The Prize Pool Network

1

The Prize Pool Network

PoolTogether has gone through four iterations. Each iteration has supported more assets:

  • The first version accepted Sai
  • The second version accepted Dai and USDC
  • The third version introduced a prize pool factory, so we had many prize pools for many different assets.

The latest version of PoolTogether, version four, significantly improves prize distribution. We can offer a virtually unlimited number of prizes. However, we currently only accept USDC.

We could create more prize pools, but the infrastructure is complicated. Doing so would be extremely cumbersome. PoolTogether still treats each asset separately: if you deposit USDC, you win USDC. Prize liquidity would not be shared between assets, so new assets would have to bootstrap their liquidity.

Protocols grow by accepting many different assets. USDC is often only about 20% of a protocol’s TVL. PoolTogether can unlock significant growth by accepting more types of assets. Users want to hold their favourite tokens; not just stables.

The Prize Pool Network unifies prizes across all prize pools; allowing us to offer a massive grand prize and easily on-board new assets.

This new model introduces three major changes:

  • A unified prize model: no matter what asset you deposit you will have a chance to win the same prize as everyone else.
  • Yield is sold for prizes: To ensure that we have prize liquidity denominated in a single token, the protocol will swap the yield tokens for prize tokens.
  • Odds are distributed by staking: Odds will be distributed across the prize pools through POOL staking.

Let’s dive in!

PoolTogether Primer

Before we begin, let’s review two important concepts: prize pools and prize distributors.

Prize Pools

A “prize pool” is a pool of assets deposited into a particular yield source.

Users deposit funds into prize pools to win prizes. Their chance to win is proportional to their deposit size. Prize pools can be entered and exited at any time. There are no fees, so users will always get back their full deposit. Prize pools are non-custodial (as before), so only users can move their funds.

Prize pools are specific to an asset and yield source. For example, we may have an Aave V3 USDC prize pool and an mStable USDC prize pool. We aren’t limited to stablecoins either; our prize pools could accept any token that generates yield, such as wBTC and wETH.

Deposits
Deposits2048Ă—1172 52.1 KB

Prize Distributors

Each day there is a prize “draw”. The draw encodes the prize distribution, which contains thousands of prizes. Users claim their draw prizes from the Prize Distributor. Prizes are all denominated in the same token, but there are differing tiers of prize rarity and size.

Prizes
Prizes1632Ă—1204 35.1 KB

Unified Prize Model

The prize pool network unifies the prize model such that users will be eligible to win the same prizes no matter what asset they deposit. Each prize pool will use the same Prize Distributor. This means that:

  • Prize liquidity is shared among all prize pools, allowing us to offer a single massive prize
  • Infrastructure is much simpler, allowing us to easily on-board new assets.

Single Massive Prize

By combining the yield from all prize pools we can have a single massive prize. For example, let’s say we’ve accrued yield across a variety of prize pools:

Asset USD Value of Accrued Yield
USDC $2000
DAI $1500
wETH $500
wBTC $350
CRV $80
Total $4430

Individually the pools wouldn’t have a very large prize. However, when combined together the prize is much more compelling.

Note: there is more than one prize. The $4430 would be broken up into tiers of different rarity and size.

Easily On-board Assets

Unifying the prizes simplifies the addition of new assets because there is no need to bootstrap prize liquidity. By allowing prize pools to tap into shared liquidity they get instant access to a prize. In the above example, we see that the CRV prize pool only generates $80 in yield. If we weren’t sharing prize liquidity, then that pool would only have an $80 prize. When the pool was first created it wouldn’t even have a prize!

Which Token to use for Prizes?

An open question for the new network is which token should be the prizes be denominated in? The arguments for a variety of prize tokens will be presented in a separate post. It’s important that we have an open discussion to discuss the merits of each token so that the PoolTogether community can come to a consensus and move forward.

Regardless of which token we use, the same mechanics will apply: users win prizes, yield is liquidated for prizes, and POOL holders stake to distribute odds.

