diff --git a/docs/index.mdx b/docs/index.mdx
index 4ebc0460..3f22509b 100644
--- a/docs/index.mdx
+++ b/docs/index.mdx
@@ -15,9 +15,9 @@ import GridBlock from "@site/src/components/GridBlock/GridBlock.tsx";
 
 If MEV remains centralized, we cannot realise the promises that crypto offers.
 
-SUAVE unifies coordination in a decentralized way. SUAVE does not compete with other blockchains: it is intended to aggregate bundles of transactions and coordinate the building of blocks that ultimately change the state of other chains.
+SUAVE unifies how we coordinate without forcing any one mechanism onto everyone. SUAVE does not compete with other blockchains: it is intended to aggregate and coordinate all the things that ultimately change the state of other chains.
 
-SUAVE is a permissionless and open marketplace for mechanisms. SUAVE enables anyone to innovate on auctions and other mechanisms for managing orderflow or filling blockspace, which enables us to (among other things) _decentralize block building_ for all blockchains. Such "MEV applications" can provide a sustainable [defender's advantage](https://vitalik.ca/general/2016/12/29/pos_design.html) against centralizing economic forces both within and beyond crypto. 
+SUAVE is a permissionless and open marketplace for many different mechanisms. SUAVE enables anyone to innovate, design, and deploy new mechanisms, which enables us to (among other things) _decentralize block building_ for all blockchains. Such "MEV applications" can provide a sustainable [defender's advantage](https://vitalik.ca/general/2016/12/29/pos_design.html) against centralizing economic forces both within and beyond crypto.  
 
 Reducing the economic centralization pressure on other networks enables:
 
diff --git a/docs/reference/builder-solidity/index.mdx b/docs/reference/builder-solidity/index.mdx
index 0a567df8..ecf759a0 100644
--- a/docs/reference/builder-solidity/index.mdx
+++ b/docs/reference/builder-solidity/index.mdx
@@ -14,7 +14,7 @@ import AlignItems from "@site/src/components/AlignItems/AlignItems.tsx";
 
 Builder solidity is solidity with some additional precompiles intended to support MEV applications. Using these precompiles, builder solidity contracts can define how SUAVE nodes run computation, without that computation being done on chain. That is, SUAVE nodes can accept encrypted data from users, do stuff with it privately and only reveal the results, not the inputs.
 
-## Why is this cool?
+## Verifiable logic, private data
 
 Builders have no incentive to share the algorithms by which they currently build blocks, so it is hard for searchers to know whether their bundles will be built in the order they submit them, or manipulated. This means that searchers are more likely to send their bundles to the few builders they trust, or only those who consistently build winning blocks (in order to avoid leaking alpha to everyone). This results in block builder centralization over time.
 
diff --git a/docs/reference/builder-solidity/worked-examples/mev-share.mdx b/docs/reference/builder-solidity/worked-examples/mev-share.mdx
index 4f0d32cc..51ac9ba9 100644
--- a/docs/reference/builder-solidity/worked-examples/mev-share.mdx
+++ b/docs/reference/builder-solidity/worked-examples/mev-share.mdx
@@ -20,7 +20,6 @@ What do I need to do, and what happens next? We'll describe the steps first in n
 
 1. I take my signed ETH L1 Uniswap tx, and encrypt it using the public key of a specific SUAVE node.
     1. I decide whether I want all my transaction information in there, or just - for instance - the address of the pool contract I am trading on, but not the direction of the trade. 
-        1. We call this Ethereum L1 transaction a "hint", because it need not reveal _everything_ an ordinary L1 transaction would.
     3. I wrap my L1 transaction in another transaction to the MEVShare.sol on SUAVE.
 2. I send the transaction in (1.ii) to the SUAVE node I encrypted it for.
 3. The SUAVE node sees this transaction, sees that it contains additional data (the encrypted L1 transaction), and passes it to the MEVM running in that node to process confidentially.
diff --git a/docs/reference/index.mdx b/docs/reference/index.mdx
index 1bd4e6c8..2a4436d3 100644
--- a/docs/reference/index.mdx
+++ b/docs/reference/index.mdx
@@ -29,7 +29,15 @@ The high level plan is:
 3. **Decentralized execution**
     - geographically and administratively, using threshold MPC, FHE etc as they become viable
 
