MakerDAO FriDAI Brunch Series, 20 Jan 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: FriDAI Brunch n8: A Deep Dive into Prediction Markets

Description

For the new episode, we invited some experts operating in the "magical" word of Prediction Markets:
Ish Goel, CEO & Co-Founder at PlotX
Rafael Suarez, Business Development Manager at Gnosis
The topic is very fascinating and we will learn how & why it became popular lately and the challenges they have faced.

A

B

A

Everyone and welcome back to the friday brunch series happy 2021 full of expectations, and we hope uh we it will be good, especially from uh purple I mean in defiant crypto. It has been a good year, 2020, but not outside this world. So, let's hope for 2021.

A

Today we will talk about a brand new topic which is not new, but maybe it's new for uh most of the people um and it's super interesting and it's a prediction market and we have two special super special guests, uh my friend rafa from gnosis and my new friend ish from plotex.

A

uh So this is these two projects work in the prediction market space. So I just want to uh allow our our guests, who present themselves starting from each uh and I will mute myself.

B

Thank you so much maria I'm very excited to be on this call thanks for the opportunity, um so very quick background on myself guys. I've been an entrepreneur for almost 14 years now uh the crypto bug kind of caught me back in 2015-16. That was the time when I started getting connected in the london uh blockchain ecosystem um got to know about ethereum. For the first time um I ended up winning the london blockchain week, hackathon in 2017, and that's when I met you card from nexus mutual.

B

So I joined him as a cto at nexus for around three years, we've built nexus from scratch, um and I essentially moved on in feb 2020 to build onto my idea of what what we think of prediction markets as and that's when we started building up a team and the idea behind plot x really came in. We spent about six months to launch our beta back in october, 2020 and uh yeah. That's that's a very quick introduction of uh what we are doing.

C

Okay, well, I guess I can take the mic now and well. First up again thanks maria and really nice to meet you ish, uh so well, um as ish mentioned, I've also been like in the blockchain space for some time. uh My journey started with bitcoin around 2014.

C

and well. Bitcoin didn't get me only like into like the blockchain rabbit hole, but also in general, like into financial markets and trading and whatnot. uh So then, yeah after, like uh yeah working for some uh for some years in different startups as business development role, in mainly around like the data industry, focus around light market data.

C

I started working for gnosis in 2018 and yeah. It's been a great journey like pushing products across like different categories um and well diagnosis was originally known, like for prediction markets, but like right now we also offer one of like the biggest um or well the biggest like multisig on ethereum and additionally, we are also developing several text protocols.

A

Cool thanks guys um just one more question to uh ice to break the ice, even if it's already breaking broken. So what is your personal challenge every day we start again from ish.

B

Okay, I think that's an excellent question maya, so my you know the day-to-day juggle specifically with plot x, uh that we kind of try to work on is essentially to improve our incentive alignment mechanisms um for users for traders and also, I think, to make prediction markets more capital efficient to bootstrap liquidity. I think this is one challenge which rafa would also agree with me on. I think this is. This is one of the biggest issues that we are juggling with and trying to fix on a day-to-day basis.

B

C

Rafa yeah. Well, I can only echo like similar things like yeah making in the end like products that uh that can like have a more like broad adoption, because, of course, like prediction, markets benefit a lot like from liquidity, uh not only for like a nice like uh ux and like a good like experience as someone like quick clicking around, but also as like uh yeah. In order to have a like an accurate prediction or aggregation of sentiment.

C

It is like quite necessary to first like have like a liquid market and then, of course, like one issue. uh That is not only like part of like liquidity. uh Pardon me like uh about about like texas and prediction markets is having a yeah like an easy ux that like enables uh more mainstream users, which I think is like a challenge that anybody working on the ethereum space is facing. Right now.

A

Yeah, I think you guys mentioned some of the challenges that we all have in this space because ux and make people understand what what we do it. Actually they are destroying so um okay, but now we can start the conversation about the topic and, as I as I said before, maybe not everybody know what is actually a prediction mark I mean what what we are talking about.

A

So, if um rafael, you can give um a definition, how how would you define a prediction market in few words, because I know that you like to talk a lot.

C

Well like in I mean, of course, it's kind of like a broad subject, but uh I think, like a concise definition uh that I like uh uh I I mean I I personally like defining prediction markets as an asset class in which the outcome of an event is tradable, uh that's kind of like what I would say is the single sentence, because of course we can go like very broad and like uh there can be like a market around like, uh for example, prices.

C

I believe products is quite a focus on that, and but also around like political events and whatnot and in general, just anything that you can like formulate as a question uh can be like uh yeah traded in a prediction market. So, just once more as a concise, like one sentence thing, I would say that it's uh yeah an asset class in which the outcome of an event is tradable. uh I don't know if you would have something to add to that ish or if you would agree or debate.