In the rest of this post you’ll see the term prizes and prize tokens. This is intended to refer to whatever token we decide to use.

Yield is Sold for Prizes

We would like to denominate the prize in a single token, but we are accruing yield in all sorts of tokens. This is why we liquidate the yield tokens for prize tokens. The prize tokens are then fed back into the prize distributor.

Users will be able to buy the yield tokens, but we expect that eventually bots will consume any available arbitrage.

Edit: Liquidator Algorithm is now posted, including explanation, diagrams and graph

There will be a follow-up technical post that outlines the details of the liquidation algorithm.

YieldLiquidation
YieldLiquidation2416Ă—1282 82.3 KB

Distributing Prizes

No matter what asset a user deposits they will have a chance to win the same prizes. However, assets have different yields. Some assets will have a higher APR, while others will have a higher TVL. How can the protocol fairly distribute the odds of winning across deposits?

POOL token holders stake on prize pools to allocate odds. In return, the stakers earn rewards.

Staking
Staking1726Ă—1151 28.6 KB

Staking on Gauges

Each draw has a total number of picks. These picks form the “probability space” of the draw. Each pick is a chance to win. Technical details here.

Staking determines how picks are distributed across prize pools. The picks allocated to a prize pool are then split among the depositors in that pool.

The number of picks that a prize pool is allocated depends on the size of its stake:

fraction of picks = staked amount / gauge stake total

The gauge stake total is a number that is manually configured by governance. This number allows governance to ensure that the odds are fairly distributed across chains as well as across pools. If a prize pool gauge has 200 POOL staked on it on Optimism and a prize pool on Polygon has 100 POOL staked on it, then the Optimism prize pool will have twice as many picks as Polygon.

The gauge stake total also allows governance to control the prize frequency. By increasing the number, prizes will become less frequent as pools will receive a smaller fraction of the total picks. This gives governance the option of configuring very large prizes that occur infrequently. Imagine we have a $1,000,000 prize that occurs once per year! This will give us so much more freedom when designing the prize distribution.

Staking Rewards

When users stake on prize pool gauges they also receive rewards: stakers split a portion of the proceeds from yield sales for that prize pool. The portion for the stakers is the staker cut. Governance will manually set the staker cut globally across all liquidation pairs.

For example:

  1. Let’s say someone has staked 100 POOL on a prize pool. The total staked amount on that prize pool is 300, so they have 33.3% of the staked amount.
  2. While the user is staked, the prize pool liquidates yield and earns 1000 prize tokens.
  3. The staker cut is 10%, so 100 prize tokens are distributed to the stakers. The staker in question can claim 33.3 prize tokens

Like prizes, users will need to manually claim their staking rewards from time-to-time. We’ll make sure the action can be batched alongside withdrawal, and that they can claim for multiple pools.

Incentives Balance Odds

Staking rewards should naturally balance the odds. The more yield a pool gets, the higher the APR is for stakers. Assuming stakers seek the highest APR, the effective staker APR should be consistent across the prize pools. This means that the staked amount will be proportional to the total prize pool yield:

Assuming:

  • Prize: $1,000,000
  • Gauge Stake Total: $500,000,000
Asset Deposits (USD) APR Yearly Yield Staker Cut Staker APR Staked POOL Odds
ETH $5,000,000 0.38% $19,000 10.00% 5.20% $36,538 0.01%
wBTC $23,400,342 1.00% $234,003 10.00% 5.30% $441,516 0.09%
USDC $35,000,000 2.26% $791,000 10.00% 4.90% $1,614,286 0.32%
FEI $10,000,000 2.29% $229,000 10.00% 5.00% $458,000 0.09%
Total Odds: 0.51%
$1,273,003 Daily Prizes: $5,101

Notice how the Fei and wBTC prize pools have different TVL and APR, but the same odds; this is because their yield is approximately the same, but stakers will make sure their returns are optimized across the prize pools.