-Technically, we will achieve this with the "MEVM": our adaption of the EVM with a new runtime and APIs. These APIs enable you to leverage new precompiles in what we call "builder solidity", which is solidity as you know and love it, but extended so that you can create MEV applications. 
+There are many reasons why we think the market will evolve toward many chains sharing a unified sequencing layer which is specifically intended to keep power meaningfully decentralized in each specific chain:
+
+1. Block builders who only operate on a single domain will find themselves increasingly disadvantaged due to cross-domain MEV.
+2. There are efficiency gains for users from aggregating and clearing their preferences inside the same auction.
+3. The credible neutrality of an open marketplace in which many parties share their views, strategies, and opinions gives SUAVE an information advantage on centralized builders.
+4. Enabling computation on sensitive data (user orderflow) in a permissionless setting is hard. By solving it once, we can amortize the cost across the ecosystem and provide better solutions more cheaply than any individual participant could.
+5. Because of how fundamental transaction sequencing is in blockchains, only another decentralized system can provide the necessary security and credible neutrality.
+
+Building unified infrastructure for expressing diverse values is the best way to give domains control over their own validation guarantees and to ensure smaller domains stay decentralized in the face of centralizing economic forces outside their control.
 
 ## *Refers to notes
 
diff --git a/docs/sidebars.js b/docs/sidebars.js
index b4e6ffdf..aece1560 100644
--- a/docs/sidebars.js
+++ b/docs/sidebars.js
@@ -58,7 +58,7 @@ module.exports = {
       link: {type: 'doc', id: 'understand/index'},
       items: [
         'understand/meaning',
-        'understand/fairness',
+        'understand/welfare',
         'understand/power'
       ],
     },
diff --git a/docs/understand/fairness.mdx b/docs/understand/fairness.mdx
deleted file mode 100644
index a4e6fe15..00000000
--- a/docs/understand/fairness.mdx
+++ /dev/null
@@ -1,86 +0,0 @@
----
-title: Fairness
-description: We define fairness as 'the maximum payoff received by any individual is not so much greater than the average welfare of every individual', and show how this definition can apply to communicating information effectively in distributed systems.
-keywords:
-  - understand
-  - suave
-  - theory
-  - fairness
----
-
-import Video from "@site/src/components/Video/Video.tsx";
-import AlignItems from "@site/src/components/AlignItems/AlignItems.tsx";
-
-# To be fair
-
-1. How can we ensure that people can participate in the value they create by virtue of the way information about their intentions is acted on? 
-
-    - _We can build systems that account for the time it takes to communicate price._
-
-2. How can systems account for the time it takes to communicate price? 
-    
-    - _By cultivating shared, permissionless, open marketplaces in which it is profitable to make credible commitments about how you will act on information you receive._  
-
-3. Will such systems be fair? 
-
-    - _If we define fairness as "the maximum payoff received by any individual is not much greater than the average welfare of every individual", then an open marketplace can help us discover more fair orderings because people are more likely to share their information with those who give them the greatest reward, which can only be done by those who both maximize revenue **and** distribute it well._
-
-4. Is fairness a worthwhile goal? 
-
-    - _Intuitively, yes. Logically, we can ask, what does "fairness" really mean? If we extend the ideas on the [previous page](/understand/meaning), a fair environment could be one in which **your capacity to make meaning matches your buying power**._
-
-# Cui bono?
-
-The second claim we'll make in this section is that **we should anticipate how the media we use move messages so that no-one is unfairly manipulated**. We've already looked at how we communicate meaning and value, now we will turn to the time it takes for communication to occur.
-
-Blockchains can be described as verifiable commitment devices. Such devices enable us to [craft prosocial coordination games](https://www.kernel.community/en/learn/module-1/dreamers/#an-infinite-stock-of-games), yet they do not totally remove uncertainty about the payoffs available to different actors, which can skew the outcome of any mechanism we use to discover price, making it unfair. This is because there is some period of time between each block, and so we cannot make commitments and achieve consensus about them (i.e. make them credible) within one block.
-
-This game-theoretic fact about intra-block time introduces a whole spectrum of uncertainty about the payoffs associated with actions taken on information received between blocks. Tarun Chitra explores this thoroughly in the video below:
-
-<Video src="https://www.youtube-nocookie.com/embed/WYH7n4M016A?start=14626" title="Towards a Theory of MEV: Uncertainty" />
-
-One of the key questions Tarun poses in this presentation is, "How do we **achieve a balance** between maximizing revenue (which is good for the stability of systems, especially if they rely on arbitrage), and making sure that the expected welfare for all participants - on average - is not so far away from the maximum payoff."
-
-In order to analyze this question, Tarun starts by looking at fairness in terms of specific applications, rather than the system as a whole. In particular, he shows that applications all have a payoff function, which depends on a set of permutations (i.e. how you order transactions). Such payoff functions can be "smooth" (i.e. the order of transactions doesn't lead to significant differences in the payoff available if there is sufficient liquidity). He proved this in the context of AMMs, where the maximum and expected payoff are only separated multiplicatively by a factor of `log(n)`.