B

Anyone, I think I agree with it. I would just like to kind of add a very dumbed down version, for uh you know I guess the audience which may not understand uh these. uh Some of these financial jargons.

C

Because you know I care if I I'm not a financial expert.

B

I am I'm a techie, I've learned, um you know being in the new fast pace for five years now or four years now. I think there are a lot of I try to kind of dumb it down for myself, and that makes it easy for me, so I think uh prediction markets the way I look at them. I think they are paid polls. I mean you need to pay to enter a poll, um unlike the free polls which are um you know, which you can run on twitter.

B

So, like somebody's running a twitter poll where you know, people are asking questions and there are three four or five options. Now they are gameable. You know you can hire people to kind of vote for you, but a prediction market. uh All that rafa said is absolutely on point, but for me uh you, I also think of them as polls where to enter a position. You need to pay for it and there is a reward. If you make the right judgment, you earn money.

B

um So I think uh that's a dumb down version of how I think prediction markets function.

A

Cool, thank you guys. I think it's pretty clear, so um we can go ahead and my next question is for- and is this: when do you think prediction markets um got popular and and what is the connection between defy and prediction markets, because we know that defy exploded last year? Basically, um and it started just two years ago- and I want to know if prediction markets follow the same path or it's.

A

B

Oh absolutely um definitely I'll try my best to answer that and would love rafa's input as well. So I think uh prediction markets became popular quite early in time. Right I mean I was having a chat with one of my colleagues in the team and he explored a very interesting story uh back from 1900s.

B

So there was a you know. That was the time I think when a prediction markets really started coming, people started realizing the importance of the concept of wisdom of crowds. So you know they would uh people would gather to.

B

You know assess the weight of uh an ox uh and they would people would come together to you know kind of, uh say: okay, this is the weight of the ox so like there was a story about 800 people coming together, and you know pointing out what the what the weight of the box could be uh and turns out that there was a there was a scientist which kind of studied the data and realized that the mean of the inputs, which the crowd had given were actually more accurate than the uh even the experts who were talking about this is the weight of the ox um and also the winning person out of the 800 people.

B

So the idea really was that you know when there is a crowd which is uh coordinating towards pointing out what a certain event um could settle at. uh That always is better. uh I mean the theory is that the wisdom of crowd is better than individuals, individual experts, which kind of talk about it. So I think it's it started like that, and then you know the very famous who wants to be a millionaire.

B

um You know uh that that that we've seen growing up uh that channel and that series where you know there is an audience poll for questions which uh people are stuck about, that you, the participant is stuck at. uh There is a study that about 91 percent of the times the audience is always correct. So I think you know these are the forms of prediction markets that have been. uh You know the wisdom of crowd. To be honest, not exactly prediction markets that have.

C

B

C

B

Space, I think uh nurses and augur uh have been. You know the front runners to bring the idea of prediction markets uh in the crypto space. I think you guys have done some really fancy stuff uh some amazing stuff and leading the charts to uh bring these uh ideas in the crypto space. I think some really really good um stuff has been done so yeah. That's that's about the popularity now. Do you agree with that rafa? Do you think uh this is something which uh makes sense.

C

Yeah absolutely absolutely like I I mean, echoing uh what you said uh I mean voting on on elections has also been like something that was like on register since, like the 19th century um and then yeah, I would say that more recently, uh like the um I mean, I believe, even even prior to that, there were like uh some like people that would like bet on um trying to see who was going to be voted as the next pope.

C

I think that's one of like the oldest presidents, um but going like um a little bit like more recently and as someone like pointed out in the chat uh like there's, also robin hansen and robin hansen's theories throughout, like the 90s that I think kind of, like popularized, uh like the the term in general, as well as, like the the logarithmic market, scoring rule that I won't get like too much into details for that, but kind of like ways in which it would be easier to kind of like aggregate uh liquidity.

C

So this could be tradable and but yeah, I would say that kind of, like uh it was uh in trade in like the early 2000s. uh What kind of was like one of the more like recent and like popular uh like digital prediction markets um and then yeah?

C

I would say that that kind of like started creeping a bit more into like what we currently associate more with what prediction markets are and, and then yeah eventually uh like gnosis uh within, like consensus and augur, were the first ones that were kind of like bringing it to the crypto space.

A

B

A couple of more, uh I think uh I mean if you have time I would love to answer: how does it fit in uh d5 and what is the importance, as you mentioned, do you think.

A

B

B

The way I look at um you know, prediction markets uh for them to be successful. I think there are three uh main elements um you know when we talk about wisdom of crowd, I think the crowd itself should be diverse when they are uh coming on to uh you know predicting on a specific event, uh it needs to be a diverse set of crowds. If we have just degens coming out and predicting on, let's say uh you know, it doesn't really make sense right, so the crowd needs to be uh diverse.

