Bitcoin desire sentiment has fallen to an eight-year low, a contrarian signal that historically aligns with market bottoms rather than tops.
Michael Sullivan, an engineer and author, applies machine learning to individual X accounts to track granular Bitcoin emotions and moods over time.
He examines how entry eras shape lasting narratives, why pro-BIP 110 cohorts show strikingly low conviction, how individual tracking avoids bot pollution, and why boredom plus infighting often mark optimal accumulation zones.
Timestamps:
01:44 — Conviction Isn’t Bullish or Bearish
07:05 — Why Individual X Tracking Beats Bots
09:53 — Desire Peaks Flag Bull Market Tops
11:49 — Bitcoin Desire Hits 8-Year Low
16:27 — New Bitcoiners Angriest Right Now
21:00 — Bitcoin Entry Era Shapes Your Views Forever?
22:48 — BIP 110 Backers Show Strikingly Low Conviction
25:33 — Pro-BIP 110 Group Lives in Its Own Bubble
28:39 — BIP110 Brigading Creates Fake Consensus
31:57 — OGs Optimistic, Plebs Stay Angry
37:13 — X Algo Shift Sparks Bitcoin Optimism
39:57 — Why Sentiment Metrics Fail for Trading
41:55 — Boredom and Infighting Signal Bitcoin Bottom
Links:
- Michael’s X: https://x.com/SullyMichaelvan
- Bitcoin Sentiment Weekly on Substack: https://sentimentsully.substack.com
- Michael’s novel, Blood of the Bourgeoisie: https://www.amazon.com/Blood-Bourgeoisie-Michael-R-Sullivan-ebook/dp/B0FRZTM49Y
Stephan Livera links:
- Follow me on X: @stephanlivera
- Subscribe to the podcast
- Subscribe to Substack
Transcript:
[00:00] Stephan Livera: Hi everyone, welcome back to Stephan Livera podcast. Joining me on the show today is Michael Sullivan. Michael has been doing some really interesting work on sentiment analysis, and I believe you actually also authored a book as well. I don’t know as much about that one, but, definitely wanted to chat with you about the, the sentiment analysis and what you’re doing there. but yeah, first of all, welcome to the show, and, yeah, let’s, let’s hear a little bit, from you on, on your journey
[00:27] Michael Sullivan: Yeah, thanks so much for having me, Michael Sullivan. Pumped to be here. so this sentiment analysis thing has been a huge rabbit hole for me, but a little bit of my background is that I’ve been an engineer for like 13 years, I’ve been a novelist and an author for about a decade, and I’ve been really obsessed with language. Like, I’ve always kind of noticed the words that people choose to use, how it evolves over time, these kinds of things. And then during my time in Bitcoin, I also have kind of realized like how much narratives play such a
[00:57] Michael Sullivan: The kind of narratives that evolve around tops, the kind of things that people get angry about and latch onto around bottoms. I just always thought this was interesting. So as of, about like six months ago or so, I decided to just like really analyze some of my own language, specifically looking at my ex data over time and see if there’s any trends there, see if I said anything interesting, and kind of look at some of the dumb stuff I’d said around tops, some of the, the things I was talking about around bears, and figured out it was actually very fascinating, and I’ve kind Like general emotions and moods of the crowds that we can kinda get into if you want to, but it’s been really interesting, and just tons of depth there that I didn’t anticipate when I first started doing the work, but it’s been really, really fun.
[01:38] Stephan Livera: So what was counterintuitive? What was something that you didn’t expect?
[01:44] Michael Sullivan: So the granularity of different emotions has kind of been something I’ve gotten much deeper into as of late. Right away when I first started doing it, I just kind of looked from like a broader, like the way you do sentiment is like with machine learning, basically you can see if an individual chunk of text is more optimistic, more pessimistic, these kinds of things. So I did it in a really basic way initially, where I was just kind of looking at people are happy Happy or sad, sort of like the fear and greed index where it’s really rudimentary, it’s like just a binary, yes or no, and there’s just like one emotion. But what’s been so fascinating, it was kind of counterintuitive, was how granular different emotions can get and how they don’t always interrelate in ways you might expect. And to give a more concrete example of that one, the conviction is one of the most interesting emotions or moods that I’ve looked at because it’s non-directional, whereas like people can be really convicted around tops, people can Convicted around bottoms, people can use more hedging language and be like more, less convicted around these same time periods, and like how those things interrelate with other moods is something that was not obvious to me right away, but it’s been one of the most fascinating parts of this work.
[02:53] Stephan Livera: So I guess let’s put that into Like in practice, does that just mean people might get overconfident in the bull and overbearish in the bear market? Like, what, what, what, what are some typical things that you will notice in, when you analyze that data?
[03:09] Michael Sullivan: So there’s like a reflexivity kind of to sentiment or emotions or narratives where, when people are optimistic, they’ll disproportionately latch on to things like the strategic Bitcoin reserve is one of the bigger examples of this around the end of twenty twenty-four, twenty twenty-five, where people talked about there being like a persistent nation- State buyer. and this is the kind of thing that people really latch onto and have higher conviction around. During those times when price is good, everybody’s more optimistic, and then the same kind of narratives, even though that’s still moving forward right now a little bit, and there’s still stuff coming out, Besant’s talking about occasionally, is almost dead. There’s like negativity around it. The same exact narrative, even though it has changed slightly, it’s, in a slightly different context now, but like there’s a ton of negativity to around it. So that kind of thing has You can actually see the same narrative evolve based on where the crowd is at mood-wise and how much the crowd’s mood does relate to the price, ’cause there’s definitely a huge correlation there.