-
-However, payoff functions in the context of _liquidations_ are discontinuous. If the price drops below a certain point, everything happens, but until such time, nothing happens. So, if you can permute the transactions from a specific application such that the order results in the price dropping below the point at which a liquidation occurs, then you can achieve payoffs which differ significantly from the average. This is unfair, by our definition. It also means that liquidations are the fundamental basis for payoffs: any payoff functions in blockchains can be written as a sum of liquidation games. It's worth reflecting deeply on this point:
-
-> **People taking leverage serves as the basis for the set of all payoffs**. 
-
-Moreover, the maximum and expected payoff in the context of liquidation games is separated multiplicatively by a factor of `n!`. This is a much bigger problem if we're trying to craft fair systems that do not advantage those who can act on price information they receive first.
-
-# Fairier transforms
-
-Tarun asks another question, based on the above: "Is there a way in which, given a payoff function, we can determine the difference between the average and maximum payoff such that we can provide some guarantees on fairness?" He suggests that an analogue of Fourier transforms can assist us do exactly this.
-
-If all payoffs are the sum of liquidation games, we can decompose them in the same way we can separate out different frequencies in a complex waveform. The "high frequency" waves are equivalent to sets of transactions in which there are very few permutations that can trigger a liquidation. The "low frequency" waves are equivalent to sets of transactions where many different permutations can trigger a liquidation. This allows us to analyze how constrained we are in realising a particular payoff (lots of high frequencies `=` constrained, lots of low frequencies `=` not constrained). 
-
-We can go further by looking at the Fourier coefficient. If a set of permutations is large, and the Fourier coefficient over that set is large, then the value of the payoff function approaches the average across all permutations. If the set is small, but has a very large coefficient, it approaches the maximum value. Tarun also introduces the Uncertainty Principle which, in this case, states that if you make a function really sharp (highly specified/exact), the Fourier transform of that function cannot also be sharp. This leads us to three facts:
-
-1. Liquidations generate all payoff functions
-2. The Fourier transform provides a way of going from a given payoff to that payoff represented in the basis of all liquidations
-3. The Uncertainty Principle provides a bound on the expected value over the maximum value
-
-These three facts together give us an explicit way to quantify fairness!
-
-Note that the "fairness" we are quantifying here is premised on a very context-specific definition: [there is no universal definition of fairness](https://www.youtube.com/live/krlAqKsdLkw?si=i-L_OZFQTHYJ7y1C&t=1439), just as there is no "ultimate meaning", which is why MEV will always exist to some degree. However, if each application can order transactions based on the above general mathematical insights above applied to their specific context, we can meaningfully reduce MEV.
-
-In practice, this means that to avoid maximally unfair payoffs, we need to have a sufficiently large set of possible orderings _and_ there needs to be a large degree of overlap between the possible orderings and the Fourier coefficient. One way to achieve this is an **open marketplace for mechanisms** at the application level. 
-
-# Generally suave
-
-This is exactly what SUAVE is: an open marketplace for applications that determine specific ways of ordering transactions, such that we can provide approximate fairness guarantees _in general_, i.e. across any new use case that occurs indefinitely into the future.
-
-It's important to understand the _general_ nature of SUAVE. As we have stressed, MEV is a general phenomenon which arises from the time it takes to communicate. People have been aware of this for centuries, but no-one has attempted to build systems that can account for it, because no-one had the sort of verifiable commitment devices (blockchains) we enjoy.
-
-Bitcoin represents a new medium with/in which we can coordinate the creation and distribution of money. Ethereum represents a new medium with/in which we can coordinate _in general_. However, the commitments this requires are public, which means that the information they contain can be manipulated in the time it takes to verify a new block, creating unfair outcomes for those seeking to coordinate.
-
-**SUAVE represents a new medium with/in which we can coordinate how everyone coordinates**. SUAVE is the metacognition for every other coordination system. And we do mean every other coordination system: everything from blockchains to traditional exchanges to Google AdWords suffers from some version of MEV.
-
-If you reflected on that fact that people taking leverage serves as the basis for all payoff functions, you might begin to sense how deep this runs. Derivatives are weapons of mass destruction not because of some intrinsic flaw in their own structure, but because the market mechanisms used to trade them cannot account for the way that liquidations skew payoff functions. SUAVE fixes this.
-
-<AlignItems horizontal="center" vertical="center" sideMargin={0}>
-<div>
-  <img alt="Fairier Transformers" src={require('/static/img/fairier-transformers.png').default} />
-</div>
-</AlignItems>
\ No newline at end of file
diff --git a/docs/understand/index.mdx b/docs/understand/index.mdx
index 71afe55b..149334d8 100644
--- a/docs/understand/index.mdx
+++ b/docs/understand/index.mdx
@@ -13,14 +13,14 @@ Maximum Extractable Value (MEV) is a general phenomenon. It can occur anywhere i
 
