Prize Adjustment #2

[Leighton]

It’s time for us to look again at the prize distributions! A few data points:

• As of this writing there has been 87 claimable draws and $911,160 in prizes claimed, there is an additional 139,000 of additional outstanding prizes that can be claimed. (all calculated using @drcpu excellent data set!)

• During this 87 days, deposits have generated $136,129. $71,332 has been in the form of USDC which has gone directly into the prize distributor contract and $64,807 has been in the form of Matic, Aave, and Avax token incentives which are currently held by the protocol.

• This means net outflows over the first 87 draws was $775,031. If you divide that by 87 (the number of draws) you can see that our average net outflows per day is $8,908. This is a rate of $3.25 million annualized.

Operating at a net loss initially was expected and was why governance voted through the initial $1 million prize subsidy. However, now that ~77% of the initial $1 million has been depleted*. We need to take action, broadly our options are:

1. Vote to disburse additional USDC from the treasury to continue funding prizes

2. Vote to decrease total prizes to match inflows

So what should we do now?
It appears we’ve more or less topped with growth in the last 2 weeks (reference @sarfang excellent dashboard but note it does not include Avalanche deposits).

I believe some of this is a function of the holidays and marketing being down but regardless, my personal opinion is that since growth is slowed and USDC is a scarce resource it doesn’t make sense to continue funding at the same delta we are currently. We can ramp back up at times but right now I think it makes more sense to slightly scale back while we get out some of the newest tech.

So I’d propose scaling prizes to target a delta of around ~$4,000 USDC per day. That’s a solid ~40% reduction in the current rate of depletion. Much more sustainable and prize APRs still higher than the underlying yield sources.

What exact distribution we should do is likely requires a second post. It’s more complicated than it first appears because of the prize cap and statistical distribution. However, I’ve put together this doc as a starting point with one idea. This was built using @drcpu data on how many claimable prizes are available for each tier.

Longer Term
I don’t think it’s a fair expectation to achieve prize sustainability in the near term. But we should use any subsidizes wisely always ensuring they increase growth with an eye towards ultimately being sustainable.

This post is primarily talking about the short term considerations but I briefly want to mention our toolkit for the longer term to achieve prize sustainability.

• Decrease prizes
  The simplest but least attractive option.

• Increase deposits
  This seems too obvious but there are actually lots of ways to potentially increase deposits, an incomplete list: move sponsorship into V4, partner with other protocols, venture funds, banks, exchanges, etc. that want to deposit sponsorship, use TWAB rewards to incentives more deposits (via POOL distribution or any other token), launch on more blockchains, create a more engaging prize distribution strategy that attracts more deposits, integrate NFTs into the protocol making the experience more engaging, partner with Neo-banks to deposit into the protocol, etc. Main point here is there are a LOT of ways we can increase deposits and most of them don’t rely on increasing prizes specifically.

• Increase yield generated on deposits
  This should also not be over looked. Our yield on deposits is quite low, we could double it just by recursively using Aave rather than supplying on only one side. As the prize pool network grows, this will be more of an option!

Action Items:

• Determine how much to reduce net out flows by
• Determine new prize distribution strategy

• Support a new prize strategy to reduce outflows
• Oppose a new prize strategy to reduce outflows

*It’s worth noting that because 1) ~50% of the total yield is in non-USDC tokens and 2) there has to be enough USDC stored on the different chains we are going to need to outlay much more USDC than we are actually giving out. Any analysis should always keep this in mind. We could of course claim and sell these non-USDC tokens for USDC but personally I think the diversification and potential upside is nice. Ultimately, I think treasury committee should decide.

[Tjark]

It seems so far the Prize APY has not done as much marketing as it could. Going back to general responses people often seem to think our value proposition is too good to be true and one of the reoccurring questions is “how can this be sustainable?”.
With this in mind I agree it could be worth going another direction that is cheaper and find a better use for the funds we’re saving in that process.

[RogueMerse]

NFT market is booming and probably is going to be considered a parallel market then cryptocurrencies, we are going towards “tightening” fears and so the crypto markets, could be a smart move to start thinking in creating a NFT collection that lets owners participate in the pooltogether protocol with advantages and lets the protocol profit from it’s sale and royalties.
Just thinking out loud here. I love this project but i haven’t seen the publicity and hype it should have

[DaBoom]

I’d like to see
1…2500
48…100
192…10
768…5

Total = 13,060 giving a 4000/day decrease while hearing the feedback on the 1$ prizes for the community

[underthesea]

The poll shows it is clear the community agrees that as the 1 million prize bonanza comes to an end we should re-evaluate the treasury spend, opting for less of a subsidy. I think it would do us well to reframe the change in prize distribution to be “continuing to offer returns 2-3x Aave” instead of “reducing the subsidy 40%”, even if on-net the result is the same.