[04:06] Stephan Livera: And in terms of methodology, are you mainly doing this on X? Are you looking at other, you know, Bitcoin Reddit, Bitcoin Talk, general, news?
[04:18] Michael Sullivan: So I went into this, originally wanted to do it for podcasts, ’cause I’m a huge fan of podcasts in general. I’ve consumed a lot of them, learned so much through ones like yours. And what I came to pretty quickly is that there’s some technical limitations of doing it this way. You can absolutely still diagnose like chunks of sentiment. I, I know some people that have been doing the similar things where they’ll look at how somebody’s sentiment tracks over the course of a podcast, what they’re talking about. But the problem is, like, imagine Michael Saylor, ’cause he’s a One week, and then he won’t do one for another three weeks, and then he’ll go on a big podcast run. There’s like four that all come out within a period, and like, you have basically no insight to what he was talking about in that period of time there between those chunks. What’s really, really valuable about, Twitter data is that people will just post almost every single day, sometimes multiple times a day, and consistently, so you get really good insights to like, where their mood is at over time. I’ve always thought about this kind of research in a similar chain stuff do, where it’s like the, being able to really track a UTXO over time with granularity is one of the things that makes such a valuable way to, it’s just such a novel, cool way to explore markets, right? I applied a lot of those same ideas to sentiment analysis, where because we can track a person’s mood over time and they’re tweeting almost every day, there’s a little bit more variability there, it’s not quite as pure as the UTXO set, for example, but because of that, you can really get granular and see how stuff is
[05:47] Michael Sullivan: Posting every day and they’re more optimistic, you can set a baseline for their average optimism levels and see when that optimism goes higher or optimism goes lower, and then look at that over time.
[05:58] Stephan Livera: Would you say this analysis is English or Western biased. Like as an example, there’s probably not that many Chinese Bitcoin people on X, right? As an example.
[06:08] Michael Sullivan: A thousand percent it is. It’s a hundred percent, central on that. The way– So I didn’t invent this style of analysis at all. Like this has been out there for quite a while. People have done, like, the machine learning models that I’m actually using to look at every individual chunk of text have basically been trained on Twitter data specifically, which is why it was a pretty good marriage here, but also on Reddit stuff, Amazon reviews Sentiment analysis in the past, but this is almost all English language, so it, it very much is biased to that domain. Fortunately, it is considered to be a lingua franca or whatever, so like there is a pretty good amount of people in other countries that are speaking about it in English, so it does work well that way. But there’s definitely opportunities to explore this more broadly. And then I’m curious,
[06:52] Stephan Livera: are you seeding it based on, let’s say, known individuals? Like you pick, you know, Michael Saylor and some other well-known people and then kind of seed it based on Are you just like looking for people just talking about Bitcoin generally?
[07:05] Michael Sullivan: So this is actually where a huge part of the work has happened is in figuring out the who and how to cohort them. where ag-again to reference how Fear and Greed does it, because that was always kind of the baseline people have conceptualized with sentiment, they did something where they basically just scrape the broader web, they’ll look at all these sources like Reddit for just like how much people are talking about Bitcoin, what context they’re doing it in. That has some really Terrible limitations in the sense where there’s all these bot accounts out there now. It’s never been worse with AI. There’s also these accounts that are incentivized by the Twitter algorithm to just like post for
[07:41] Stephan Livera: engagement and GM and whatever.
[07:43] Michael Sullivan: A hundred percent, that stuff is out there constantly, and that really pollutes your dataset, and that’s part of the thing that gets picked up in some of those indices. So what I took was the kind of opposite approach of this, where I did it from exclusively an individualized basis, so only picking up tweets for like individual folks and then looking Looking at their mood over time like I described before and seeing how that changes, and then kind of cohorting that together with other people that are of a similar mindset and looking how them, how their perspectives change, and that gives us a lot Pure form of data in a way, because you don’t get those bots, you don’t get the engagement farming accounts, and you also have a little bit less, influence by algorithms. So for example, like if, if something’s going around the algorithm one day, and then like the, the who would totally shift ’cause like a different group of bitcoiners might be more favored by the algorithm. But if you look at them in the sense of an individual, you kinda get that same person over time, which might be biased by the algorithm in their own little way, but you still get that person
[08:41] Michael Sullivan: Of the new monster, this is led.
[08:43] Stephan Livera: Interesting. And so,
[08:46] Stephan Livera: you were talking a little bit recently about, the concept of desire and low desire being a strong buy signal. So what are you seeing there on desire?