B

So what happened I think uh to have borderless permissionless ecosystems are very critical, um so I think uh the first element for via decentralized prediction markets. I think that's the biggest uh reason why we need it. um The second uh big uh you know point for a prediction market to be successful is that there should be incentives for people to participate in the market. You know it shouldn't be, like you know, um just participate for the sake of it. There should be an incentive and also disincentive for not doing the wrong thing.

B

um So um again, um when users participate in prediction markets, um they get rewarded for making the right prediction. Now, where are these predictions settling and who is deciding? um uh What sort of rewards should be given so again uh in the default space on ethereum in the public blockchain ecosystem? These things are done on smart contracts and there can be nothing more transparent than that with respect to settlement of rewards which happen on chain.

B

I think the third.

B

Important important factor for prediction markets to be successful is uh the fact that people should make their own decisions, while uh you know predicting on an event and shouldn't be influenced by herd mentality. You know which typically happens on off chain words. uh When you go on twitter and you see a poll, you typically, I mean.

B

Unless you see uh what the results are, you try to kind of choose what others are saying um for that to happen once again uh on chain, d5 prediction, markets kind of uh you know the cost of apping into somebody else's predictions keeps getting higher uh when you enter into prediction markets, because the stakes keep going up and down. So I think these are the two three um elements which haven't been able to answer uh which haven't been answered by centralized counterparts.

B

uh The way it can be answered in the diva ecosystem- and I think that's where it really fits in more specifically for plot x, I think uh vr uh I mean the idea which we came up with, uh was to only focus on price discovery um and price action of uh crypto assets and to be able to provide that uh you know ability for traders and investors and also builders tech builders to be able to see uh the signaling of their uh crowd is thinking that a specific b5 asset or crypto assets are going.

B

So I think that's where we fit in uh sorry. If I spoke too much with rafa over to you to talk about it as well, no.

C

A

I think in general uh what is that?

A

The centralization is a fundamental to striving this in this uh in this space as well, because it had a lot of value. So I I think we can rafa when you want to to add uh some something. uh You can also start to explain a bit more of your project, so we can because issue already mentioned something about plot x. We can also explain about nazis, so we will focus a bit more on your own.

C

Projects sure absolutely I mean regarding how they fit in in d5. I think uh ish already like gave something that like echoes a lot, but uh what I will say uh that yeah, just like through permissionless innovation. uh Like permission, uh uh sorry, permissionless prediction markets managed to enable uh like yeah, having like um an opinion of like a larger crowd with skin on the game uh so yeah. I think uh they benefit defy in like giving a like more reliable predictions and then b.

C

Another thing that uh I believe was like pioneered by like taos tag, but that we have also like built upon is, I think, like prediction markets or in general things where you have like um skin in the game, make you like act uh more or like incentivize you to actually participate, especially when it comes to like governance and decision making. That tends to be sometimes somehow sometimes I I mean, as you will know, like maria sometimes like.

C

If you go to like the maker forum, then uh sometimes it can be like completely empty people, don't vote, and this and whatnot, and so so I think, yeah, that's one of the things that they will like bring the most of uh to d5 that basically they will increase participation, but moving on now to uh like gnosis and and like what we are doing. um So.

C

Basically, when it comes to prediction markets um already for a while, we stopped focusing on like aiming to ship like a final product with like a front end, and what we actually do is that we focus on like building the infrastructure and like the infrastructure and all like the tooling required, and to yeah to face all the challenges of uh of decentralized prediction. Markets right, which are something like ish has already mentioned, like uh yeah being like permissionless. So there's like a diverse crowd or issues with liquidity and whatnot, and so yeah.

C

That has led us to build like the uh what we call the conditional tokens framework. um So basically yeah. This is like a very modular, or you can even say, like low-level framework, uh for anybody to just like, go and like build their prediction markets right. The cool thing here is that, like you, can determine, of course, like the outcomes uh you can choose up to like 256 outcomes, and you can pick of course, like any kinds of like resolution dates.

C

You can select like whatever kind of uh oracle or arbitrator that you want so basically uh like. Even uh you could even use like, for example, auger uh as uh as an oracle or you could even, for example, just use like one giving an address and have like a centralized decision maker right and what has this kind of like uh matured into well? There's already like a few teams building like some very interesting prediction, markets on our framework and the probably the two biggest ones like one is omen.

C

uh So basically, omen is uh a product build on on this. On, like this, the conditional token framework- and uh this is uh hosted and maintained by the dx dao and some of like the interesting features about this- is that, for example, it uses like a uni swap style amm right. So basically yeah you provide like 50, 50 liquidity and then, like trading, can be kind of like seamless right. Then it uses like reality, oh as an oracle and clarissa as an arbitrator.