 One of the [first known exploits in the modern era took place in 1834](https://www.schneier.com/blog/archives/2018/05/1834_the_first_.html) between Bordeaux and Paris. Communication is never instantaneous (leaving aside quantum entanglement for now), and those who receive information first have an advantage because they can act prior to anyone else, giving them greater ability to extract value from it.
 
-There is a deep link between time, information, value, meaning, and power. If we investigate the intersection between informtation theory and economics, we can make a case that [information is the new money](https://youtu.be/vi-rVTFTb6s). Money is a _verifiable record_ for the quantification, storage, and/or exchange of value just as information quantifies, stores, and/or communicates meaning.
+There are deep links between time, information, value, meaning, and power. If we investigate the intersection between information theory and economics, we can make a case that [information is the new money](https://youtu.be/vi-rVTFTb6s?t=675). Money is a _verifiable record_ for the quantification, storage, and/or exchange of value just as information quantifies, stores, and/or communicates meaning.
 
 Moreover, time seems to be like money because there is a duration between receiving information, assessing its value, and constructing meaning. Duration advantages those who receive information first, or can act on it most quickly. Advantage means that power accrues to more sophisticated experts over time, often in the form of more money. If the way we communicate–and the media we use–do not account for this in their design then they tend to be unfair, resulting in the centralization of power irrespective of how they function.
 
 We seek to understand MEV from first principles and so we will assess the above outline through three different lenses:
 
 1. [**Communication theory**](/understand/meaning) (valuable information and meaningful communication).
-2. [**Mathematics**](/understand/fairness) (what is actually fair?)
+2. [**Mathematics**](/understand/welfare) (what is welfare?)
 3. [**Politics**](/understand/power) (power and the problems of the public)
 
 ## Goals
@@ -31,30 +31,20 @@ We want to study MEV from first principles, but with less proofs and more relata
 
 > This section does not present answers, or the "right" way of thinking about MEV. It presents interesting and different ways to think about what MEV really is and why it matters.
 
+For instance, MEV can also be characterised as a phenomenon that occurs if there is an ["overlap in user preferences"](https://www.youtube.com/watch?v=l_gkFntA1cA&t=1126) and some actor responsible for allocating resources to those overlapping preferences. This lens may lead us to more specific behavioural economic models than the framework described above, and we welcome any explorations in that direction. You will likely end up investigating [how power functions](/understand/power) in either case, but the different trail you blaze to get there will be valuable to us all.
+
 ## SUAVE specifics
 
-If you read the [welcome](/) page, you will see that SUAVE is specifically intended to create an open marketplace for mechanisms. 
+SUAVE is specifically intended to create an open marketplace for mechanisms. 
 
-Well designed mechanisms for the exchange of information can help ensure that honest actors profit more than dishonest ones, aligning our incentives around shared ownership rather than private economies of scale.
+Well designed mechanisms for the exchange of information can help ensure that rational actors profit more, aligning our incentives around shared ownership rather than private economies of scale.
 
-SUAVE is our attempt to empower users and maximally decentralize public blockchains by providing "unified infrastructure for diverse expressions of value". We ask that you discuss the validity of this phrase through the three core claims we'll make in this section:
+SUAVE is our attempt to empower people and maximally decentralize public blockchains by providing "unified infrastructure for diverse expressions of value". We ask that you discuss the validity of this phrase through the three core claims we'll make in this section:
 
 1. People should be able to participate in the value they create by virtue of the way information about their intentions is acted on.
 2. To do this, we must anticipate how our messages move through our media so that no-one is unfairly manipulated.
 3. This amounts to programming money (and the power associated with it) in increasingly democratic ways.
 