Flow of Funds

Let’s summarize the above changes in a flow of funds diagram:

FlowOfFunds
FlowOfFunds2415Ă—1901 151 KB

Here you can see how tokens flow through the protocol:

  1. Users deposit funds
  2. Funds generated yield
  3. Yield is liquidated for prizes
  4. Prizes are fed into the prize distributor, with a portion going to staking rewards
  5. Winners claim their prizes from the prize distributor.

V4 Migration

The existing PoolTogether V4 system can be migrated to support the prize pool network. The only changes required are to:

  • Swap out the Prize Distributor
  • Attach Yield Liquidators to the prize pools

Existing users will not need to move any funds; they can immediately take advantage of the new network.

The prize pool network requires the POOL token for staking, however. The POOL token must be securely bridged to chains on which prize pools reside. This is not the case with Avalanche, so it will not be supported at launch.

Rollout

PT Inc’s tentative rollout schedule:

Date Description
June 13 Testnet launch
June 27 Audit begins
July 11 Production Launch

In a few weeks PT Inc is going to roll out a prize pool network testnet. This will be an opportunity for the community to try out the new app. The testnet will allow us to work out any bugs and get a feel for the system. We’ll be able to smooth out any rough corners.

Once we’re happy with the contracts, we’ll send them off for audit. We’ll take a full weeklong audit for this new code.

With the smart contracts largely solidified, it’ll be time for us polish the app with the final design and develop our launch plan.

Finally, we launch!

Launch Plan

There are a number of questions that governance needs to decide on before our production launch. I will help coordinate the decision making as we get closer to launch.

In the meantime, it would be worthwhile to start thinking about:

  • Which asset should be the prize?
  • Which networks to deploy to (Optimism? Polygon?)
  • Which networks to shut down (Avalanche? Ethereum?)
  • What is the initial prize distribution? (we can do large infrequent prizes now)
  • Which assets to launch with (can be all of them if we like)

Which Asset Should be the Prize?

We still need to decide which asset we want to give away. There are basically two front-runners for the asset: POOL and USDC. More on this soon.

Which networks to deploy to?

The new prize pool network requires the POOL token. The POOL token is available on Polygon on Optimism, so it will be easy for us to launch on those chains.

However, the POOL token is not available on Avalanche. We will not be able to deploy the network to Avalanche.

Another aspect to consider is Aave V3: it has not been deployed to Ethereum. It seems having Ethereum L2s is now considered sufficient, so it may be enough to have Optimism.

Which networks to shut down?

In light of the above, perhaps we shut down the Avalanche and Ethereum prize pools.

What is the initial prize distribution?

The new network will allow us to have large, infrequent prizes. The launch is an opportunity to come out swinging with $100k+ prizes that occur once a quarter, or six months etc.

It would be worthwhile for the Treasury Working Group to start thinking about the prize distribution. Instead of considering how to divvy up the interest accrued over one day, we can consider much longer timelines. How could we divvy up interest accrued over six months?

What assets to launch with?

With the latest iteration of the prize pool network code, it’s becoming clear that we will incur very little overhead from launching new pools. We may just want to launch with support for all assets on Aave. Nonetheless, it’s worthwhile thinking about which assets we want to promote.

The assets that have POOL staked will be the ones with the highest odds of winning; so token holders can coordinate by staking on the best assets.

Summary

The new prize pool network will allow PoolTogether to scale indefinitely; incorporating new assets and chains without incurring significant overhead.

POOL token holders will be able to engage and govern the system in a decentralized way, and incentives will ensure the protocol is balanced.

We’re incredibly excited about this direction for the protocol, and believe this can take us to $1b+ TVL.

13 Likes
2

I think POOL should be the prize assuming it means the protocol would be buying back POOL and distributing. This would get us in a position where POOL has a lot of trade volume and should hopefully lead to some growth.