I have spoken with several people about a new tiering and distribution recently and came to the conclusion that there are two contentious issues.

1. Should we continue $1 prizes
2. Should we change the prizecap from 1 to 2

I suggest we have polls for each of these issues. I’m not sure if we should use a new thread or have those polls in this thread. In my opinion contention in the DAO can be a good thing. Let’s engage!
Votes on polls should be considered from known-members of the DAO. There should be clear consensus to change the status-quo, like 60%+ to eliminate the current $1 prizes.

In regards to the proposed tiering in the google docs; I found through simulations that it did not proportionally reduce the net result for whales. I think this is an unintended outcome. I also did not see how the prizes were a 40% reduction in spending, but there are a lot of moving parts there so I likely did not fully grasp your calculation method.

I think users are looking at three metrics. Total prize distribution, prize yield, and odds. Personally I would like to see us target a 15% prize yield. I think this is pretty close in-line with the 40% reduction proposed in this thread. With the prize cap and tiering we can have the 15% diminish towards 6% as the deposit amount grows to whale status ($1 million and above). By keeping a similar total number of prizes we can keep the odds in a similar range.

The prizecap allows us to keep the prize distribution big, but cut a little more of the treasury cost. In my opinion a prizecap of 1 makes the most sense. Two prizes a day just makes people wonder, why never 3? One is natural. The second prize that people win is effectively farmed yield straight from the treasury. Drcpus numbers show that despite wallet splitting the prizecap benefits the treasury and smallfish.

Lastly I think we should split the $2500 into 3x$1000 prizes and have zero at the top. I think there is consensus for more big prizes and to tilt the tiering so that more depositors are outsizing their deposits with wins. Let’s lean a little more towards a feeling of winning exciting prizes and away from yield (2 prizes a day).

[Brendan]

I’d like to add another discussion point that I think is worth covering: having different tiers per chain. This idea has already been discussed and has support.

This morning the PT Inc team discussed feasibility in terms of user experience and technical work. The team believes having the same prize amount per prize pool is essential. It’s an incredibly important meme to say “you could win $X amount of dollars every day” no matter what chain they deposit on.

However, how that prize money is distributed can be configured per-prize pool. This is important, because:

• Polygon users clearly love the $1 prizes. We constantly see tweets about it.
• A prize below even $200 isn’t worth it on Ethereum. Users may see this and feel like they’re being scammed.
• Growth has stagnated on Avalanche, even with the subsidies. It may be worth experimenting by tuning the distribution.
• L2s such as Optimism will also have higher transaction costs

There are a few smallish changes we’d need to make to the app to support different prizes per chain, but nothing terribly significant. The smart contracts are already set up for it.

It’s worth having the conversation now, as we explore changing the prize distributions. Shall we make fat prizes on Ethereum, a slightly broader distribution on Avalanche, then a dist to make fish happy on Polygon?

[BRONDER]

I’d like to add another discussion point that I think is worth covering: having different tiers per chain . This idea has already been discussed and has support.

I agree that different tiers per chain is a super interesting concept, but balancing the different pools is harder than I thought at first sight. Therefore, coming to a conclusion as a community would take us some time which, unfortunately, we do not have right now.

Prize distribution needs to be changed, and to do that we have to get things moving with some polls and opinions. Let’s go!

First topic is $1 prizes. Some people love them, some hate them. I think they are really interesting and actually improve a lot the chances of getting a prize (metric that people really look at). On a different side of things, gas prizes on Polygon are crazy these last days, which makes $1 prizes not really viable. I know that in the future it may be possible to have those prizes only on specific chains, but for now, I would focus on one prize distribution for all chains.

Should we keep $1 prizes?

• Yes, keep them!
• No, eliminate them!

