turning Variance into DebtThe Cost of Money: Turning Variance into Debt



Variance is not just a math problem, it is a liquidity problem. When an FPPS pool falls short of expected blocks, it still owes miners their daily payouts, and that shortfall must be financed. There is no free lunch. Someone is paying for everything.

Even if that liquidity is borrowed short term at 10% annual interest, the cost of money alone adds up. But the real hit is not the interest. It is the capital lock-up. The pool’s reserves or credit lines must absorb that gap until luck turns, sometimes for weeks.

Inside the DATUM Layer, those gaps shrink dramatically. Variance has always been the hidden cost of running a mining pool. The DATUM Layer turns that cost into a shared function of the network rather than a private liability, freeing pools from the constant drag of short-term debt.

Could This Be the Way Things Unfold?Could This Be the Way Things Unfold?



DATUM: Reducing Mining Variance Without the Steep Cost of Centralization will end up being more than a headline. It will be the blueprint for how mining evolves from here. The incentives are clear. Smaller pools gain stability without losing control. Larger pools gain resilience without surrendering market share. Everyone wins when risk is shared and power stays distributed. It feels like a no-brainer. The only real question is how quickly the industry will see it that way. Time will tell, but the path forward has never looked clearer.
some territories are moderated

Also FPPS pools are vulnerable to block withholding attacks.

reply

This is Steve Barbour's point, I think. I wish he would get more airtime. Someone should convince him to do an AMA here...

reply