Deploy to Optimism and Polygon as well as plan for launch on Arbitrum and maybe ZK sync.

Agree with shutting down Ethereum and Avalanche, want to keep loss at a minimum and ETH gas fee’s are killer.

I’d like to see us shift back to having larger prizes overall and include some more rare jackpot sized prizes. TWG can definitely help with this analysis.

We should allow all assets on Aave V3 as it should remain up to the individual to determine what asset they are most comfortable with. If they deposit USDT they will get the same amount of USDT back, we are not responsible if USDT loses it’s peg.

7 Likes
3

Great Post, thank you @Brendan,

Can I suggest that you create a separate governance post for each one of your questions so that we can properly thread discussion/debate for each one?

2 Likes
4

Yes! I will follow-up with posts for each category of question. I’ll make sure to outline the options, as I see them, and offer some evaluation criteria.

2 Likes
5

I’m a bit worried about the parameters that governance has to set. Particularly, the gauge stake total feels so hard to get right, and even harder at launch.

Basically, it seems that the gauge stake total should be similar to the total amount of pool that is staked across the different gauges. If this value is higher than that, prizes will be more infrequent, of it’s lower, prizes will be more frequent.

Obviously, this has to be according to the prize distribution. If the prize distribution is distributing double the amount of prizes compared to what it’s being generated, then it intuitively makes sense that the gauge stake total should be double the amount of the total POOL staked in all the gauges.

The problem is that if a large amount of POOL is staked or unstaked in a short period of time, the calculations done to get a proper prize distribution will be in vain. This can lead to 2 different scenarios:

  1. The protocol distributes less prizes than expected if a big amount of POOL is staked. If this goes to an extreme, it can lead to people withdrawing their deposits because of low yields. Fortunately, people withdrawing will actually increase the expected yield and balance everything (avoiding an huge loss of TVL).

  2. The protocol distributes more prizes than expected if s big amount of POOL is unstaked. Although this scenario should help attract more TVL because of the higher prices, this would come at a cost to the treasury or the POOL token sell pressure (depending on the asset we choose to distribute).

I feel that our governance should act really quickly in these scenarios, which is something that with the current possibilities feels hard to do.

Should we have an emergency committee? (Kind of like the one that Curve protocol has). I don’t know if that’s necessary, but it’s something to think about.

2 Likes
6

The TWG will draft a formal response after all members review, but I have some preliminary, personal thoughts to share.

Governance

If we are to shut down the Ethereum deployment, then does that mean we have a cross-network governance solution? Because if that’s not the case, we’re completely divorcing the governance experience from the prize pool network. This will create friction if most of our users are on Polygon, Optimism, Arbitrum, etc. and hold no POOL balance on mainnet. Because governance becomes crucial in the new tokenomics, barriers to participation will have significant conquences.

I know that you’ve spoken about an EIP for this solution, @Brendan. Is there a governance solution that allows the community to participate across the prize pool network?

Networks to Deploy On

I’d vote for:
*Optimism

  • Abritrum
  • Starkware and/or zkSync
  • Polygon

I’d advocate for closing down Avalanche, as yields there have compressed a bit since last summer. I’d only advocate for closing down the Ethereum deployment if there’s a governance fix for cross-network participation.

If we’re changing the cadence of prizes, an Ethereum deployment could likely make sense if we have larger prizes. Small daily prizes have been killing the experience on mainnet but larger prizes that justify any gas cost will improve the experience.

Which Asset to Add

After a lot of thought, I’m not a fan of using POOL as the prize. Because POOL is a volatile asset, it makes the user journey much more confusing and the POL aspect across networks becomes a cumbersome burden for the DAO. I also don’t think that swapping assets for something like USDC as the prize would make much sense either. Considering there’s a fee split going to POOL stakers, this means that we’re essentially directing reserves to POOL stakers, while (at current) the protocol is not profitable.

