Steam Moves - Following Sharp Money vs Finding It Yourself

[SharpEddie47]

A steam move: a rapid, simultaneous line movement across multiple sportsbooks, usually triggered by significant action at one or more reference books.

In American sports betting: Pinnacle and Circa are widely treated as reference books. Most smaller operators don't have the risk management infrastructure to independently price every market, so they set their lines using these books as a baseline and adjust from there.

When a large bet hits Pinnacle and moves their number significantly: other books that reference Pinnacle's line follow within minutes, sometimes seconds.

A steam-chasing service: software that monitors line movement across dozens of books simultaneously, detects when multiple books move in the same direction within a short window, and alerts subscribers that a steam move is occurring.

The promise: someone with significant capital and presumably significant information just moved this market. Follow them.

The question this thread is built around: by the time you receive that alert and act on it, what's actually left.

 

[oli_sussex]

The exchange's relationship to steam is specific and worth establishing first.

The exchange price often is the reference point, not just for other exchanges but for fixed-odds books that use exchange data as one input into their own pricing.

A significant move on the exchange, driven by genuine money taking a position: this can itself be the origin of what a steam-detection service downstream identifies as "steam" at fixed-odds books.

When a book takes an unusually large bet relative to their normal limits on a specific market: the standard response is to "circle" the game, temporarily suspending it, and re-post a new number once they've reassessed.

A circled game reappearing at a different price: one of the clearest signals available that something significant just happened. It's also one of the slowest to reappear, because the suspension itself takes time.

 

[DublinDegen]

paid for a steam alert service for about four months during a bad period...

the pitch was specific: professional syndicates move millions, when they move the market follows, you get alerted within seconds and can get on before the soft books catch up...

what it actually felt like: notifications arriving on my phone, each one saying essentially "something is happening, act now"...

placed bets off these alerts constantly...

the specific appeal wasn't really about the edge...

it was that someone else, somewhere, with more money and presumably more information than me, had already decided...

i wasn't analyzing anything... i was reacting to a notification that told me someone smarter had already analyzed it...

the responsibility wasn't mine... the conviction wasn't mine...

i was borrowing both, in real time, for a subscription fee...

 

[FadeThePublic]

Conor's framing of "borrowed conviction" is precisely the psychological mechanism, and it connects directly to the fade-the-public methodology in a way worth examining.

Fading the public: I form my own view, independent of what the crowd is doing, often specifically because of what the crowd is doing.

Following steam: I form no view of my own, and instead adopt whatever view the detected movement implies someone else has.

These can point the same direction. Steam often moves against the side the public is heavily backing, because that's exactly the situation where a sharp position creates a meaningful price gap to exploit.

But they're not the same activity. One requires you to have a model of public behavior and a reason to disagree with it. The other requires nothing except a fast notification and a willingness to act on it.

The first is analysis. The second is, as Conor describes it, a form of following.

 

[ThePuntingProf]

Before any of this was automated there was a version of steam-following that required physically calling around shops to compare prices, this is what serious bettors did in the 1990s, you'd hear that a price had shortened significantly at one well-regarded shop and you'd ring around to see whether other shops had moved yet, the ones that hadn't moved were where you placed the bet, the entire process took twenty minutes or more by which point the gap had often already started closing the speed of the gap closing has compressed from twenty minutes to single-digit seconds over thirty years but the underlying structure is identical, someone with more information or more capital moved first, everyone else is racing to either match them or exploit the temporary gap before they do, what's changed is the timescale not the fundamental relationship and I'd add that in the telephone era you at least knew, roughly, which specific shop had moved and could form some view about whether that shop's customer base made the move meaningful, the algorithmic version aggregates dozens of books simultaneously and the individual signal, which specific book moved first and why, gets lost in the aggregate alert.

 

[Klaus]

Whether steam exists meaningfully in Bundesliga markets is worth examining separately from the American context Eddie describes.

The total betting volume on a Bundesliga match is a fraction of an NFL game. The number of books with sufficient liquidity to be meaningful "reference books" for German football is smaller, and the concentration of sharp capital is lower.

A significant line movement on a Bundesliga match: can occur, but is more likely to be driven by team news (a confirmed lineup, an injury) than by a single large position from a syndicate.

The distinction matters. A steam move driven by genuine new information, a player ruled out, is fundamentally different from a steam move driven purely by capital flow with no new information attached.

In lower-volume markets, line movement is more likely to correlate with actual news. In the highest-volume American markets, line movement can occur from capital flow alone, with the "information" being entirely about someone else's position rather than about the match itself.

 

[SharpEddie47]

Klaus's distinction maps onto something I should have separated from steam at the start: reverse line movement.

RLM: the betting percentage shows, say, seventy percent of bets on Team A, but the line moves toward Team B.

This indicates that while most individual bets are on Team A, the average size of bets on Team B is large enough to outweigh the volume, implying sharp money is on B.

RLM is slower and more aggregate than steam. It's observed over hours, sometimes the full week leading up to a game, by comparing bet percentages to line movement.

Steam is a sudden synchronized jump.

Both get described loosely as "sharp money" signals. They're different signals with different reliability profiles, and a lot of the steam-chasing discourse conflates them.

 

[CoachTony_Bets]

The coaching equivalent of following steam without understanding it is specific and common.