C

So basically it has its own like flavors. As into like what the team like wanted to have like the dx dao team, uh in terms of like their mission to create kind of like a decentralization maximalist prediction market and then another example would be like poly market that uh they built like their beta. Using um like the conditional token framework, uh we recently launched uh nosis impact. That is like a plugin that you can use um for.

C

Like snapshot, um you know snapchat, like the off chain, uh vote, signaling app and then basically yeah like it creates two markets uh uh that are like basically the same question with different assets. And then you can like calculate like a possible price impact right. So uh just to sum it up uh like what, um uh like general or like the brief explanation of like what we're doing uh when it comes to prediction markets. It's those is uh we're focusing on like building yeah like the core infrastructure and developer tooling.

C

So anybody, including if it's like, ish or like somebody in any other like tao or project um it can like just set up their prediction markets and whatever like flavor they want to in terms of like decentralization and whatnot.

A

Cool, it's amazing. I knew a bit about. I mean we, I remember last year when it was last year that you guys launched the multisig right.

C

uh So so yeah the uh the notes is safe. Multisig uh like yes, but like the original multisig uh has been around since 2017.

A

Remember that you should nice and what about plot x alicia, um I uh I I actually. uh It was the first time that I heard about your project, because I I'm not into into this this uh word um but uh explain me how I mean the idea came up and what is your competitive advantage.

B

Absolutely uh so we are fairly new maria just about three months old, so I guess a lot of people don't know about us just starting our journey here. uh So um I think what portex is doing. We've learned a great deal from gnosis and augur and we are trying to adopt a little bit of a different approach um to how we are approaching, how we are building a prediction market protocol and also an interface to kind of interact with. So obviously, the protocol is permissionless.

B

Anybody can build a user interface on top of it, but what we've done is that our current protocol is limited to settling uh prediction markets that are driven by an api. So we are looking at anything which can be settled via an api coming in from decentralized oracles like chain lake, um so these are only price um price asset prices uh which are being predicted upon. uh We've intentionally stayed away from.

B

uh You know, predicting on stuff, like um american elections and um stuff, like that, uh we feel um having a focused approach, which is very um native to the crypto space is what we thought might uh resonate more with users. uh So far, it's been a good experience. We've obviously identified some of the issues that um have been on the platform, with respect to obviously uh liquidity and a couple other areas. Ethereum gas costs have been uh paying uh for almost everyone, um but yeah. That's that's really the product mariah.

B

You are able to predict on uh markets like what is the price of bitcoin in the next four hours or by the day or one period. Similarly, any other erc20 token uh can be predicted upon, uh we don't add uh or control um which markets are predicted upon. There is a decentralized governance, on-chain governance layer where the community can vote on, which assets to add.

B

So if they need to add an asset like link, if they want to predict on link usd anybody, can start a market and start predicting, so we use an amm as well, so the markets are created using an amm along with three options: there's a long shot and a neutral position on every market. All three options and the pricing for those options are calculated based on an on-chain amm.

A

B

Follows the zero-sum.

A

B

So all three options settle between them, so um there is no concept of liquidity providers. As such, as rafa mentioned uh that you need, I mean the idea is that the players or the users that come in they predict on any and they provide liquidity on any of these options and it's a zero-sum game between the three options that are in front of you and they settle automatically without any intervention.

B

um So yeah, that's that's really what uh it's a little bit of a different approach. We are, we are exploring and experimenting with how it works.

A

But it's nice to see some differences, and I mean both of the projects are really interesting and I think uh I I am quite ignorant about the topic, but I think the approaches could be very well functioning. uh I mean gnosis is there for two years or three years now, so it it just. It has been being proven that he is functioning but also plot x.

A

I see a bright future, um we can go ahead and I want to ask um again I mean we already talked about the the the project and what you guys are doing. What are your competitive advantage and your, but I want to please mention a real use case of your project for them for the users I mean for the final users so very like, as you are talking with your mom.

B

Let me go for it, and rather you can add on to that. So I guess the top two use cases for us is crowd, driven um driven asset valuation. So, whatever asset is there um just like in the old days uh the ox the weight of the ox was valued by uh the audience which was voting for it. um In this case, we feel the real asset. Valuation can eventually happen from the crowd when the crowd comes in. um I think second is also a signaling tool. I think I think, of plot x.

B

The vision of plot x is to become one of the finest uh signaling tools out there for investors, which is also related to uh asset valuation, to be able to see how historically an asset has performed and how, in the future, the asset could go based on where people are placing their predictions. So I think these are the top two use cases. We are also exploring um the possibility of plot x acting as a hedging tool.