Due to the considerable unintended consequences of using POOL as prizes, this is likely not the best way forward. Of course, we can still distribute POOL through the TWAB contract to bootstrap certain prize pools on networks where governance is present. Such an approach would perhaps lead to greater participation in the protocol and POOL staking.

Instead, I’ve thought that a new token might be a better approach–one that doesn’t come with the limitations an existing token like POOL has. You’ll likely point out the flaws in this early thought, but is there a possibility we can issue a stablecoin that is backed by deposits and can be redeemable for interest generated? If we can somehow allow a user to delegate the chance to win to an address before deploying into a platform like Aave v3 or Curve, or separate the PTUSD token from the chance to win, which could function like debt in Aave but with the ability to delegate to someone else, then we can explore more composable solutions for PTUSD.

Potential benefits:

  • Issuance can be controlled through prizes, with PTUSD being redeemable for interest accrued; upon redemption, PTUSD would be burned
  • If we can separate the chance to win from the token itself, we can make PTUSD more composable and allow people to borrow against their PTUSD or use it in other yield sources. This would make deposits more sticky, as people can always retain the chance to win while leveraging their deposits elsewhere in DeFi.
  • Since liquidity would be maintained within the protocol, the issue of POL across networks wouldn’t be an issue either
  • With greater deposits and interest generated kept in the prize pool network for longer, we can help grow TVL and the prize pool network faster by essentially compounding assets since PTUSD would represent the underlying interest
  • This use would likely create more interest in POOL and we can always create a redemption mechanism that allows people to always be able to redeem 1 PTUSD for $1.05 worth of POOL. Doing this would allow us to buy back the underlying interest while distributing more POOL to those who are actively participating within the protocol

I’m sure there is more overhead that would go into this approach, but it would likely have a better outcome that trying to make a volatile asset like POOL the unifying token. Distributing a volatile asset as a prize goes further away from the “No Loss” ethos the protocol was built upon, imo. I’d also love to clear the composability hurdle, as that will allow for more sticky deposits.

Prize Distribution

Without specifics on the composition of assets or yield sources, I believe it would be difficult to postulate potential prize tiering. Depending on the blended interest from the various assets in the prize pool network, it’s likely worthwhile to shift back to weekly prizes. However, if we can have multiple cadences such as weekly AND monthly prizes, that would make a big difference in how anyone will run some game theory on the prize pool network.

Note on Communication

I’d like to note that I’m always thankful for all the hardwork the PT Inc. development team does. You all ensure the PT community can deposit and enjoy a gamified savings account.

That being said, I feel as though the initial discussion about the new tokenomics was presented in a way where POOL holders would have a say on whether they wanted to pursue this approach. Instead, I’m under the impression from this post that the new tokenomics is moving forward and POOL holders will be able to vote on the features present within the new tokenomics at launch.

Now that we’re in a bear market, any choices that compromise the user journey within the protocol may cost us significantly in the long run. I’m sharing this opinion out of an abundance of caution, as I don’t want us to take the left-hand path at such an important juncture.

Conclusion

Looking forward to more discussion and hearing your feedback, @Brendan. I have deep respect for all the hardwork you and every engineer on the team has put into your work in the last six months. I’ll always try to maintain an honest, open dialogue as a passionate contributor. That being said: I’m always happy to be proven wrong :nerd_face:

1 Like
7

Question: Prize Distribution

Initially, we need to estimate the gauge stake total based on total deposits and estimated staker APR. If we assume that stakers desire a certain APR, we can calculate how much POOL will be staked. For example, let’s assume that staking will stabilize at 5%:

| Asset | Deposits (USD) | APR | Yearly Yield | Staker Cut | Staker APR | Staked POOL $
| — | — | — | — | — | — | — | — |
| ETH | $5,000,000 | 0.38% | $19,000 | 10.00% | 5.20% | $36,538 |
| wBTC | $23,400,342 | 1.00% | $234,003 | 10.00% | 5.30% | $441,516 |
| USDC | $35,000,000 | 2.26% | $791,000 | 10.00% | 4.90% | $1,614,286 |
| FEI | $10,000,000 | 2.29% | $229,000 | 10.00% | 5.00% | $458,000 |

We can see there will be a total of $2,550,340 worth of POOL staked. This gives us a baseline against which we can set an initial value of the gauge stake total.