[08:56] Michael Sullivan: So I kinda went into this, I described how I’ve been doing this, but I, after I got this dataset, I was like, “Alright, I’m just gonna dive in and see what’s there, what relationships exist, what correlations are out there.” I’ve definitely been approached by people that want me to like give off the data for like individual day trading things to see if stuff’s spiking. And, but the thing that’s been most interesting to me personally is looking at like broader emotional regimes, ’cause if you zoom out, you can really see periods of time High levels of anger, for example, and, and that’s the kind of like broader lens that’s been most interesting to me ’cause it’s like a broad market psychology kind of thing. but one of those, the maybe the most interesting one of those was desire. and, and what to, to describe what desire is, ’cause again, there’s a ton of granularity with moods, that’s like people expressing wanting language. It’s, it’s like, it’s an emotion that’s not very based in human action. It’s kind of just like, “I want this thing
[09:53] Michael Sullivan: It’s very, very commonly peaking, this kind of wanting language, towards the end of bull markets, like it really is– like looking at the chart historically for desire over time over the last like eight years of Bitcoin, it peaked the most towards the very end of bull markets preceding runoffs. It’s not like some perfect predictor with like a hundred percent hit rate or anything like that, but the increased levels do flag there, and then the inverse of that is Oftentimes, marking of like relatively lower ranges for Bitcoin. And again, I think it’s kind of embedded into some of the principles of human action where, like, you don’t actually- If you’re just professing that you want something to happen, it’s very different from actually doing a thing, taking action, buying something, or stacking or whatever, you know? It’s kind of this emotion that shows up once you’ve already done the work and you’re just professing your desire for something to happen. And I didn’t, I didn’t go in with a story to tell around that at all. It kind of went like, I just looked this up, found this repeatedly in my research going back even into like the twenty sixteen era, I was like, “Wow, what is happening here
[11:01] Stephan Livera: That’s interesting. And so, I guess,
[11:05] Stephan Livera: just make sure I’m understanding you correctly. So, let’s say over the last eight years, you know You found that like in the, the top of the bull market, so like the kind of December twenty seventeen or the sort of mid twenty twenty-one or, let’s say October twenty twenty-five-ish or was it a bit, a bit above that, where we were like at those peaks, you’re saying that was where most They were mentioning desi– the concept of desire, and then typically what you were saying is on the other side of it, kind of the bottom of the bear market, so we’re talking like December twenty eighteen, we’re talking late twenty twenty-two, like, and maybe, maybe even recently, that’s when the desire is low. Is that, is that a summary of what you’re saying? Or a correct summary? You
[11:49] Michael Sullivan: spot on there. You spot on there. The, the Twenty twenty-five example was a really good one of this because it was just like peaking, it, towards the end of October twenty twenty-five, right before we finally saw the sell-off. right now, desire is actually at the lowest it’s ever been in eight years. people are really not talking about Bitcoin the same manner that they have. And, and this gets kind of weird and conceptual, ’cause it’s like psychological moods and that kind of thing, but these moods do different, do really change in different ways, like, and they kind of have different relationships to price at A good contrast here is something like excitement. again, the con– you asked me counterintuitive things I’ve learned from this, like one of the things is like how moods can relate to time. And one of them is excitement, where like comparing something like desire or optimism is a really good contrast here. If you’re optimistic, it’s like thinking way out in the future, like I’m very positive about the future of humanity or the future of myself personally. Excitement is an emotion that happens more in the here and now. Like if you have a vacation you’re taking in six months
[12:52] Michael Sullivan: In three days, you’re like really pumped about it. and these emotions definitely like flash in that way too, where you really see excitement spike very early on in bull markets historically, like right away. When we really, like the most excited peak I have in recent years is when we finally broke around the 100K range. Like there was huge amounts of excitement out there, everybody was talking about it.
[13:16] Michael Sullivan: Stuff like optimism usually peaks a little bit later than that, and then stuff like desire actually tends to peak after either of those two at the very end. And that kind of relationship’s really cool to me, ’cause I, I just never thought about it in that manner before I started exploring this more deeply. I was just like, “Oh yeah, vibes are good, people are pissed, they’re angry or whatever.” But there’s a relationship between these different moods, and they actually do evolve over time, and Bitcoin’s been just a really interesting, like, like- Group of people to examine to see those things happen, ’cause there’s just some, so much market psychology going on in this domain.
[13:50] Stephan Livera: It’s another question on the desire component. Is there anything in terms of what people are desiring? Is it just number go up or is it more like the, you know, Bitcoin price is really high, so everyone’s looking at, you know, house prices, the classic thing, or what’s, what’s that American platform, was it, whatever, that American platform? that people use to check pri-property prices. Is it like that kind of desire or what, what, what desire are we talking about here?
[14:14] Michael Sullivan: You can’t really delineate between them, right? Like, people, what’s kind of tricky about this is it, it can be subjective what you associate them with, where like,
[14:24] Michael Sullivan: there’s two different ways I can do this. I, I’ve looked at specific narratives, so you can just look at every time people are mentioning Bitcoin, are they using more excited language around Bitcoin or are they using more angry language around Bitcoin? And that But what I’ve found to be, maybe more useful, because you get more granularity of data, is looking at everything a person talks about because their mood isn’t just influenced, like They’re not just talking about Bitcoin all the time, people are pretty diverse in other interests. So if you look at every single thing they say on one of these platforms, you still get these mood changes, and by doing that, you do pick up on these more subtle things because, like, and initially you might not think like, “Oh, if you’re examining everything they’re saying, it’s not as tightly correlated to Bitcoin,” but it for sure is. There’s like, especially if you’re examining Bitcoiners who are primarily in these communities talking about Bitcoin almost exclusively, you do get this kind of stuff
[15:17] Michael Sullivan: Precise differentiation if it’s like them talking about desiring for a yacht or desiring for a new home versus just desiring the coin price to go up. I could do a deeper analy-analysis on that, but I haven’t yet. Gotcha.
[15:30] Stephan Livera: Now, what about anger? I know this is something that came up in your analysis, and you were– At one point, you were saying people are the angriest, Bitcoin is the angriest that they’ve been.