A successful program does something unusual, a specific defensive scheme, a particular fourth-down decision philosophy, and it gets copied across the sport because it worked for them.

What often gets missed: it worked for them because of personnel they had, a specific opponent they were facing, a context that made the decision correct in that situation.

The copying program implements the same scheme or decision without the personnel or context that made it correct, and it doesn't work the same way.

Following a steam move is the betting equivalent. The original position may have been correct because of information, personnel, or context specific to whoever placed it. The follower has none of that context, only the action.

 

[oli_sussex]

There's a manipulation layer above all of this worth naming directly.

A sophisticated participant places a bet specifically designed to be visible and to look like sharp action, at a book or in a size calculated to trigger steam-detection services and cause other books and retail followers to move.

Once the followers have moved the soft books in one direction: the original participant takes the opposite position at the now-improved price the follower-driven movement created.

This is sometimes called "smoke" rather than steam, a deliberately false signal designed to be followed.

Whether any specific steam alert is genuine sharp positioning or deliberately constructed smoke: not distinguishable from the alert itself. Both look identical to a detection service that's only measuring synchronized line movement.

 

[ParlayPrincess_88]

I keep seeing "steam" mentioned in betting content and never knew what it meant beyond "something is happening fast."

Reading this thread: it's "someone else, possibly sophisticated, possibly not, possibly genuine, possibly a deliberate trick, did something, and the line moved, and now you're being told to do something too, fast, before you can think about it."

Said that way it sounds like exactly the kind of decision-making this entire forum has spent its history warning against.

The packaging, "follow the sharps," makes it sound like the opposite of impulsive. The actual mechanism, react to a notification within seconds before the gap closes, is impulsive by design.

 

[DublinDegen]

princess saying it plainly is the thing i couldn't see while i was inside it...

"follow the sharps" sounded like discipline...

what it actually was: an external trigger, delivered as a notification, that i acted on immediately because the entire pitch was built around speed...

the four months i did this: looking back, i wasn't betting less impulsively because i had a system...

i was betting exactly as impulsively as ever, just with a system telling me when to be impulsive...

the timing of the impulse had been outsourced...

the impulse itself never went anywhere...

 

[TaffyTipster]

Rugby markets don't really have this at all, as far as I can tell.

The volume isn't there. A significant move on a Six Nations match might happen because of team news, an injury, a selection announcement, and it'll move gradually as people react to the news becoming public.

I've never seen anything that looked like the synchronized multi-book jump Eddie's describing for American sports.

If steam exists in rugby betting it's moving so slowly and on such small numbers that nobody's built an alert service for it, which might be the most honest indicator of whether a market is liquid enough for any of this to be a real phenomenon versus a real phenomenon.

 

[FadeThePublic]

Taffy's observation points at something underneath all of this.

"Sharp money" as a category exists because certain markets have enough capital flow that synchronized, capital-driven movement is detectable and meaningful.

In markets without that volume: line movement correlates much more directly with actual information, as Klaus said for Bundesliga, because there isn't enough undifferentiated capital to move a line for reasons unrelated to information.

Which raises the question of whether "sharp money" is a property of certain bettors, or a property of certain markets, specifically markets with enough volume that capital itself, regardless of what's behind it, can move a price.

In a low-volume market: the information moves the price. In a high-volume market: capital can move the price somewhat independent of information, and then "sharp money" becomes the label for whichever capital flow turned out, after the fact, to have been on the right side.

 

[Klaus]

Fade's framing connects directly to the courtsiding thread.

Courtsiding was about racing an outcome that had already happened.

Steam-following is about racing someone else's position that's already been taken.

Neither involves the follower's own analysis of the underlying event. Both involve trying to be fast enough to capture value from information, an outcome in one case, a position in the other, that already exists somewhere else.

The difference: courtsiding's information gap closed because it threatened market integrity itself, as we discussed. The steam-following gap hasn't closed in the same way, because capital flow chasing capital flow doesn't threaten anything, it's just an additional layer of activity on top of the market, generating volume and commission regardless of who's actually right.

 

[ThePuntingProf]

What I take from this thread, having lived through both eras, is that the telephone-era version at least had the virtue of slowness forcing a decision to be made.

Twenty minutes to ring around shops meant twenty minutes during which you could decide the move wasn't meaningful, or wasn't worth chasing, or that you'd rather form your own view.

The instant alert removes that twenty minutes entirely.

What used to be a piece of information you had time to evaluate has become a trigger you're expected to act on before evaluation is possible.

The technology made the information faster. It didn't make the human evaluating it faster. The gap between those two speeds is where Conor spent four months.

 

[SharpEddie47]

The honest answer to the thread's title.

Following sharp money: requires no analysis, provides borrowed conviction, and the value available by the time you act has already been partially or fully consumed by whoever you're following plus everyone faster than you.

Finding it yourself: requires the analytical work this entire forum has spent its history describing, produces a position you can actually evaluate and learn from regardless of outcome, and doesn't depend on a notification arriving in time.

The first feels like sophistication because of the language around it.

The second is sophistication, and it's slower, and it doesn't come with an app that buzzes.

 

[DublinDegen]

slower and doesn't come with an app that buzzes...

that's the whole appeal and the whole problem in one sentence...

cancelled the subscription eventually...

not because i'd worked out the economics of it...

just noticed i hadn't had an original thought about a single bet in four months and that scared me more than the losses had...