The executive team will update the gauge stake total. The executive team will be able to quickly adjust the value in the event of major staking changes. Just like how we adjust the prize distribution now.

Question: POOL Participation

I want to make it clear that POOL token holders will be able to stake from day one. The governance EIP is for cross-chain proposals, which is a separate issue. The executive team will be managing the deployment, as they do now.

Question: Draw Frequency

I want to clarify that we’re not changing the cadence of the prizes, but rather statistical probability of the prizes. In other words: there is still a daily draw but prizes could occur, on average, less than once per day. I.e. there is a daily draw, but there could be yearly prizes.

Question: Prize Token

In my mind, the choice is between POOL or a widely available stable such as USDC.

  • If we give out prizes in POOL then our only problem is ensuring there is liquidity on the chains on which we operate. POOL volume would increase and open up new possibilities for POOL.
  • If we give out USDC, then it has plenty of liquidity but we wouldn’t be capturing any of the buy pressure from liquidations. POOL volume will remain low.
  • If we minted our own stablecoin, not only would we need to manage liquidity but now we’d need to manage issuance. It would add a horrific amount of complexity!

The question of which token to use for prizes is a big one! We should probably break it out into a separate thread.

Question: Prize Tiers

We don’t need to postulate! We have the existing prize pools to use as a baseline. We’d look at how much yield accrues per day, then tune the tiers and odds as we’ve done before.

For example:

  • let’s say $1000 in yield accrues per day.
  • We estimate there will be 2,550,340 POOL staked given the TVL and yield (see above)
  • Let’s say our prize distribution gives out a total of $1,000,000 in prizes.

The odds need to be adjusted so that the desired daily prizes are $1000.

estimated stake / gauge stake total = daily prizes / prize total

solve for gauge stake total:

gauge stake total = ( 1,000,000 / 1,000 ) x 2,550,340 = 2,550,340,000

If we configure the gauge stake total to be 2,550,340,000 then approximately $1000 will be given out per day for a $1m prize distribution.

Question: Decision Making

Ultimately, POOL token holders will decide whether to upgrade V4 to support the new network. This is not something we can just ram through.

The goal of this post, the upcoming design, and the testnet will be to show people what is possible. To show how we can evolve PT to become a multi-chain, multi-asset juggernaut. It’s not easy for everyone to visualize the possibilities from a few math equations or diagrams; so we’re spending resources to ensure that everyone understands the vision.

Once the voters are informed, they can decide whether or not to move forward.

1 Like
8

Now, this might become more apparent in the testnet, but how does a POOL holder know that the odds aren’t fairly distributed and how to distribute them appropriately? Conceptually, I understand that POOL token holders will balance the odds and will get rewarded for it, but I’m bogged down by the details and the how.

You touch on this below:

Can you ELI5? What drives the incentive to balance out the odds? I’m sure you’ve sufficiently answered this question with your table, but I need a bit more, lol. Before you answer this, I make an attempt to understand below.

In this example, do the players in the Optimism pool have double the chance to win the same prize as the Polygon players, therefore you want them to be equal? The incentives will push the stakers to move away from Optimism to Polygon, until they balance out? I am assuming the driver here is higher APR for staking POOL on Polygon as long as it’s disadvantaged?

I hope I’m somewhat understanding…

1 Like
9

Just for the sake of clarity: If I put up a vote to change the fee tiering for POOL stakes or put up a proposal to create a new subDAO, will I be able to participate in the Snapshot or on-chain vote with the POOL I have staked on a gauge?

That’s the aspect I was trying to capture in my comment there. For the rest, I’ll wait for separate threads to discuss further.