[15:41] Michael Sullivan: Yeah. So again, like with the desire thing, like that’s lower than ever, and anger is higher than ever. We’re at– I, I’m, I’m curious to get your thoughts on this too, because you’ve seen a lot of these cycles before, and I’m curious how it just feels to you personally, but just looking at the data, the current- Bear market is wild when it comes to the sentiment. Like, there’s low levels of desire, there’s quite a lot of boredom, and people are really, really angry. but not everybody’s equally angry either. Like, so, I mentioned the cohorting earlier, which has maybe been my favorite part of this. if you take people that have been in Bitcoin longer durations versus people that have been here shorter durations, the plebs, so to speak, are much angrier right now comparatively to the people that have been around longer. And those are always the most fascinating mood differences for me.
[16:27] Michael Sullivan: Because of the way that Bitcoin adoption works, the newer folks make up a very large chunk of the market, right? So like their mood presently does disproportionately impact things comparatively to the people that have been around longer. Not to mention that a lot of the newer folks, especially in the last couple years, got in through treasury companies, which are down like eighty, ninety percent in some cases, right? it’s, it’s, like, it’s kind of understandable to think through their lens and to empathize with why they might be really, really angry right now and so they’ve been pretty, pretty checked out. At least those are some of my thoughts on like why moods might be, might be bad, but the anger levels for sure, especially among some of those groups of people, are way higher.
[17:09] Michael Sullivan: Groups. I also did a, a fun piece a little while ago using Michael Saylor’s, cohorting that he did basically. So I’d already set up the
[17:17] Stephan Livera: capitalists, the technologists, the maximalists, and the fundamentalists or something like that, right?
[17:22] Michael Sullivan: Those are the ones, yeah. So as soon as he did that, I was like, “Wow, I’ve got my whole cohorting system set up. I have to experiment with this.” So I did that, and the angriest group of those is by far the fundamentalists right now, where the technologists They’re pretty darn optimistic and constructive right now, and the capitalists, so like the sailor folks, people involved in treasury companies, they’re actually, they’re, they’re like a little bit angry, but they’re actually like really high conviction generally and pretty constructive, whereas the fundamentalists in general that are by far the angriest and have the lowest conviction right now, which I found fascinating.
[17:59] Stephan Livera: That’s funny, ’cause this is obviously kind of like leading to, let’s say, current arguments of the day with Bip 110, right? Because it just to me, it kind of parallels a little bit, because of course, as Michael himself says, he considers himself in all of those camps, but let’s say- Most people see him as the capitalist, and he’s kind of the leader of the capitalists per se, whereas, you know, even nowadays, it seems like most of the technologists and most of the capitalists, like Michael himself and others, have basically– they’re kind of against this Pip-10 stuff. Whereas what you find is like the fundamentalist type, I guess Subjectively, I’m gonna place a bunch of the pro-one ten camp into the fundamentalist camp here. So that’s kind of my subjective, you know, finger-in-the-air read of what’s, what’s going on. What’s your reaction?
[18:50] Michael Sullivan: that is absolutely correct. That is more mostly the fundamentalists. So there’s overlaps between the cohorts, right? You kind of mentioned that Michael considers himself to be in all these different groups. I do too. I have a foot in pretty much all those camps to some varying degree.
[19:04] Michael Sullivan: but like I said, the cohorting was, has been the largest amount of work, as part of this because you kind of gotta figure out which groups people get put into. It’s somewhat subjective, so I’ve had to go through and like kind of delineate like which persons in which bucket.
[19:20] Michael Sullivan: Taylor made of how he very explicitly defines these things, so that I was able to like really craft a detailed description of what it meant to be like a newer retail pub or something like that compared to a fundamentalist compared to, BIP one ten, BIP one ten being the easiest ’cause they’re very, open about that generally. And then using that, you can actually put each of these individuals’ tweet sets through AI models and like have it help me figure out whether or not they actually do fit that description, if my initial cohorting was wrong, and then I go iteratively through Over and over again to kind of like make sure I’m accurately grouping them because, again, it’s kind of subjective initially. It’s really tricky to do this in a way that’s like perfectly consistent, but through a lot of iteration there, I’ve been able to like pretty well cohort these groups. I say just a lead into this answer of like, you’re spot on that those three groups in particular have huge amounts of overlap. I can’t remember the exact numbers, but BIP one ten is extremely highly represented among the fundamentalists and extremely highly represented among the pubs too, newer folks. Like the, if I had to give a range to the highest percentage of BIP one ten support, it would be people that probably came in around twenty twenty through twenty twenty-three era, likely. Kinda curious if that maps with what you’re, what you personally think. That’s kind of my, subjectively,
[20:33] Stephan Livera: yeah, I think so. I think that’s, and even Tonevay’s, in his kind of arguments back and forth, he was mentioning that too. He was like, “Uh, why?” He– because his point was, most of the people he was kind of, quote-unquote, battling, on the same side with in twenty seventeen, most of them were with him against one ten, whereas he saw himself as battling a lot of the one ten people nowadays, were the newer people from, let’s say, twenty twenty, twenty twenty-one
[21:00] Michael Sullivan: I, I think there’s an interesting dynamic here, slight digression, but like, I’ve, I’ve always thought that the era you come into Bitcoin really does like paint your perception of what Bitcoin is indefinitely, and, depending on how much you’re willing to change your mind about it and shift your perspective. But I, I, I’ve said that a lot in the past, but then, like, the more I think about it, the more I think like the first bear market you lived through probably also is really representative of that, and some of the narratives that went around the twenty And part of that myself to some degree.
