Steam Moves and Reverse Line Movement

What steam actually is

Steam is a coordinated price movement across multiple major sportsbooks within a short window, usually triggered by a substantial bet at a sharp originating book. The line at the sharp book moves first, then within seconds or minutes the same line shifts in the same direction at every retail book that uses that source as a reference.

Steam isn't a public bettor following the crowd. It's the market reacting to a sharp signal. When Pinnacle takes a $100K bet on the under at a major NFL total, their line moves. Within 30 seconds, DraftKings, FanDuel, BetMGM, and most of the rest of the US retail market move the same way, because they're using Pinnacle's line as their input and they don't want to be left holding a stale price after the sharps have already hit it.

The practical implication: steam is a moment of information being incorporated into the market. If you can place a bet before the steam fully propagates, you're getting better odds than the new consensus. If you can only place after the move is done, you're betting at the post-steam price, which is usually no longer +EV against the new fair line.

The mechanics of a steam move

Step by step, here's what a textbook NFL steam move looks like:

The window for a retail bettor to capture meaningful edge from a steam move is between T+1:00 and T+3:00, when major US books are still in the process of moving but haven't fully caught up. Sharp bettors with bots and direct API access exploit the T+0:30 to T+1:00 window. Retail bettors clicking through their phones can sometimes catch the T+1:00 to T+3:00 window if they're paying attention.

How to identify steam in real time

You don't need a paid scanner to recognize steam, though one helps. The mechanics:

• Sharp book moves first. Pinnacle, Bookmaker, Circa, BetCris. If the move is real steam, you'll see it at one of these before anywhere else.
• The move is at least half a point on a spread or 10 cents on a moneyline. Smaller moves are noise. Real steam clears the noise threshold.
• Multiple books move within a short window. If only one book moves, that's local action, not steam. If three or four major books all move within a few minutes in the same direction, that's steam.
• The move sticks. If a line moves and then drifts back to where it started within an hour, that wasn't steam, it was a temporary imbalance that the market corrected. Real steam holds.

Mathematically, you can think of steam as a step-function change in the market consensus probability. The implied probability of an outcome jumps by a meaningful amount within a short window, and that change persists.

Following steam vs being steam

Two different ways to think about steam strategically:

Following steam. You see steam, you place the bet on the sharp side at whatever retail book is slowest to move, capturing the gap between the new fair line and the still-stale retail price. This is technically profitable in expectation but fragile in practice. Books recognize steam-following customers quickly and limit them. The CLV is excellent (you're betting at pre-move prices), but the runway is short.

Being steam. You're the source of the move yourself. Either you're a sharp originator (effectively impossible at retail volume) or you're betting model output that happens to align with sharp action (which means your model is producing accurate prices). This is where line shopping plus a personal model converges into something that looks like sharp action even at small stakes.

For most retail bettors, the realistic strategy isn't following steam in real time. It's recognizing that the closing line is the result of accumulated steam, and that betting against the closing line consistently is what produces sharp results. If you're consistently on the right side of the eventual close, you're effectively riding all the steam moves that happened along the way.

What reverse line movement is

Reverse line movement (RLM) is the situation where the line moves against the side that's getting the larger share of bets. It's one of the most cited signals in sharp betting analysis.

Standard scenario. A team is at -3 (-110) on Sunday morning. By kickoff, 75% of bets are on that team. Public sentiment is overwhelmingly on the favorite. But the line has moved to -2.5 (-110). The line moved away from the side that's getting the bets. That's RLM.

The typical interpretation: the small minority of bettors on the underdog must be betting much larger amounts on average, because the book moved the line toward them despite the larger ticket count on the favorite. Large bets that move the line against the public ticket count are presumed to be sharp action. RLM is therefore a sharp signal.

The math of RLM

RLM hinges on the difference between bet percentage (tickets) and money percentage (handle). Most public tracking sites show both numbers.

Numerical example. Sunday NFL game. Favorite is -3 at open.

Why this works as a signal: sharp bettors place larger bets per ticket than recreational ones. When the dollar share diverges meaningfully from the ticket share, the dollar-heavy side is presumed to be the sharp side. When the line confirms this by moving in the direction of the dollar-heavy side, the inference is strengthened.

RLM caveats

RLM gets cited a lot, often incorrectly. A few things to keep in mind:

Public bet/money data is not always trustworthy. The numbers come from books that voluntarily share data, or from companies that scrape it imperfectly. Different sources can show different splits for the same game. The "75% of bets on the favorite" stat from one source can be 60% from another.

Line moves can have non-RLM explanations. An injury report drops, weather forecasts change, lineup news breaks. The line moves for an information reason, not a money-flow reason. The bet/money split at the moment of the move tells you nothing about why it moved.

Books shade lines toward public sentiment. A line moving away from the public side might just be the book correcting a deliberate shade after watching how the market priced it. That's not really RLM in the sharp sense, it's the book recalibrating.

RLM is not a strategy, it's a signal. Knowing that sharp money is on the underdog tells you something. It doesn't tell you the underdog will win. Betting RLM blindly is a recipe for chasing other people's information without understanding the underlying mispricing.

False steam

Not all steam is sharp. A few classes of moves look like steam but aren't:

Operator-side moves. A retail book gets a large unbalanced position on one side and adjusts the line to attract action on the other. Other retail books copy because they use the same data feed. The market moves but the underlying probability hasn't changed. The line eventually drifts back as the imbalance resolves.

Public squeeze. Heavy public action on a popular team causes books to move the line to limit their exposure. Looks like steam if you don't see the bet/money split, but it's the opposite of sharp, it's the public pushing the line.

Coordinated retail moves. When all major US books simultaneously update their data feed (often on the hour or half hour), prices can shift simultaneously without any sharp action driving it. This is mostly a retail-only phenomenon and the move usually doesn't show at sharp originator books.

Distinguishing real steam from these alternatives requires watching the sharp originator books. If Pinnacle didn't move, the move at retail isn't sharp steam, it's something else.

Practical takeaways

• Steam is real and tracking it does provide useful information about where sharp action is landing.
• Following steam in real time is a short-runway strategy. Books limit accounts that bet immediately after sharp moves.
• The closing line incorporates all the steam that happened during the betting cycle. Betting consistently on the right side of the closing line is the long-form version of "follow steam."
• RLM is a useful signal but it's frequently misread. Make sure you have reliable bet/money data and that you understand why the line moved.
• Watch sharp books (Pinnacle, Circa, Bookmaker) for the original move. Retail-only moves aren't steam in the meaningful sense.
• Steam-following only beats expectation if you can actually capture the price gap before retail books fully adjust. For most retail bettors with phone-based slow execution, that gap closes before they get a bet down.
• The realistic value of understanding steam is interpretive. It tells you what the sharp market thinks, which informs your overall view of which side is value, even if you're not literally betting at the moment of the move.