Thanks for the responses, @Brendan :turtle:

1 Like
10

We’ll have a nice interface showing odds and APR of each pool.

Exactly! The staking APR should balance out as people seek the highest APR for their POOL.

I see what you’re saying- yes! We’ve already added Polygon POOL to Snapshot, so users will able to signal using their tokens.

What I haven’t mentioned, however, is that we’re also going to ensure that staking leads to on-chain votes. Users will be staking by depositing into a voting power contract, then staking those tokens into the gauges. I’ve glossed over those details for now, but it’s important to consider.

2 Likes
11

oh my god, good job team :+1::+1::+1:

3 Likes
12

Liquidation Algorithm

The PoolTogether Liquidation algorithm is now available on Github.

The project includes:

  • Full explanation
  • Diagrams
  • JavaScript simulator
  • Graph of simulation results

Quick Description

The PT liquidator will allow the protocol to sell yield for another token. For example, if the prize token is USDC and yield is accruing in CRV, then the protocol would be selling CRV for USDC.

The PT Liquidator is inspired by AMMs. This means the liquidator operates on pairs.

The link above includes a diagram of simulation results. It includes a graph of the liquidator purchase exchange rate vs the actual market exchange rate for tokens:

LiquidatorPriceTracking
LiquidatorPriceTracking2394Ă—1476 187 KB

The liquidation algorithm does an admiral job of tracking the market rates!

6 Likes
13

Looks great!

I think it will be important to move off Ethereum if frequent liquidations are happening, otherwise the protocol is gonna be paying gas like crazy. Polygon is a natural starting point as well as whatever rollups we’re prepared to launch on.

Of course a lot of yield is on Ethereum L1 still, so there is a bit of a tradeoff there too. But in any case if there’s one Network to leave it’s the one that creates the most disgruntled users (Ethereum, due to gas fees). I don’t care either way about Avalanche staying or going.

I also want to commend the fact that this approach not only makes PoolTogether more appealing to other protocols/DAOs, but adds value to the POOL token.

2 Likes
14

Personally not a fan of using POOL token as the Prize Asset (I mentioned my reasoning in the Tokenomics discussion from a while back.) but USDC, makes 100% more sense, thats the whole point of POOL, would be great for the marketing to “Get payed in USDC, No funny dog coins or monkey business”

Deploy on OP and Polygon, we can add ETH back if it ever makes sense again.

Avalanche isn’t great and our Token isn’t there. so yeah take it down, and Ethereum at a low priority of deploying atleast for the moment.

2 Likes
15

This is clearly a major effort to take the protocol to new heights. I really appreciate that. I think the reasoning and direction is clear.

I have thought a lot about this over the past two months and have reservations on both ideas - one token for the prize, and of POOL driven gauges. I think it’s crucial that we enable adding assets other than USDC but I’m not convinced yet this is the best approach to do so. Tangentially, I am more in favor of a permissionless prize pool building. Like the V3 builder, where anyone can build a prize pool and we as a community are the facilitators of prize pools as opposed to curators of a prize network. I believe that is more the direction of a base layer protocol. But I also think it’s possible we try both!

I want to clarify this point. I think you are saying it’s cumbersome specifically for the prize liquidity. But as far as I understand it, the prize network concept still requires creating new pools for every asset? To be clear - I like that separate prize pools isolate the asset/yield source risk. But, zooming-out, I think the cumbersomeness is specific to the cross-chain meta prize pool concept, which creates a disconnect between prize liquidity and yield generated.

Clarifying this - we still have to bootstrap and manage liquidity per chain, it can just be one asset, right?

I don’t like using POOL or USDC for prize. If we have to choose one I say POOL because I think the reflexivity has a short-term chance to outweigh the costs. And we have more of it. From my perspective the prize pool network adds cost and complexity and I believe our design from yield to prize should be as slim as possible.