[21:33] Michael Sullivan: I think some of those narratives really got latched onto there and created a, a little bit of the spirit that still lives on inside a lot of the BIP one ten cohort right now.
[21:41] Stephan Livera: Yeah. so give us your analysis you did while we’re on BIP one ten. give us your analysis on that. I know you wrote a post on this and, maybe you copped a little bit of flack for this one, but, what was your takeaway there on the BIP one ten analysis?
[21:57] Michael Sullivan: Yeah, so I, I went really deep into this one. I, I’ve been sitting on this for a very long time. When I very first started doing this work, this was like one of the very first groups I wanted to look at because it was clearly very divisive. and it’s easy to cohort, right? ‘Cause it’s people like are putting bit one ten in their profile, so it’s like super easy to figure out what groups they go into. And I didn’t say anything for a very long time because of how angry they were. I literally looked at them and I was like so eventually I came back around to it ’cause it’s just been, it’s building up, people are getting angrier and angrier, stuff’s getting closer, and I, I kinda felt like I was doing a disservice to just not share it openly ’cause I, I really have a pretty different way to look at these things and I have a lot of like data around some of these like social cohorts, so it’s like I, I just have to share this.
[22:48] Michael Sullivan: The biggest difference, and the thing I led to peace with, was conviction. So similar to the fundamentalists that we talked about earlier, I, I cohorted people into BIP one ten versus non-BIP one ten, people that are most divide, like, are heavily pushing back against BIP one ten publicly.
[23:03] Michael Sullivan: those folks generally have much higher conviction, so people pushing against BIP one ten, the people that are fighting pro-BIP one ten are much lower conviction. And what was interesting about these two groups is like you can see their mood lines, be kind of move in tandem pretty There’s some deviations, but in general, those two groups actually kind of like respond similarly to one another and similarly to things going on the Bitcoin and the broader ecosystem. And then right around October twenty twenty-five, there is a divisive split, and those two groups, like the
[23:34] Michael Sullivan: Pro-Bip10 gets much less, much angrier, and then also much less convicted, and Anti-Bip10 gets much more convicted over that same period of time.
[23:42] Stephan Livera: So you said October 2025 there, right? Yes. Now, to be clear, that was around the time, funny enough, Plan B Forum Lugano, right? And I was on stage at this time, right? I was the moderator of the panel, and Jameson Lop asks Luke Dasher this question of, “Hey, are you gonna do a consensus change?” And he gave a very vague, kind of evasive answer, and guess what happened? Like, I think pretty much that night or the next Within the next day or two, it came bit four four four at that time, called RDTs and later would become bit one ten, that was when it was announced. So I guess in your sentiment analysis, I guess you were looking at kind of not and the filter people because obviously that’s a very strong Let’s say Venn diagram with the bit one ten people. And so when you say October twenty twenty-five is when there was a divergence there, I, I presume that was around the The creation of, and the launch of that BIP, or no?
[24:38] Michael Sullivan: Precisely. Yeah, it’s like right around that time period. It’s, I, I tried to like not be super prescriptive around exactly when that form, ’cause there’s like lots of stuff going on around that time. Core V30 was around that time too. And what’s been interesting is I’ve shared this, is that people kind of take their own narrative and map it to the conviction data or map it to why the differences are happening. So if I tell one group about the conviction thing, they
[25:03] Michael Sullivan: One ten people about the conviction thing, they’ll say that the anti-bit one ten people are just like false bravado. So, and, and they’ll like, each all claim their own reasons for why this split happened. I’ve tried to be pretty like passive, like, “This is what the data says, it happens in October twenty twenty-five,” I’ll kind of let you map your own thing to it, but, that is for sure when that inflection point happened. And that wasn’t the only inflection point either. almost every single mood that I cover in the entire piece
[25:33] Michael Sullivan: Most fascinating thing, but also comparatively to the broader ecosystem of Bitcoin, the pro-BIP10 people really seem like they’ve been in their own little, separate bubble comparatively to broader Bitcoin.