B

um There is a little bit of a capital inefficiency right now in how uh plot x is designed uh we're trying to solve that by some really cool ideas. um But yes, uh as a hedging tool is something which we see uh going forward. uh This could also be quite useful.

C

Yeah and well uh on my end, uh as as I mentioned before, like uh what I believe is like the use case and what makes like what we're building like stand out is that it's like the general low-level framework for everybody to build on know so, for example, even plottix.

C

If, for some reason, needs uh like to do something different, uh they can still go and use the conditional tokens framework and, as ish mentioned, let's say, use a chain link oracle or any kind of like a thing that can be like just translated like into like a signature or a transaction, and it can be used so basically yeah. I would say that uh I mean if I go into detail, for example, when it comes to omen, there's already markets on like isis and politics and whatnot.

C

But if we're going to like specifically about like uh gnosis, uh it is basically that that it's uh like yeah, like a tooling that everybody can use and everybody can like customize.

A

Cool, thank you guys, um and uh so I I also want to again speak with our moms and- and I want you to uh imagine that you are like in front of the computer and your mom um comes, and you know a tea and back up a cupcake and say and tell you hey guys, I I brought you some some cupcake. What are you doing? What do you answer.

B

Go for it, I would let you go for it.

A

So what are you doing you say mom. I.

C

Oh, I tell her I'm joining these calls to like answer this type of question.

C

So, basically, what uh what I would tell or what I tell because I I still talk to my mother from time to time. No, what I would say is that basically yeah we're building like the building blocks, to create this very powerful tool that will enable people to like aggregate information better make better predictions uh and, on top like uh have yeah like a stake or like something to lose. Whenever they're not doing the right prediction.

B

Ish, so I think I would. I would tell my mom that mom I'm trying to build a tool that can help you. You know, plan your finances and kind of uh you know tell you how much a potato is. Gonna cost you in times to come. um So that's what I'm gonna tell her. I think guys.

A

And then your mom asked him so guys are you able to predict with the next president of the usa?

A

I want no. This is my next question, which.

B

A

Please connect, I mean, please comment um your work related to the news like the for example, and now we had this uh crazy election in the u.s or also the covet situation. I don't know, is there a connection? How is it connected and uh who wants to start? Please go.

C

I I I'll take this one so yeah, basically uh so something that happened when it comes like. If you specifically mentioned like the us election uh well, volume across uh all prediction. Markets was quite high, included like uh gnosis uh or well like omen, which is built on gnosis, uh but also like on centralized venues like ftx, uh also like in august markets and whatnot, and I think what is like the most interesting to see. This uh is uh yeah to kind of like realize how, especially in an event like the u.s elections.

C

That was something that was very like 50 50 in terms of like nobody really knew what were like the actual odds for something. What I think was very important was that, like maybe in this situation, prediction markets like if you were looking at them before the election, weren't telling you like something that was very relieving afterwards.

C

But what I found the most interesting, especially as you were, watching, like all the different like uh pricings or like probabilities, that prediction markets were pricing throughout like different like protocols and venues was that basically, they were like aggregating. The information like in a way faster, like way faster than any other kind of method right, like the normal polling that people would do or this or that would like take ages to produce kind of like an effective update.

C

While we would always be seeing um like whether you were checking like an omen or ftx or, as I said like, if you went like cross venue, this was even like more interesting uh because yeah, you would see, uh I don't know there was like whatever news or this state came in, and you would immediately see like the reflections on like prediction, markets, pricing like uh changing and so like yeah. I think uh in this election like they proved to be uh yeah. Quite an interesting and amazing tool to just like aggregate yeah.

A

C

Like sentiment and information.

A

Okay, do you want to add something english.

B

Yeah, I think, uh rafa you really summed it up really well. I was closely monitoring how all the prediction markets were reflecting the results of the u.s elections before they were declared. So I think some really strong use case on that side. um Now, from from a platex perspective, we are not really um uh I mean for now. We are not predicting on the news. I think that's something which we've stayed away from.

B

Those are subjective items where uh settlement is a little bit of an issue, so we've tried to kind of uh make it easier for us to design uh price based uh settlements. So I think, uh from a news perspective, the kind of markets that can run on plot x's. What will be the impact of where will be the price of you know, bitcoin. If we then wins the uh you know election, something like that.

B

I mean those are kind of which can run on plot x, but we haven't uh tried that angle, so I don't have much to say about it, uh but yes, I've been closely monitoring it and monitoring it and I think um yeah they do have a really good impact.

A

Cool, thank you guys. um There is also a question from the audience which actually was the next question that I was thinking to make, um and it is about the challenges that you are facing uh and now and what do you think are the main challenges in the future. You already mentioned liquidity but like let's elaborate a bit.