Second topic is the prize cap. We have been with a 2 prize cap since the beginning of V4. For those that have not thought much about it, what the prize cap is effectively doing is decrease the expected yield from depositors with higher amounts, while not affecting small fishes. From some simulations, I would say that people with less than 10K are basically not affected, and the more you go up from there, the more it may affect you. However, this does not mean that whales are not getting prizes, in fact, right now they should expect over 10% APR, which is still great.

What should we do with the prize cap?

• Decrease the prize cap to 1
• Keep the prize cap in 2
• Increase the prize cap to 3 or higher

I’ve tried to be “impartial” the best I could to check what people think and get the voting going. I will express my opinions cleare once we get some answers.

[Gonbatfire]

@BRONDER

Decrease the prize cap to 1

Reducing the yield of big players while retaining the yield of small depositors does seem like a mouth-watering solution to all our problems, the problem is, can we actually do that?

Sometimes actions have the opposite effect of what the actual intent is, this reminds me of what I see everyday on “overreaching” governments, example:

• Gov passes a law to create a new 25% tax on the “rich” “to build more hospitals”
• Their intent is to “not affect regular people”

What happens next:

• The capital moves somewhere else where it maximizes it’s profit
• There is now more incentive to evade taxes

And now even if the gov had the true intention of building more hospitals, they now collect less tax income due to the above, which actually pushes them further from their objectives and worsens everyone’s situation (Now add some money printing and you have my home, Argentina)

What would actually happen if prize cap were to be decreased?
Based on my personal experiences:

• Investors would move to a more profitable protocol
  I don’t see this is a problem since they would move away when V4 becomes sustainable anyway (because APR will be >4%) PT is not an investment, it’s a savings account.

• Affected users would have more incentive to evade

Now this actually worries me, right now the 2 prize cap is not 100% effective, because people “evade” by splitting wallets, still, the prize cap does bring us some “income” (not claimable prizes) because some few users don’t actually bother splitting (altho they could with little consequences)
But what if prize cap decreases? Sure, some people may still not bother, but a higher “tax” (APR decrease) will 100% raise some eyebrows, and incentivize a higher number of people to evade, and once they evade once, they are more likely to keep doing it (creating multiple wallets, maybe start looking into scripts, and so on to the point the prize cap is nullified)

Therefore if this incentivizes enough people to split, we could end up in a situation where we actually have less unclaimable prizes, than if we had left the prize cap as is. And enforcing an stricter control on the cap is not really option since unlike govs we don’t have monopoly of force

[BRONDER]

I voted to eliminate the $1 prize in the distribution. This elimination is not permanent, and here is my reasoning behind it:

Right now we will have only one distribution for the 3 networks that we’re currently at. Prizes that are more expensive than gas fees are not hidden, which basically means that $1 prizes will bring frustration to users on other network apart from Polygon.

The deal breaker for me however is that Polygon fees right now go over 50 cents and unfortunately I do not see this as an exceptional occasion, so we will need to get used to it. Giving prizes where more than 50% of the prize is used for gas fees is basically really not efficient. I would much rather have a POOL drip or something of that sort than $1 prizes (just given how things are right now).

The obvious option that comes to my mind is making those $1 prizes into $2 prizes, however, it is not really sustainable to have 3,000 $2 prizes at the moment. With the DPR introduction we will be able to have a “more configurable” number of prizes and we will probably be able to make $2 prizes a reality. However, for now, I’ll eliminate them.

Regarding the prize cap, I would set it to 1 instead of 2. I absolutely agree with the behavior exposed by @Gonbatfire , however, I think that there is 1 factor that you are not considering. The factor is that we are not collecting taxes, but quiet the opposite, we are giving out money as a marketing action. Therefore, it is up to us to decide how to give that money.

Honestly, I do not want to give treasury money to whales, but instead I want to attract thousands of smaller depositors that can have a good experience with PoolTogether and spread the word. A 1 prize cap allows us to achieve that behavior while also giving actually a really good yield to whales. Also, a 1 prize cap is also great for our treasury, as it allows us to have more dropped prizes, which actually go back to the treasury improving long term viability.

The decision, although it may seem controversial for some, is clear to me. Once the incentives finish, the prize cap will probably get eliminated (unless some other compensation mechanisms are created).

[Gonbatfire]

@BRONDER

Therefore, it is up to us to decide how to give that money.