-## General application
-
-There are many reasons why we think the market will evolve toward many chains sharing a unified sequencing layer which is specifically intended to keep power meaningfully decentralized in each specific chain:
-
-1. Block builders who only operate on a single domain will find themselves increasingly disadvantaged due to cross-domain MEV.
-2. There are efficiency gains for users from aggregating and clearing their preferences inside the same auction.
-3. The credible neutrality of an open marketplace in which many parties share their views, strategies, and opinions gives SUAVE an information advantage on centralized builders.
-4. Enabling computation on sensitive data (user orderflow) in a permissionless setting is hard. By solving it once, we can amortize the cost across the ecosystem and provide better solutions more cheaply than any individual participant could.
-5. Because of how fundamental transaction sequencing is in blockchains, only another decentralized system can provide the necessary security and credible neutrality.
-
-Building unified infrastructure for expressing diverse values is the best way to give domains control over their own validation guarantees and to ensure smaller domains stay decentralized in the face of centralizing economic forces outside their control.
-
 ## Why?
 
 This section is structured as above because MEV is a **general phenomenon**. Much of the activity around MEV in blockchains so far has been characterised by bots built to exploit short-term opportunities.
diff --git a/docs/understand/meaning.mdx b/docs/understand/meaning.mdx
index e32cbed2..aaf5933f 100644
--- a/docs/understand/meaning.mdx
+++ b/docs/understand/meaning.mdx
@@ -13,7 +13,7 @@ import Video from "@site/src/components/Video/Video.tsx";
 
 # The meaning of it all
 
-We established in the [introduction](/understand/index) that MEV is a general phenomenon which arises from the time it takes to communicate. If communication is not instantaneous, then those who receive information first can act on it prior to anyone else. 
+We established in the introduction that MEV is a general phenomenon which arises from the time it takes to communicate. If communication is not instantaneous, then those who receive information first can act on it prior to anyone else. 
 
 Why is the ability to act on information first more valuable?
 
@@ -32,7 +32,7 @@ There are at least three important points to note about the previous sentence.
 
 The reason all of this matters is that the meaning we make of/from any given information is inseparable from the [value](https://www.kernel.community/en/learn/module-1/value) we assign it. In order to understand what is actually being extracted, we must first perceive how meaning is made (and therefore how it can continually be remade). This is not just a philosophical exercise: if we refine our understanding of meaning, we can more readily identify relevant information and valuable means of acting on it at different times.
 
-Metaphor is the primitive through which we make meaning. We equate one thing with another, and so perceive the value of it relative to some prior experience/perception. Metaphor is not just a literary phenomenon. **E = mc<sup>2</sup>** is a metaphor. The concept "electron" is a metaphor. We chain these metaphors together in order to create shared understanding: "data" is a metaphor for unstructured bits; "information" is a metaphor for structured bits; "knowledge" is a metaphor for interpreted bits; "wisdom" is a metaphor for understood/realised bits. "Bit" is also a metaphor, though one which lacks a stable referent, because metaphor - just like the meanings we construct with it - is itself a strange loop.
+Metaphor is the primitive through which we make meaning. We equate one thing with another, and so perceive the value of it relative to some prior experience/perception. Metaphor is not just a literary phenomenon. **E = mc<sup>2</sup>** is a metaphor. The concept "electron" is a metaphor. We chain these metaphors together in order to create shared understanding: "data" is a metaphor for unstructured bits; "information" is a metaphor for structured bits; "knowledge" is a metaphor for interpreted bits; "wisdom" is a metaphor for understood/realized bits. "Bit" is also a metaphor, though one which lacks a stable referent, because metaphor - just like the meanings we construct with it - is itself a strange loop.
 
 Strangeness does not mean we cannot learn to apply our understanding in practical ways. In particular, we can notice our dependence on time and interaction when making meaning, and the way this links to the value we can extract from receiving any piece of information first. 
 
@@ -48,7 +48,9 @@ However, as the old adage goes: "Price is what you pay. Value is what you get."
 
 Does programmable money provide a means of addressing this problem? We know that [systems which only have win-lose transactions](https://github.com/norvig/pytudes/blob/main/ipynb/Economics.ipynb) amplify inequality, and we're looking for ways to quantify and distribute the added value of information as it is ordered by the people who receive it first. Essentially, we're trying to program money such that we can participate in win-win transactions, where the shared wins come from accounting for the time it takes to communicate price.
 