I am curious about @BraveNewDeFi 's idea to create a stablecoin. I agree that composability is something we should really be trying to tap into.

I’m skeptical POOL holders will participate. They literally do not one click vote on Snapshot to renew their own rewards for a free 9% return. Twice now. How will this system motivate them?

I think this is in conflict with V4s premise to be on every chain. Requiring POOL liquidity becomes a hurdle to expansion.

If having to choose one I would shutdown Avalanche before Mainnet. I want to give Mainnet a fair chance to have traction before abandoning it. I’d really like to see us add Arbitrum. From my perspective I think we are already running late on the opportunity.

Am I wrong in thinking the Prize Pool Network will enable asset expansion but limit chain expansion?

I think giant infrequent prizes that occur at random times is going to be a challenge. We can target one per quarter and see four. Or none. Variation is common and we see that with the grand prizes now. It’s further complicated by the fact we do now know when they will happen, so hyping them becomes a hard problem. How do you see less frequent prizes being messaged?

Does the prize network allow us to change the draw frequency? I think drawings every day are too frequent and create the wrong expectations for depositors. Premium bonds award once a month. I’d really like to see us push forward in a direction that aligns with the mentality of saving as opposed to farming. I would like to see draw frequency changed to once or twice a week.

I’d love for us to reflect more on V4 as a community. I’m guessing PT Inc has done a lot of this, but as a community I feel we have not. I see The Prize Pool Network as a do or die on the meta prize pool concept, which I personally am not so sure is the right direction. There are complexities and costs that I think need to be considered and the way I’m looking at it the Prize Pool Network adds to those.

1 Like
16

Hey Brendan, thank you for putting this together! I had a couple of questions about the liquidator you outlined here:

  1. It seems suboptimal that the liquidator price is always slightly higher than the market price—this would mean the protocol will always be eating fees if this liquidation mechanism were used. From the simulations you guys ran, do we have a preliminary idea of how much the protocol is giving up relative to if it were able to liquidate at the actual market price over some time period?
  2. Would it not be possible to just liquidate the yield drectly in existing AMMs? This would obviously entail slippage+AMM fees, but both slippage and AMM fees are already implicitly taken into account when using the liquidator you outlined (since others/bots have to arbitrate, and they would have to incur these fees directly).

I also wanted to ask a couple of questions to follow up on staking $POOL:

Does this mean that from day 1 of the Prize Pool Network being live (assuming it passes an on-chain vote), that $POOL holders will be able to stake on prize pool gauges on every chain that the Prize Pool Network will be live on? If not, would it be possible for you to provide more color on exactly which chains $POOL holders will be able to stake from day 1 of the Prize Pool Network going live?

1 Like
17

Hi Yuan!

Yes! When the market is relatively stable the liquidations can be tuned to be 99% efficient.

Using an external AMM is not ideal for a number of reasons:

  • It creates a dependency on an external system, which may be different from chain to chain
  • Forces us to bind to a particular source of liquidity, but market liquidity may change.
  • Requires liquidity in the market to exist. If we want to onboard hundreds of tokens then each may have liquidity in different markets, or not at all. By letting the arbers come to us we don’t dictate where they get their liquidity from.

Yes! We’ll be able to stake POOL right away. It looks like staking will be available on Polygon, Optimism, and possibly Ethereum (depending on what gov decides)

3 Likes
18

Hey,
Théo from mStable here. Thanks for this very interesting write-up @Brendan
Bringing a little bit of colour to why using mStable as a Yield Source to power the Prize Pool makes sense:

  • Save APY (mStable Savings account product) has been outperforming competition over time on stable assets with roughly a 10% 18 months-Average APY (Dune) without adding any risk
  • Save APY is organic and generated from three different revenues sources (Lending on Aave & Compound, swapping fees & external rewards liquidation)
  • mStable has been live for more than 25 months and processed more than $5bn in transaction volume

Happy to answer any question Pool Community might have :slight_smile:

2 Likes