[25:45] Stephan Livera: Right. That was another interesting point you raised in your post where you were basically saying, kind of, they got a lot of engagement on their posts compared to the people who were anti-10, and I think you were also saying they– the people pro-10 spent a lot more time on that issue compared to the anti-
[26:04] Michael Sullivan: Precisely. And, a lot of people yell at algorithms and echo chambers and say that it’s just like completely algorithmically driven, and I kinda reject that a little bit. I, I think that as much as we like to villainize algorithms, they’re trying to give us what we want. this is, particularly top of mind right now ’cause Twitter just changed their algorithms quite a bit to show us people we actually follow. Wild concept. It’s been amazing. but Just be– like, it’s not like the, the algorithm, which is, like, people often frame the algorithm as this like godlike thing that decides everything that you see, but what it’s really trying to do at the end of the day is show you what you wanna see. And I think that fundamental, like,
[26:46] Michael Sullivan: like thing the algorithm is– algorithm is doing is kind of driving this to some degree, but it’s not like intentionally making the people more divided. A lot of people in the Bip one ten camp are like obsessively interested in Bip one ten. A lot of people outside that Fighting against it have a lot wider and more diverse r-array of things they’re talking about, a lot more like from the treasury companies to like really technical discussions. It’s pretty wide, whereas generally the people talking about BIP10 really heavily talk about BIP10 and almost, almost only BIP10. So I don’t really think like some, some people thought when I shared that data it might be like bot farms ramping up BIP10. I don’t think that’s what’s happening. I think there’s just a lot of intense focus around just that single issue and that
[27:30] Michael Sullivan: On this, I, I covered this in the piece too, but, you know, if people that are deep into that BIP one ten cohort talk about BIP one ten, their posts get higher amounts of replies, higher likes, and higher retweets than anything they say that doesn’t include BIP one ten. The inverse is basically true, people that are against BIP one ten, so their audience doesn’t give a damn, in my opinion. Like, generally, there’s like, I, I covered one with Lynn Eldon in particular, ’cause she had one of the more interesting examples of this
[28:00] Michael Sullivan: Passionately following her, I’m one of them, I, I love reading her stuff, and she’s known for just like ratioing people all the time. This is like a common thing, this happens like, at least like once a week, if not more. So it was crazy to me seeing that, like, Lynn goes in these comment sections talking about Bible in ten, when she was kind of roped into this debate around it, and Mechanic and Luke are both ratioing her, and she’s getting almost no engagement. And again, I don’t think this is some like, Almost at all, where the Pro-BIP10 folks care about this issue really strongly and came in to watch and support their people and just disproportionately like it heavily.
[28:38] Stephan Livera: It’s so funny the things you’re saying, it really, it, it matches up with my own experience, right? Engaging with this stuff because sometimes, there are times where I’ve gone in to kind of colloquially fight a battle about BIP10, and I’ve, I’ve generally gone in with not with the expectation that I’m going to ratchet these people, because I know that, because I know they will go
[29:00] Stephan Livera: Group chats or whatever, and they’ll be like, you know, like that Matrix thing, oh look, the guy’s coming, he’s fightin’, Morpheus is fighting Neo, kind of thing, and then they’ll all parade in there and hit like and retweet on it, and then they’ll kind of, you know, start, kind of brigading, effectively brigading a thread, right? And that’s like an online thing, it’s not just a Bitcoin thing, it’s like an online thing, and so in highly engaged groups, that happens. And to your Lynn, think about Lynn. Her audience is obviously Bitcoin stuff, but also macro, like she has a very big macro base, and a bunch of those people might be interested in her take on, you know, bonds or the Iran war or like AI and tech stuff, and they’re not so focused on like BIP one ten, so they’re not gonna like get in the comment section or to kind of support her by liking and retweeting her or whatever, right? So it’s just a very different mindset, isn’t it?
[29:49] Michael Sullivan: And you kind of think about what that incentivizes. Like, it Like how I’ve fallen into this trap historically too, where I’ve had stuff that I’ve talked about that’s blown up more. And like, we all like to think we’re unbiased about these things, we just talk about whatever. But like, if you are constantly getting positive engagement and feedback from people, if you write about something, you’re more likely to write about that thing again, post about it again, talk about it again. The opposite’s also true. Like if you get canceled for something or people yell at you for something, you get a bunch of hate, you’re not likely to post about Perspectives. I, I’d like honestly to have done more of that discussion with friends and group chats and texting and whatnot, like deeply forming our opinions, because I know if I post anything about it publicly, I’m gonna get a bunch of people in my replies yelling at me for my perspectives, so it kinda disincentivizes queer,
[30:41] Michael Sullivan: Clear open debate around this thing, and I, I think that’s something that’s, that was maybe the most interesting finding of the piece, to be honest with you, like this is happening a lot, and I think that, in my personal opinion, the people that are really pro PIP 110 dramatically underestimate how many people that are against PIP 110 have strong opinions about it, but just haven’t voiced them as much ’cause they don’t wanna get wrapped up in all this heated debate. There’s a lot of that out there. and the other thing that might be
[31:10] Michael Sullivan: Holds a great deal of Bitcoin, probably. I kind of speculating slightly here, but there is a huge economic perspective around this, right? Like, like one note isn’t equal to one vote. Like, there is a very big economic factor at role-at play in this debate, and some of these people with really large audiences from a long time ago, like, they, they really are reflective of some of the larger Bitcoin holders in the space, in my opinion. So, that’s a big element that I think gets lost in this. It’s not just this like really Loud, smaller minority. It’s like, there’s a lot of elements here, and the economic thing plays a big role.
[31:47] Stephan Livera: Yeah, I think that’s it. on the question of OGs versus plebs, can you explain a bit of your thinking there on, on where they diverge?
[31:57] Michael Sullivan: So some of the most stark ones are with the excitement, optimism, these kind of things. I, I just actually looked at this like three hours ago, ’cause I, I kind of like digest the data, look through it again. There’s tons of compute that has to go into doing this, so I’ll kind of like run, run things overnight, look at it fresh in the morning. right now, we are having a huge divide, across like OG is getting more optimistic Not really excited yet, but, but less angry where the pubs are really angry. This is happening right now despite the fact that Bitcoin Twitter, in general, got a really positive algorithmic shift, which is part of the reason I wanted to update this and look at it a bunch because it’s kind of fascinating that the algorithm changed so much.