Unfortunately I do not see it this way, maybe we “should” be able to decide, but imo the reality is that we cannot. There is not an efficient way in which we can possibly differentiate users in a pseudonymous blockchain, where anyone can create a new identity with a few clicks on Metamask, best we can do is play along with “cheaters” and not try to wake up more of them

[Leighton]

I think eliminating the $1 prizes right now would be a huge mistake. Here’s why:

• 62% of all daily prizes are $1. We take these away we now have 1,212 wallets per day that use to get something now getting nothing. How do you think people will feel this makes the protocol better? Hypothetically they have a better chance for larger prizes but there is only $1,212 in $1 prizes per day. So that means any addition to larger prizes will be tiny and not make a meaningful improvement in odds.

• We’ve gotten universally positive feedback on the $1 prizes. I have yet to see a single actual user complain about winning $1 prize. I’ve seen people complain about it in theory but actual users have been happy to win them.

• 95% of all our users are on Polygon – we should optimize for the best distribution where the largest amount of our users are. Optimizing for Ethereum makes zero sense. Even with a prize distribution optimized for Ethereum users still have to pay hundreds of dollars in gas fees. It’s not where growth will happen.

• $1 prizes are not about the financial returns or gas fees. They are about making the protocol fun to use and proving that it does work! Has @Gonbatfire said, as soon as his girlfriend won a $1 prize she wanted to deposit more. People are using PoolTogether because they want to have fun and get a return – lets make sure that happens.

• If our logic is that we should eliminate all prizes less than gas fees than we would need to eliminate all prizes less than $100 because that’s gas fees on Ethereum.

A better solution than eliminating 62% of all prizes would be to have chain specific distributions. However, we are not ready for that right now so lets keep this optimized to make the protocol fun, engaging, and useful to our largest amount of depositors.

[Brendan]

I 100% agree on all of your points regarding the $1 prizes

However, this point:

@Leighton

Even with a prize distribution optimized for Ethereum users still have to pay hundreds of dollars in gas fees.

This is a big assumption. Yesterday we saw a deposit of $1000 even though we are displaying the deposit, claim and withdraw gas costs each of which is $120+. Saying that users will never use Ethereum is just not correct.

I don’t want to abandon Ethereum or Avalanche. I think we need to tailor the prizes to be different on those chains to make those users happy. When we add an L2, we’ll see higher transaction fees than Avalanche.

Otherwise, if we refuse to tailor the prizes per-chain, then we might as well shut those pools down as they aren’t cost-effective.

As discussed, we can still have the same headline- “$100,000 in prizes per day no matter which chain!” (or whatever the denomination). However, we can break down the amounts into appropriate chunks.

We’ll have to revisit this when we use an L2 anyway. Might as well tackle it now.

[Gonbatfire]

$1 prizes bring a ton of value to the protocol, ESPECIALLY outside developed countries (Let’s not forget this!!) so I can agree with you on that

However I do think that Polygon being $50 cents to claim right now is an issue we need to tackle, but instead of just throwing $1 prizes away, how can we make them work more efficiently? Gasless experiences? (Biconomy), non-custodial secondary balances? Subsidizing?

[underthesea]

@Leighton

62% of all daily prizes are $1. We take these away we now have 1,212 wallets per day that use to get something now getting nothing.

Does this calculation make the false assumption that all $1 prizes go to a single wallet that wins a single prize? I ran a query on draw 95 and the number of unique wallets that won a single $1 prize was 457. I’m not sure how you derived 1,212. Also I think you are assuming that if we eliminate $1 prizes that automatically deducts that entire prize count from the total. If we do not use the first tier, maybe we can change the bitrange? Part of the possibilities I don’t fully understand.

My number one problem with $1 prizes is the uncertainty of Polygon gas going forward. Yesterday people were complaining about the cost to claim on Polygon. Literally yesterday. Our claim transaction uses a lot of gas. If I thought that Polygon was going to be 40 gwei for the next two months on average I’d be happy to have $1 prizes. I don’t think that will be the case, and I’d hate to see us assume it will be. The number one complaint I’ve seen about Pool Together over the past year is the gas cost. Why would we put ourselves in a position to continue to suffer from that fate?

User experience matters a lot. I have onboarded a lot of people to V3 and V4. When the gas spam issue came up a few weeks ago I received texts of people withdrawing from V4. I’ve seen this happen on BSC. Low cost networks have historically not been sustainable. Polygon will still be relatively low cost and our claim will cost 80cents because its fundamentally high on gas usage.

