Attention vs Sentiment: Why How Much Beats How Positive
TL;DR: Sentiment asks how the crowd feels, which is noisy, ambiguous, and often a reaction to price that already happened. Attention asks where the crowd is looking and how fast that is shifting, which is measurable, harder to fake, and the cleaner market signal. How much beats how positive.
When people try to read the market from social data, they usually reach for sentiment, is the crowd bullish or bearish? It feels like the obvious question. It is also the wrong one to lean on. The more useful signal is attention: how much something is being discussed, how fast that is changing, and how broadly it is spreading. In short, how much beats how positive. This post explains why, and what attention actually looks like when you measure it properly.
Sentiment vs attention at a glance
| Sentiment | Attention | |
|---|---|---|
| What it measures | How positive or negative posts read | How much, how fast, how widely something is discussed |
| Question it answers | Is the crowd bullish or bearish? | Where is the crowd looking, and is that shifting? |
| Failure modes | Sarcasm, irony, memes, bot hype, reacts to price | Single-source echo chambers, raw volume without breadth |
| How easy to fake | Easy at small scale | Harder, especially with breadth across sources |
| Direction in time | Often lags price | Often leads or coincides with price moves |
| Best framing | Context, weighted lightly | Core signal, broken into volume, velocity, breadth |
Two very different questions
Sentiment and attention sound similar but ask completely different things.
Sentiment asks: how does the crowd feel about this? Positive, negative, neutral. It tries to read emotion.
Attention asks: where is the crowd looking, and is that shifting? It measures focus, not feeling.
That difference is everything. Emotion is ambiguous, easy to fake, and often a reaction to price that already happened. Focus is concrete. You can count it. And a shift in focus tends to come with, or even slightly ahead of, the moments that matter, because something has to draw eyes before anything else can happen.
Why sentiment is the weaker signal
Sentiment polarity is genuinely hard to measure and weak even when you get it right.
It is noisy. Social posts are sarcastic, ironic, meme-coded. "Going to zero, buying calls" can be a joke, a hedge, or real conviction, and a model rarely has the context to tell. Emojis flip meaning between communities. Coordinated hype and bots distort the mood.
It is backward-looking. Sentiment often turns positive because the price went up, not before. A lot of impressive sentiment models are quietly measuring the past with extra steps.
And the research bears this out. We covered the evidence in detail in does social media sentiment predict stock prices, and the recurring finding is that sentiment polarity is a weak, unstable predictor, while volume and search trends carry more signal. The mood is the noisy part. The attention underneath it is the useful part.
Why attention is cleaner
Attention has three properties that make it a better signal than sentiment.
It is measurable. You can count mentions. There is no need to guess whether a post is happy or sad, you just register that the post exists and is about a given asset.
It is harder to fake at scale. A single account can post a thousand bullish messages, but genuine, broad attention across many independent communities is much harder to manufacture than a sentiment score.
It is less ambiguous. "This asset is being discussed 5x more than yesterday" is a fact. "The crowd is 73% bullish" is an interpretation stacked on top of noisy inputs.
The three dimensions that actually matter
Attention is not just one number. Done properly, it has three dimensions, and reading them together is where the signal lives.
Volume. How much is being said. The raw level of discussion. Useful, but on its own it favours the assets that are always loud.
Velocity. How fast that volume is changing. A stock going from quiet to busy is more interesting than one that is permanently busy. Velocity catches the shift, not just the baseline.
Breadth. How widely the attention is spread across different, independent communities. This is the one most tools ignore, and it might be the most important.
Why breadth beats raw volume
Here is the distinction that separates real attention from an echo chamber.
A ticker getting hammered with mentions inside one corner of one forum is one thing. The same ticker suddenly appearing across many separate communities at once is something else entirely. The first is a few loud voices. The second is attention genuinely broadening, more independent groups arriving at the same focus around the same time.
Raw volume can be a single viral post or a coordinated push. Breadth is much harder to manufacture, because it requires many unconnected places to light up together. When you weight attention by how broadly it spreads, you filter out a lot of the noise that fools volume-only tools.
This combination, volume and velocity and breadth read together, is what we call Social Heat. It is not a sentiment score and it is not a prediction. It is a clean measure of where attention is concentrating and how that is moving.
What attention does, and does not, tell you
Be clear about the limits, because this is where honest and dishonest tools part ways.
A surge in attention does not mean an asset will go up. It means something is happening, more eyes are on it, the conversation has shifted. That is real and useful information. It is awareness, not a trade signal.
The right framing is information, not instruction. Attention tells you the market's focus is moving toward something before it necessarily shows up in price or mainstream coverage. What you do with that is your call, alongside your own research and everything else you weigh. Anything that turns a mention spike into a green "buy" light is overselling, which is the trap we wrote about in why most AI stock-prediction tools sell you noise.
Bottom line
Sentiment asks how the crowd feels, which is noisy, ambiguous, and often a reaction to what already happened. Attention asks where the crowd is looking and whether that is shifting, which is measurable, harder to fake, and genuinely informative. How much, how fast, and how broadly beats how positive.
That is the principle Orpail is built on. We measure attention across stocks and crypto, volume, velocity, and breadth combined into Social Heat, presented as an honest lens on where the market's focus is going, not as a forecast. If that is the signal you want, you can get early access here.
Orpail provides informational and educational data about publicly available social and news activity. It is not investment advice, not a recommendation to buy, sell, or hold any security or digital asset, and not a prediction of price or performance. Social attention is one lens among many. Always do your own research.