C

Yeah, well uh I mean when it comes like uh to liquidity, uh basically yeah and I'll. Try to like make that like something like very uh in, like a very simple analogy like uh basically in the end, all kinds of markets need, like a buyer and a seller right like if you go to like the market to buy a potato, you need somebody that will be willing to sell you a potato right to continue with, like the the potato example.

C

So I believe that, like with uh amms like uni swap and whatnot uh like, we have managed um already, like kind of like a way in which it's guaranteed, that people will be able to like participate and like buy or sell either like a yes or no token or an above, whatever price token and but like uh what, I think is an additional challenge to bring like this liquidity, and this might be like slightly technical. But I think it's still nonetheless very important. Is that especially like with binary uh prediction markets?

C

uh It is kind of like difficult to and to replicate like the payout that that binary prediction markets do so. Basically, uh what happens when somebody is providing liquidity, uh whether it's like potatoes or like amazon stock? Is that usually they like to have like a mechanism in which they can, like? You know, protect themselves from like forms of like price variation? No, and they usually do this like with some other financial product uh or well.

C

If it's like derivatives, they can like take opposite positions and whatnot, and so basically, I think one of like the challenges that especially like binary prediction markets are facing is that they represent like this payout. That is similar to like a binary option, and since that is kind of like difficult to replicate and liquidity providers tend to have uh yeah issues using like a more financial lingo like hedging their their liquidity right.

C

So I think that is like uh one of like the the core issues that is like limiting uh like people from joining as a liquidity provider, and then, eventually, that makes like prediction markets, not so liquid, which then limits, like both the participation and like the uh the value uh that they bring. And so I would say, like that's the main one.

C

Apart from like what we have mentioned, um that is kind of like external of like what uh plot x or gnosis or any other prediction market does is like the ux and like the costs of uh of ethereum.

C

And nonetheless, I think there is like quite a lot of uh progress when it comes to like new synthetic assets and payouts and whatnot, and I think a lot of them um are going to like help us yeah, like protect liquidity providers like make it actually more attractive to uh to just like, go and make a prediction market usable.

B

Yeah, absolutely, I think I completely echo with uh you know what rafa just said uh from a liquidity perspective.

B

Prediction markets are typically capital inefficient and I think that's what we are trying to solve for by doing different implementations of.

C

B

How how they settle and how liquidity is attracted? I think rafa briefly touched about the other issue, and- and I would like to kind of expand on that um prediction. Market is only as successful as the number of people who participate in it, because you know the crowd. Wisdom is only as good as uh I mean. If five people are trading in millions of dollars or a prediction market, it doesn't make any sense right.

B

So it is obvious that uh the transaction amounts are going to be small and the volume of transactions need to be a lot more to be. uh You know effective uh in terms of the settlement of where the prediction market is going for that to happen. I think the underlying infrastructure um is a little bit of an issue from an ethereum perspective, um the uh the standard transaction.

B

uh You know, transactions which are happening on plot x for at least that's what I can tell um at times the gas cost is more than the transaction amount which the user wants to do. So uh that's a little bit of a challenge for us, and we've tried hard in terms of optimizing. The smart contracts on ethereum um and also experimenting with a lot of other stuff- it just doesn't work right from now it doesn't work.

B

um So I think uh the challenge that we're facing is the scalability of the underlying infra, for which we are also looking at moving to an l2 or exploring with zk roll ups. I think some couple of things that we are doing on that side, apart from the liquidity problem, which happens to be the most important one as well, so I think these two are the most uh critical ones.

A

Yeah, thank you guys, really um insightful. uh I think now we can go uh further. uh Wait! There's another question. I think so another question from the audience: it's. um Why is it important for a prediction market to be decentralized?

A

uh We already answered, but, let's repeat with permissionless ability, people to participate. I think each already mentioned it, but uh if we want to uh quickly uh give an answer to this well.

C

um Yeah, okay, well, uh like previously mentioned it, and I think like in in quite like an accurate way that basically, what what permission does uh permissionlessness or like decentralization really brings to prediction markets is that it enables like uh first, a large crowd to like attend and then second that diverse crowd, because if you just have like a small population section from uh let's say, let's say you're trying to like predict the elections and then you're. Just like asking people from you know: inner city, los angeles, yeah,.

A

It's biased, it's a it's an auto profiling. I mean it's a like in every statistics.

A

As far as you can go uh on with the sample, the uh the better for the for the p value, or whatever of this statistic.

C

Exactly in the end, it's just like a matter of like the law of large numbers right, like the more uh the more like elements you have like in yourself, well, uh the more relevant like statistical results you will have uh but yeah. I don't know if you have like anything to addish, because I think you had already said it. Yeah.