[32:35] Michael Sullivan: So it’s interesting that that’s still happening, ’cause I thought maybe we’d see a giant change and much more positivity. That’s not happening yet. that’s one of the bigger divides is around some of those moods. There’s, there, there’s been some specific historical inflection points that have really, his-proven out to be
[32:53] Michael Sullivan: I don’t wanna put too much of a story behind it or say that the OGs are always more accurate, ’cause it’s not like I’m measuring them saying, “Oh, we’re gonna have a price run tomorrow.” But there is definitely a historical example where you can see that, specifically in summer chop, chop, solidation, shout out James Check for the great narrative, around that period of time. lots of sideways action around the sixty K range, right?
[33:17] Michael Sullivan: Pwbebs got really angry, super bored. And towards September, the OGs started getting more excited, more optimistic, but Pwbebs didn’t, they stayed really flat. And there’s a couple inflection points like that historically where people that have more market experience start to get a little bit more excited when we run up towards the tops of ranges, and it’s not a hundred percent correlation with like accurately predicting anything, but it’s really fascinating how and when those two groups diverge because it kind of Like, like in any other domain in the entire world, like the longer you’ve been doing something, the better you get at it, typically, with some caveats, but like in general, like the more time you spend in something, the better you get at it. And maybe you don’t get better at trading Bitcoin inherently over being in it that long, but you do get better at stomaching volatility, recognizing patterns, and like this thing is so consistent and going up forever, Laura, over a long enough duration of time, like there’s certain inflection points where it does seem like people that have been They’re like less impacted by some of the negative things in the bears, they’re less impacted by the euphoria in the bulls, and they kind of are like a counter signal in a really positive way comparatively to the phebs, which oftentimes you generally wanna fade, like they get disproportionately angry at bottoms, disproportionately excited and optimistic at tops.
[34:30] Stephan Livera: Yeah. And I guess just for the definitions here. So when you say plebs, are we referring there to people who have not been around for a long time? Is that the main thing? Because there, there are other– ‘Cause sometimes people use the definition more of like just, you know, low followers versus high follow account. But to be fair, there are people who have like a low follow account and they’ve been around for a while, but they’re more like an anon. They’re not trying to be like a public name and face, et cetera. So how do you kind of
[34:59] Stephan Livera: My age and tenure, or are we going by, you know, stack size or perceived stack size? Like, what, how do you, how do you slice it?
[35:06] Michael Sullivan: Yeah, that’s a really good question. I, I, I do it by perceived time in Bitcoin, ’cause some of the most intelligent people that I follow have like three hundred followers. It’s, I really don’t think follower count is a good representation of dividing people in that way. I, I could do it, I haven’t done it yet, but like, but especially in the Bitcoin space where people have anonymous accounts
[35:29] Michael Sullivan: I just give up, I get a new account. I know they’ve been around for a very long time and have really intelligent opinions about things, but they just like to stay relatively anonymous. There’s lots of that, so I just don’t think that it’s a great way to group people. what I do is time in Bitcoin.
[35:43] Michael Sullivan: Basically, like the newer folks versus the older folks is essentially what I’m referring to. And this is tricky too on X, because like your account doesn’t accurately represent how long you’ve been around because of that too, where you might be getting new accounts, you might be switching the time. You might cycle
[35:57] Stephan Livera: through a few NIMs on the to– over that time.
[36:00] Michael Sullivan: So again, this is something that like AI has been really helpful with. Like, I still have a large subjective aspect to this where I have to go through and look at stuff in detail, but because I have so much of the individual tweet data and people have talked about historically, like what they’ve done, if they’ve been in for a specific amount of time, you can see when they talk about Bitcoin, sometimes their initial purchase prices, you can kind of divide these groups. And it’s not gonna be perfect because, like, I, not for– I can’t do this for some people. Like, A rough idea they’ve been on, they’ve talked about it a lot, so it’s much easier to group folks. then the newer folks are a little bit easier too, ’cause you can kinda see like, like the, the very often talk about when they entered the space and that kinda thing. kind of a long-winded answer, but like the, the, the super quick version is basically just the time duration, as I think that’s kind of the better way to do it comparatively to some of those other ways. Yeah.
[36:51] Stephan Livera: Now it’s come a couple, come up
[36:59] Stephan Livera: We are more likely to now actually see them, on our feeds. So what are your takeaways on that? Have you seen any big sentiment shift, like there was a lot of people saying Bitcoin Twitter is back? So what have you seen in the data?
[37:13] Michael Sullivan: Yeah, so I’m really pumped about this in general. I’ve been, loving it recently. It’s been fun seeing some folks I haven’t interacted with as much. I’m really curious how it evolves, ’cause X has made some changes like this in the past and then it’s reverted over time. I’ve actually already seen this start to happen to some degree where I didn’t see a single news item right away for like the first twenty-four hours, I’m starting to see some of those things filtering a little bit more. So I think it’s gonna be an evolving thing.