-How can we do this? SUAVE is one option: create an open marketplace for mechanisms that enables anyone to develop applications which strive for a balance between maximizing revenue and returning that revenue to the people whose information created it.
+What does this look like in practice? A win-lose transaction is easy to understand: if my option is "in the money", then the person on the other side is losing money. I win, they lose. However, if a searcher finds a profitable backrun transaction based on information I share with them about my trade, it is a win for that searcher, but it is also (directly) a win for me because some of the additional revenue generated from aggregated information about all transactions is shared with me, **and** it is (indirectly) a _win for everyone_ because we get to transact in a more efficient system. This is new and not possible in the traditional system.
+
+SUAVE is a unified environment for the expression of win-win values. That is, without permission, anyone can create mechanisms which both create the greatest win for themselves while _simultaneously_ ensuring that the people whose information made their win possible also win. The incentive in such an environment is to maximise both efficiency and welfare simultaneously. 
 
 If the systems we create do not find this balance, then those who can access information first interpret it in ways which assume that they deserve payoffs significantly different than the average. We may not consciously agree to this, but every time we take a loan, or buy a stock, or interact with money in general, we give our implicit agreement. In this, we get stuck playing unbalanced games in which we are [both victims and perpertrators](https://www.kernel.community/en/conversation/power-less#defenders-advantage), all hoping we can manipulate things such that the lack of balance works in our favor.
 
diff --git a/docs/understand/power.mdx b/docs/understand/power.mdx
index 423dc55f..ea27b66f 100644
--- a/docs/understand/power.mdx
+++ b/docs/understand/power.mdx
@@ -54,7 +54,7 @@ Such contracts might help us “perfect the means of communication of meanings s
 
 That sounds nice, but what does it actually mean to perfect the communication of meanings? In our context, it is nothing other than enabling people to make credible commitments, while crafting a system which can order responses to those commitments in ways that simultaneously create the most collective value _and_ distribute it fairly. 
 
-The mechanisms that do this generally look like auctions, which discover price in uncertain environments, and result in the exchange of information (“meanings”) for commitments (“payments”). Most importantly, the method that the auction uses to determine the price can be transparently and publicly encoded in a builder solidity contract, though the actual bids individual users submit need not be (indeed, _must_ not be for optimum functionality).
+The mechanisms that do this generally look like auctions, which discover price in uncertain environments, and result in the exchange of information (“meanings”) for commitments (“payments”). Most importantly, the method that the auction uses to determine the price can be transparently and publicly encoded in a builder solidity contract, though the actual bids individual users submit need not be.
 
 Public process with private inputs is a _fundamentally democratic idea_. It may be the only concept capable of defending true democracy in the modern age: verifiable mechanisms for collective decisions; privacy for the individuals who contribute to the decision-making process. This is not a new idea. In fact, Dewey says exactly the same thing: 
 
@@ -110,7 +110,7 @@ This is one reason why it is so impactful to work on mitigating MEV. Given their
 
 Working on MEV is how we participate in a public which transcends any particular political agenda such that we can maintain the meta-political agenda of permissionlessness. That is, the realpolitik of the public created by MEV is just this: any power structure is potentially acceptable so long as it does not preclude the ability of anyone to suggest and implement an alternative.
 
-Traditional politics contests power via the application of either violence or patience. Permissionless marketplaces for mechanisms which order transactions such that their indirect consequences change do not require either protest or occupation. The use of any given mechanism is not determined by rhetoric or personality: it is determined by efficiency and fairness. As Dewey says,
+Traditional politics contests power via the application of either violence or patience. Permissionless marketplaces for mechanisms which order transactions such that their indirect consequences change do not require either protest or occupation. The use of any given mechanism is not determined by rhetoric or personality: it is determined by efficiency and welfare. As Dewey says,
 
 > _“Democracy, then, entails a kind of openness in which its substantive meaning—that is, what concerns it addresses and what ends it pursues—is always in the process of being determined._
 
@@ -144,7 +144,7 @@ Our mission is to illuminate, democratize, and distribute. Just like Dewey, we u
 
 Such harmonic operation depends on the deliberate inquiry of each citizen into the kinds of transactions whose effect indirectly influences their lives, and the associated search for others similarly affected with whom it is possible to create solutions. Working on MEV–and using SUAVE–is nothing if not a deep and deliberate inquiry into the indirect consequences of transactions and a search for the others striving to defend humane and meaningful value(s). 
 