B

That's like right absolutely and uh apart from the diversity of the crowd, I think the second point which I touched upon was the transparency in settlement of the market. I think uh that's a very critical, a centralized exchange. You never know how they settle uh the market and uh they typically decide the algorithm. For you know, settling the market also.

C

Another element of.

B

Prediction markets is the probability.

C

Of an option, pricing.

B

That that are available in front of uh you know the public, so in case.

C

Of permissionless.

B

Ecosystems, open source ecosystems, the calculation and the formula for uh calculating the probability and the price of an option is transparent, which is not the case with centralized uh systems. You don't know how the probability is being calculated. So I think these are the two main reasons why a permissionless prediction market makes a lot more sense.

A

uh I think now uh we are ready to have if you guys want to show a bit uh something uh of your your project. I know you should have a demo to show.

A

Okay, cool, please: let's go ahead and then rafa you. You want to try again.

C

We're having some issues with the screen sharing, but we'll give it a go after planets.

B

I'm going to do a very quick screen share guys. We are looking for active feedback to what we've built and would love to uh I mean if you have some questions and if you have some feedback, uh please do uh give us. So um can you see my screen uh rafa and maria yeah? Okay sounds good sounds good, so this is where you land. When you go to app.lotics.io, I'm gonna quickly. Take you through what this screen looks like. um So these are the assets that you can predict on.

B

At the moment we have markets running on uh bitcoin and ether. If you want to add more markets, uh then, as a community member of plot x, you can go to our governance section, uh which is an on-chain governance section where you can create proposals in terms of which uh assets. Do you want to add uh if the community votes on those um I mean, if they like uh to add those assets to be predicted upon uh the assets automatically will start coming here.

B

So there is no manual intervention with respect to how the markets are decided upon or created in the beginning, once the asset appears here, the amm of plot x starts creating uh markets on a four hourly basis, daily basis and weekly basis.

B

So we are experimenting with uh the the time slots of market uh markets guys earlier we had a one-hour market, but we realized um due to the uh scaling issues on ethereum our users when our unable to get the time to enter a market, because at times you know the gas cost would be so high that they wouldn't be able to get into the market and then eventually the market would go without any liquidity. So we stopped them. So now we have a four hour daily and weekly ones.

B

uh We might also experiment with the monthly markets, but the idea here is to kind of have uh consistency in the market duration. So we don't want to. I mean the earlier idea was.

C

B

You know to create markets which are six months long, for example, uh to to predict or predict upon the next american election or the american president. If the market is six six months.

A

Or one year long, so our.

B

Thinking was it's a little um and the ux for the end user is a little difficult with respect to when the market is going to get settled and they would like to trade their options in the beginning. um So these are the market uh timings that you see on the daily market. This is the live one which is running and the time to enter the market is 1 hour, 30 minutes, after which the market moves into settlement. So um the the market is fairly straightforward.

B

What will be the value of ether usdt uh tomorrow, which is january 16th at 9 30 pm my time so whenever you log in on the um on this link, this time will reflect.

C

A

Sorry guys, I was muted, disappeared, no yeah, I think he's reconnecting.

C

And we can give him.

A

Yeah, okay, so he disappeared completely. Let me invite him again on the screen.

C

Well, if you want meanwhile, there's, I see a question here.

A

There are two questions both about oracle's. Actually, the first one is here in the chat and the second one is: um why is important for a prediction market to be decentralized, with permission, less ability for people? No sorry, this is the improvement. So how can you define defend against some malicious oracle that is feeding wrong information to the system.

C

uh Yeah, well uh I mean there's, there's different methods, no like, as I mentioned when it comes to like the way like our framework does. It is that it lets, like the user use like any kind of like arbitrator, that they want right uh so, like uh there's different methods that have been used uh when it comes to there's like a centralized solution where you could just have.

C

You know like uh an entity that will just like uh yeah sign, what's the correct outcome, uh but then there's also like uh what augur originally did that there is kind of like a dispute system where, like people just like, take like their um their assets and then define like uh what uh um and through like that, like staking and voting process, and you define what is like the correct outcome right similar to like how clarisse does it, um but yeah I mean going back, I think we can go back now. That is she's back.

C

I was just trying to questions.

C

We can, if you want, we can jump like back to your demo.

B

B

I don't know where I lost the connection, so please.

C

We're scrolling down like through the charts- I remember okay, so.

B

Yeah, so the current um participation which, which is on the market, is around 210 dollars. It's a zero-sum game between these three options that you see, the options that are created are created automatically using the amm that is programmed on the smart contracts. So it looks at the price of ether usdd um at the time of market creation and then looks at some technical parameters for creating three options uh in which people can predict. So this is a short option. It's a neutral option. Then there is a long option.