I really urge people to think about the implications of our subsidies effectively going to Polygon validators. That’s not effective in my opinion. We have to assume gas will be a problem. It’s always been a huge problem for us.

If we proceed with $1 prizes, what will be our threshold for gas pain where will we take other action?

[Leighton]

Does this calculation make the false assumption that all $1 prizes go to a single wallet that wins a single prize?

This is based on @drcpu data here. That data show that on average, every day there is 1,212 claimable $1 prizes.

I think getting clarity here is important, because like you said, if it really is just 457 wallets per day – that’s a good bit different.

[BRONDER]

The feedback of the previous polls just tell me that we’re actually pretty divided in opinions. Therefore, I will post 3 options that I believe could suit different opinions. All of them will work with the 2 prizes cap:

Option A:
$0 - 1 - $0
$1000 - 3 - $3,000
$0 - 12 - $0
$50 - 48 - $2,400
$10 - 192 - $1,920
$5 - 768 - $3,840
$0 - 3072 - $0
$11,160

Option B:
$0 - 1 - $0
$1000 - 3 - $3,000
$200 - 12 - $2,400
$0 - 48 - $0
$10 - 192 - $1,920
$0 - 768 - $0
$1 - 3072 - $3,072
$10,402

Option C:
$0 - 1 - $0
$1000 - 3 - $3,000
$0 - 12 - $0
$50 - 48 - $2,400
$10 - 192 - $1,920
$0 - 768 - $0
$1 - 3072 - $3,072
$10,402

Which option do you like best?

• Option A
• Option B
• Option C

[drcpu]

Let me be the savior here, it all depends on how you interpret the data.

Focusing on draw 95 because that’s the one that @underthesea picked.

There are 457 wallets that win a single $1 prize:

SELECT COUNT(*) FROM prizes WHERE draw_id=95 AND claimable_prizes='{99999999921875}';

There are 288 wallets that win two $1 prizes:

SELECT COUNT(*) FROM prizes WHERE draw_id=95 AND claimable_prizes='{99999999921875,99999999921875}';

There are 264 wallets that win one $1 prize, combined with another prize:

SELECT COUNT(*) FROM prizes WHERE draw_id=95 AND 99999999921875=ANY(claimable_prizes) AND claimable_prizes!='{99999999921875}' AND claimable_prizes!='{99999999921875,99999999921875}';

There are 334 wallets that win one or more prizes which are not $1 prizes:

SELECT COUNT(*) FROM prizes WHERE draw_id=95 AND NOT (99999999921875 = ANY(claimable_prizes)) AND claimable_prizes!='{}';

There are, in total, 1343 wallets winning one or two prizes:

SELECT COUNT(*) FROM prizes WHERE draw_id=95 AND claimable_prizes!='{}';

As usual, the truth lies somewhere in the middle. The two top queries are people that would not have won any prize if $1 prizes were to be eliminated, which amounts to 745 wallets. The two middle queries show there are 598 wallets that would have won a prize regardless of $1 prizes existing or not. In total, this means there are 1343 wallets winning a prize (as shown in the last query).

Note that there is a small discrepancy with the data shown here which highlights only 1326 winners because that analysis takes duplicate wallets over multiple networks into account, but that’s too complicated for simple queries.

We can of course still debate about how many of those wallets are unique users, but it is definitely safe to say that eliminating $1 prizes would result in only around half of the wallets winning a prize (assuming draw 95 is an average draw, which it seems to be).

[DaBoom]

…Option A is eerily similar to my suggested prize tier above. What’s your end game here Bronder? haaha j/k, I would like to know why you chose Option A to have 48 50$s instead of 48 100$s. Seems 48 50$s is better for the protocol, so if that’s the case I like it! Still want to know, just in case there is more to it. As always, thanks for steering the ship!

[BRONDER]

You basically nailed it yourself. Note that option A still distributes a bit more funds than options B and C. Going for 48 $100 prizes would also mean that distributions B and C would need to have other prizes for the 3 of them to be 3 balanced options.

In terms of simulations there is really no difference given that the number of prizes would be exactly the same for the 2 options. Of course with $100 prizes the average APR would be a bit higher, but that cost would come exclusively from the treasury and I believe it is best to save those expenses for now.

[Leodisc19]

Many people who today win prizes of $1, have the feeling of being closer to the big prizes, if the minimums return to $5 the small fish will spend many more days without receiving a prize, I think this will lower the expectation of many participants.