[37:42] Michael Sullivan: My timeline got significantly more optimistic after this change, but not everybody’s has, which is kind of like one of the tricky things about this, and one of the reasons I found this individualized way of looking at this much better, because it’s really, it’s really tricky to just like go and scroll on X and get an accurate representation of where all Bitcoin’s at, because we’re all seeing our little echo chambers, all of us are. We all have our biases, we all have our way we interact with this thing, we all have different people we follow. So it can be challenging to do that,
[38:13] Michael Sullivan: Again, I think back to the Bib one ten thing, if you can, if you like that a lot and respond to it a lot, you’ll see disproportional amounts to that, and you’ll start to think that that’s more of a thing than it actually is in the broader space. So by looking at individuals, you kind of remove a little bit of that noise, so you can still see how they’re interacting, how they’re commenting, how they’re talking, despite how their algorithm is changing. With that said, though, there is, there definitely a positive mood shift. I think specifically accounts
[38:42] Michael Sullivan: I didn’t study this super deeply yet, ’cause I’m kinda just like living through it. I wanted to like do a quick overview, but I haven’t like done a deep dive into this yet. It’s kinda just like my off-the-cuff thoughts so far. I’m almost certain though, that accounts that are older, because they’ve had more followers, because in like twenty twenty-one era, there was a lot more like engagement on Bitcoin Twitter in general, these folks have had a massive surge of interactivity again, contrasting that with some of the smaller accounts that might only have a couple thousand followers or
[39:12] Michael Sullivan: Decrease in reach comparatively, because there were some folks that were getting tons of engagement as they talk about whatever their pet topic is, and people kind of have ignored the followers, whatever. So I think there’s kind of like a shift that way. I don’t really have a great take on it outside of that, and kind of just observing right now to be honest. That’s fair enough.
[39:29] Stephan Livera: I mean, it might, it might be a bit early to have something concrete on that. So, I guess in terms of other ways that this analysis, you have to like, I guess,
[39:41] Stephan Livera: You know, could people spoof this kind of thing? you know, I guess, what’s that– I think there’s another one. I think it’s called Goodhart’s Law, like a metric that gets used and then eventually it stops being a good metric. So I’m curious how you’re anticipating any of those things?
[39:57] Michael Sullivan: Yeah, I mean, I’ve heard that. I think it’s likely to happen where all things eventually, like the market finds a way to eliminate any kind of edge that anyone has. I am just kind of studying right now, seeing what exists out there. I’m not being too, I’m not actively like not trading on this or anything like that yet, which I know is part of the thing I’ve, I’ve been talked to quite a lot about from people, but We’ll see how this evolves. There, there’s so much granularity to the moods, and there’s so much contextualness to this where, when I first started doing it, I was really looking for like some singular metrics I could watch, and I, I found a couple things that do seem higher signal that I’ve shared, like the excitement, desire, those sorts of things. But there’s a lot of noise in this data. Like for example, like if I was gonna go tell somebody that, like, “Oh, geez, we’re really excited right now, you should go buy Bitcoin,” like, I think that would be a really, poor recommendation because there’s so much noise. Like the algorithm just shifted. If I were to say that right now, like, there’s so much potential things that could cause that. So that’s why I’ve taken a much broader lens because there’s so much like day-to-day noise in the data. However, the broader lens
[41:04] Michael Sullivan: If the crowd is angry versus disapproving versus bored, like some of those things are interesting, so I’ve been, I’ve been looking at a much broader range framework rather than something that people are like, “Can you use this precise trading information?”
[41:16] Michael Sullivan: There’s also some stuff, narrative analysis too, that I’ve been doing, to see, and it could just be
[41:21] Stephan Livera: that, you know, as Bitcoin goes along, there are issues of the day, whether right now that’s Bitcoin ten, in the future it could be some other thing, and maybe that’s like a sentiment analysis topic that you could research and talk about and that kind of thing. So, it’s kind of, there will be, there’ll always be things that just come up in Bitcoin, right? ‘Cause it’s, that’s just what Bitcoin is. so
[41:46] Stephan Livera: Takeaways that you think, you know, the average Bitcoiner who’s trying to stack and stay sane, any, thoughts on how they should think about these sentiment patterns?
[41:55] Michael Sullivan: Yeah, so again, don’t wanna be too prescriptive in the crazy short term, but man, if things don’t look amazing when you zoom out, like, again, like almost every emotional regime that’s been reflective of bottoms historically is just like flashing insane buy signals right now. Desire being super low, people being bored and checked out, like optimism declining, the newer- People getting disproportionately angry and infighting, the divide between Bitcoiners talking about different things, like these all seem like bad things, but one of the most interesting findings of this is like the deeper I’ve got into this and looked at some of these bad moods, the more I’ve been like contrarian optimist, because like these, these things are always what you would expect to happen, deep in bear markets. And some of the most particularly bullish things are like the boredom levels, like people are just checked out, highly apathetic, and then the infighting, some of these things
[42:46] Michael Sullivan: Interesting, because Bitcoin is having some really positive momentum in more like technical metrics, like the rate of the decline slowing. this isn’t my specific area of expertise, but like historically when some of those things do start to slow, and yet still the retail people are more checked out than ever, like that is so represent-representative of like really fantastic entry points to Bitcoin. So if you’re somebody who’s dcaing or buying right now, I just think it’s a brilliant time presently. I don’t know what’s gonna happen in the short term, not claiming we’re about to like moon from Usually have some news item that drops and we teleport down to new lows, but man, I feel like it’s a really, really constructive time to be buying right now personally, no financial advice.
[43:26] Stephan Livera: Well, it’s fascinating analysis, I mean, the psychology, the narratives, the stories that people can tell when they’re looking at this stuff. So listeners go and check it out, follow Michael, his X is Sully Michael Van, and, Michael’s Substack is sentiment sully dot substack dot com. Michael, thanks for joining me today.
[43:43] Michael Sullivan: Yeah, thanks so much for the time, everybody If you want to follow this stuff more closely and just share my thoughts all the time, thanks for the time.