-This deliberate inquiry has led us to try and quantify fairness, and understand the relationship between time, communication, value, and meaning. Next, it will require that we consider how to engineer low-latency networks that preserve privacy. These three inquiries combined might help us realize the ultimate vision presented in _The Public and Its Problems_:
+This deliberate inquiry has led us to try and quantify welfare, and understand the relationship between time, communication, value, and meaning. Next, it will require that we consider how to engineer low-latency networks that preserve privacy. These three inquiries combined might help us realize the ultimate vision presented in _The Public and Its Problems_:
 
 > _“The highest and most difficult kind of inquiry and a subtle, delicate, vivid and responsive art of communication must take possession of the physical machinery of transmission and circulation and breathe life into it. When the machine age has thus perfected its machinery it will be a means of life and not its despotic master. Democracy will come into its own, for democracy is a name for a life of free and [enriching communion](https://www.kernel.community/en/learn/module-7/giving).”_
 
diff --git a/docs/understand/welfare.mdx b/docs/understand/welfare.mdx
new file mode 100644
index 00000000..9cb0eda5
--- /dev/null
+++ b/docs/understand/welfare.mdx
@@ -0,0 +1,86 @@
+---
+title: Welfare
+description: Fairness may occur when 'the maximum payoff received by any individual is not so much greater than the average welfare of every individual'. This definiton can help us quantify how efficiently we communicate value in distributed systems.
+keywords:
+  - understand
+  - suave
+  - theory
+  - fairness
+---
+
+import Video from "@site/src/components/Video/Video.tsx";
+import AlignItems from "@site/src/components/AlignItems/AlignItems.tsx";
+
+# To be fair
+
+1. How can we ensure that people can participate in the value they create by virtue of the way information about their intentions is acted on? 
+
+    - _We can build systems that account for the time it takes to communicate price._
+
+2. How can systems account for the time it takes to communicate price? 
+    
+    - _By cultivating shared, permissionless, open marketplaces in which it is profitable to make credible commitments about how you will act on information you receive._  
+
+3. Will such systems be fair? 
+
+    - _If we define fairness as "the maximum payoff received by any individual is not much greater than the average welfare of every individual", then an open marketplace might increase welfare because people are more likely to share their information with those who give them the greatest reward, which can only be done by those who both maximize revenue **and** distribute it well._
+
+4. Is welfare a worthwhile goal? 
+
+    - _Both 'fairness' and 'welfare' are laden terms with complex histories. However, using our quantifiable definition of fairness, we can instead ask, what do we expect the result of greater welfare to be? Our suggestion is that optimizing for welfare requires creating open and contestable power structures. Power is what this whole section revolves around, and quantifying welfare might help us test how [decentralized power is in any given system (next page)](/understand/power), as well as whether [your individual capacity to make meaning matches your buying power (previous page)](/understand/meaning)._
+
+# Cui bono?
+
+The claim we'll make on this page is that **we should anticipate how our messages move through the media we use so that no-one is unfairly manipulated**. We've already looked at how we communicate meaning and value, now we will turn to the time it takes for communication to occur.
+
+Blockchains can be described as verifiable commitment devices. Such devices enable us to [craft prosocial coordination games](https://www.kernel.community/en/learn/module-1/dreamers/#an-infinite-stock-of-games). However, credible commitments do not totally remove uncertainty about the payoffs available to different actors, which can make it difficult to quantify whether a given application is fair in the sense we have defined it above. Uncertainty about payoffs largely stems from the period of time between each block in wich we cannot make commitments and achieve consensus (i.e. make them credible).
+
+Intra-block time actually introduces a whole spectrum of uncertainty about the payoffs associated with actions taken on information received between blocks. Tarun Chitra explores this thoroughly in the video below:
+
+<Video src="https://www.youtube-nocookie.com/embed/WYH7n4M016A?start=14626" title="Towards a Theory of MEV: Uncertainty" />
+
+One of the key questions Tarun poses is, "How do we **achieve a balance** between maximizing revenue (which is good for the stability of systems, especially if they rely on arbitrage), and making sure that the expected welfare for all participants - on average - is not so far away from the maximum payoff?"
+
+In order to analyze this question, Tarun starts by looking at welfare in the context of specific applications, rather than the system as a whole. In particular, he shows that applications all have a payoff function, which depends on a set of permutations (i.e. how you order transactions). Such payoff functions can be "smooth" (i.e. the order of transactions doesn't lead to significant differences in the payoff available if there is sufficient liquidity). He proved this in the context of AMMs, where the maximum and expected payoff are only separated multiplicatively by a factor of `log(n)`.
+