B

uh Then there's an option price for all the three options that you see and you can simply click on predict. uh It talks to you about what amount of prediction you want to make. You can make the predictions in plot token or ether or b plot, which is uh also an erc20 token. um If you choose to do it in ether, it shows you the balance of your metamask.

B

I don't have plots right now um and you just enter the amount.

C

B

Ether you want to spend on making that prediction and click on predict. So that's how you enter the market at the time of prediction. It also shows you: what is your potential winning um for that specific market? If you do win- and this is obviously different for different options, so in this case, majority of the predictions have been either made on the short position or the long position, so the neutral position is the one where the winnings would be maximum.

B

So if I, for example, uh predict with point one, it shows me uh that my potential winnings could be 175 dollars. If I do win it so um yeah, I guess that's. That's really what uh the interface looks like with respect to predictions, uh then there is a leaderboard where you can see what markets are are under. You know which stage whether they are live or they're in settlement or settled um the total predictions that have been done.

B

What sort of participation has been there and then there is a my account section as well, which talks to you about what is your account with respect to what predictions you've done so um yeah? That's that's really what the interface is about.

B

A

Think amazing, and I really appreciate the very nice and easy ux there I mean. I think it's very important to have something that um people understand. Users understand quickly- and um this was uh was really nice to see, and I mean pretty straightforward.

C

It's not working right, you're, not sharing, okay. Well, then, I think all of you can check it out, and I mean uh what I think the dxdo and the omen team did great is that they made it like uh yeah kind of like very like pure and adaptable to uh to all kinds of questions.

C

uh So please go and click on it and you'll see like that. You can sort the markets uh out of like categories. uh You can also like filter how many markets you're seeing and whatnot, and once you click on the market yeah, you can see like the different outcomes that you can buy. You can see like which oracle or which dispute mechanism is being used.

C

um So yeah, please, go ahead, take a look and uh regarding um yeah, regarding that small question there, like you, can do binary markets and then all kinds of markets as well. Oh thanks a lot.

C

um So yeah I I mean there, for example, can you please click on that? One um ish like the first one um that uni one.

C

So, basically yeah here uh you have like some general information. This again, it's like a price market. uh There's like two markets on uh on this specific question so like, for example, you can change the the currency you're trading it on. This is like a price.

C

uh uh This is uh regarding liquidity, mining on um on uni, so for example, this is quite uh one of those like very nice and exciting cases, because this is like a prediction market on a specific or a possible like governance outcome, uh so you know everybody or every different participant can play it like in a different way.

C

You know, for example, if uh let's say you're like a vc with a large allocation of like uni, uh then you know you can have like some additional information uh that might help you like take the right answer and at the same time you like aggregate information for like most people and then also, if you are like uh yeah like a uni holder, and you don't want it, then you can also like vote no and whatnot and and then yeah.

C

It gives you some like general information as on like, what's the current liquidity, the volume and, as I mentioned, like the oracles and whatnot uh so yeah, I mean that's uh just one of the different uh protection markets that have been built on top of the conditional token framework, so check it out and I can send some other things yeah so that question graham oh yeah, you're asking to er ish so I'll. Let you answer that one.

A

So there's a question.

B

Yeah yeah, I saw the questions excellent questions, graham thanks a lot for those uh yeah. Absolutely I mean it's not about us offering it. Graham, you can go on the plot x, uh uh governance and uh you you can. You can suggest which market you want and just provide an api. So on trade volumes I think it's an excellent um market to have what will be the trade volume of bitcoin, probably um by a specific date, so you can do that just provide us with.

B

I mean just provide uh the api link, which you would uh want to settle it via or if there is a decentralized oracle available for it, and it's that's it I mean the markets would start getting created and we could we could work on it. Yes, absolutely.

A

Cool, so I think guys we can wrap up. We perfectly fits in one hour and I would like to thank you a lot for this great conversation and, as usual, I learned a lot.

A

I used to ask friday branches with the topics where I'm not at all uh informed or like expert, so I I can learn myself and so please, if you want to suggest a new topic or you know somebody who would like to speak in the friday branch. Please let me know- and I want to thank again the guys that today spoke who are ish goel from plotex, ceo and um rafael, um suarez from moses business development. You can reach them out on linkedin and you can reach me on wherever, because I'm happy.

A

I will not hear my uh telegram my telegram handle if you guys want to put here some link as well. Please go ahead and that's it okay! Well, you can share. The link, will be the recording of this talk. So if people uh in 10 minutes uh click on this link, they will uh they will find the the recording.

C

That's perfect, okay! Well, thanks a lot maria and thanks for everybody attending and great meeting you ish and yeah.

B

Absolute pleasure meeting you as well rafa thanks a lot maria for this opportunity had a lovely day and.

A

Don't forget to uh to us and thank your mom for the great.

B