+However, payoff functions in the context of _liquidations_ are discontinuous. If the price drops below a certain point, everything happens, but until such time, nothing happens. So, if you can permute the transactions from a specific application such that the order results in the price dropping below the point at which a liquidation occurs, then you can achieve payoffs which differ significantly from the average. This is unfair, by our definition. It also means that liquidations are the fundamental basis for payoffs: any payoff function can be written as a sum of liquidation games. It's worth reflecting deeply on this point:
+
+> **People taking leverage serves as the basis for the set of all payoffs**. 
+
+Derivatives really are weapons of mass destruction. Tarun's proof - which shows that the maximum and expected payoff in the context of liquidation games is separated multiplicatively by a factor of `n!` - is powerful becauseit explains **why** derivatives are destructive. Everyone knows leveraged positions are more risky. However, when we look through the lens of MEV, we can see that taking a leveraged position creates systemic risk because it increases the chance that the payoff sophisticated users can get will differ extremely from the average, which shifts the global incentive structure toward private ownership (manipulation for individual profit) and away from shared stewardship (reciprocity for mutual flourishing).
+
+# Fairier transforms
+
+Tarun asks another question, based on the above: "Is there a way in which, given a payoff function, we can determine the difference between the average and maximum payoff such that we can provide some guarantees on fairness?" He suggests that an analogue of Fourier transforms can assist us do exactly this.
+
+If all payoffs are the sum of liquidation games, we can decompose them in the same way we can separate out different frequencies in a complex waveform. The "high frequency" waves are equivalent to sets of transactions in which there are very few permutations that can trigger a liquidation. The "low frequency" waves are equivalent to sets of transactions where many different permutations can trigger a liquidation. This allows us to analyze how constrained we are in realising a particular payoff (lots of high frequencies `=` constrained, lots of low frequencies `=` not constrained). 
+
+We can go further by looking at the Fourier coefficient. If a set of permutations is large, and the Fourier coefficient over that set is large, then the value of the payoff function approaches the average across all permutations. If the set is small, but has a very large coefficient, the payoff approaches the maximum value. Tarun also introduces the Uncertainty Principle which, in this case, states that if you make a function really sharp (highly specified/exact), the Fourier transform of that function cannot also be sharp. This insight provides a means to bound the relationships between expected and average payoffs. 
+
+In review:
+
+1. Liquidations generate all payoff functions
+2. The Fourier transform provides a way of going from a given payoff to that payoff represented in the basis of all liquidations
+3. The Uncertainty Principle provides a bound on the expected value over the maximum value
+
+These three facts together give us an explicit way to quantify welfare!
+
+Note that the "welfare" we are quantifying here is premised on a very context-specific definition: [there is no universal definition of fairness](https://www.youtube.com/live/krlAqKsdLkw?si=i-L_OZFQTHYJ7y1C&t=1439), just as there is no "ultimate meaning", which is why MEV will always exist to some degree. However, if each application can order transactions based on the above general mathematical insights applied to their specific context, we can meaningfully reduce MEV.
+
+In practice, this means that to avoid maximally unfair payoffs, we need to have a sufficiently large set of possible orderings _and_ there needs to be a large degree of overlap between the possible orderings and the Fourier coefficient. One way to achieve this is an **open marketplace for mechanisms** at the application level. 
+
+# Generally suave
+
+This is exactly what SUAVE is: an open marketplace for applications that act on the information in transactions in verifiable ways, such that the incentives in the system tend towards general welfare and away from the centralization of power.
+
+It's important to understand the _general_ nature of SUAVE. As we have stressed, MEV is a general phenomenon which arises from the time it takes to communicate. People have been aware of this for centuries, but no-one has attempted to build systems that can account for it, because no-one had the sort of verifiable commitment devices (blockchains) we enjoy.
+
+Bitcoin represents a new medium with/in which we can coordinate the creation and distribution of money. Ethereum represents a new medium with/in which we can coordinate _in general_. However, the commitments this requires are public, which means that the information they contain can be manipulated in the time it takes to verify a new block, creating unfair outcomes for those seeking to coordinate.
+
+**SUAVE represents a new medium with/in which we can coordinate how everyone coordinates**. SUAVE is the metacognition for every other coordination system. And we do mean every other coordination system: everything from blockchains to traditional exchanges to Google AdWords suffers from some version of MEV.
+
+<AlignItems horizontal="center" vertical="center" sideMargin={0}>
+<div>
+  <img alt="Fairier Transformers" src={require('/static/img/fairier-transformers.png').default} />
+</div>
+</AlignItems>
\ No newline at end of file