PrizePicks, Underdog, Betr, and Sleeper: What Changes Across Platforms

Open four fantasy apps on the same evening and look up the same guard's points. You may find four different numbers, three different payout structures, two different definitions of what happens if she leaves the game early, and one direction you are quietly not allowed to take. None of the apps is wrong. They are selling four different products that happen to share a player's name.

This article maps what structurally differs across pick style platforms such as PrizePicks, Underdog, Betr, and Sleeper, and why those differences matter for research. It deliberately avoids stating any platform's current multipliers, scoring charts, or settlement rules as fact, because those change, sometimes quietly, and the platform's own published rules are the only authority worth trusting. What follows are the mechanisms, which are far more stable than the parameters.

The entry structure is the product

A sportsbook prop is a single position with a price attached to each side. A pick style entry is different in kind: you combine several picks into one entry, and the entry pays on a fixed schedule that depends on how many picks you included and how many of them hit. Some formats require every pick to hit; others pay reduced amounts when most of them do.

This structure changes the math of everything downstream. Because picks are multiplied together, a small per pick disadvantage compounds across the entry, and a single weak leg drags the whole ticket. It also means the platforms are not pricing each pick individually the way a sportsbook prices each side; the margin lives in the relationship between the payout schedule and the joint probability of the combination. Two platforms can post the identical line on the identical stat and still be selling meaningfully different value, purely because their payout schedules differ.

Stat menus differ, and so do stat definitions

Each platform decides which stats it sells for which sports, and the menus only partly overlap. One app may post fight time and significant strikes for a UFC card while another posts nothing on combat sports. One may sell map scoped esports lines that another does not touch. If your research process lives on one app's menu, you never see the offers where another app's menu is weakest, which is often exactly where the interesting numbers live.

The subtler trap is that identically named stats are not always identical products. The clearest case is fantasy points. A fantasy points line is a bet on a weighted sum, and each platform chooses its own weights. If one app's scoring chart rewards a stolen base more than another's, the same real game produces different fantasy totals on the two apps. A projection built for one platform's chart is simply wrong for another's, even at the same posted line. Slateline models fantasy composites per platform for exactly this reason, and treats a chart we cannot verify as a reason to withhold settlement claims rather than guess.

The same player carries different lines at the same moment

Platforms set their numbers independently, update them on different schedules, and respond to different flows of entries. The result is routine, observable disagreement: a rebounder posted at 8.5 on one app and 9.5 on another at the same minute. Sometimes that gap reflects a real difference in the products, a different stat definition or payout structure. Often it simply reflects that one number has absorbed the day's information and the other has not yet, a dynamic covered in the line movement article.

For research, cross platform disagreement is itself a signal. When four operators agree within half a point and a fifth sits two points away, the outlier is either the best offer on the board or a product difference you have not noticed yet. Both possibilities reward investigation before action. This is why Slateline shows the same projection against every book's posted number side by side in Prop Grid: the spread across platforms is information the individual apps cannot show you, because each one only knows its own number.

Some offers only exist in one direction

On several platforms, certain offers are sold over only. This is common for alternate lines with modified payouts, and it appears on some standard lines too. The platform is not obligated to sell you the mirror of what it posts, and often it structurally does not: there is no under to take, at any price.

This matters more than it first appears. If your projection dislikes an over only offer, the correct output is a pass, not an under. Research tools and casual analysis alike can slip into treating every posted line as two sided, which quietly manufactures recommendations for products that do not exist. Slateline treats the sold sides of an offer as part of the offer's identity, and a projection that disagrees with an over only line is displayed as low probability on the over, never as an under recommendation. When you evaluate any tool, including ours, check that it makes this distinction.

Void rules differ, and voids change settlement

What happens when a player never appears, leaves injured in the first minute, or has her match end early varies by platform and by sport. One app may void the leg and recalculate the entry at a smaller size. Another may have different thresholds for what counts as participation. Combat sports and tennis add their own cases: withdrawals, retirements mid match, fights that end in seconds.

Void handling is not fine print; it changes the probability you are actually buying. A leg that voids on a did not play is a conditional product: you are betting on the player's stat given that she appears. A leg that grades as a loss in the same situation is a different, strictly harsher product. The same projection should price these differently, which is why Slateline's probabilities for sports with meaningful absence risk are built conditional on appearance, matching the void convention rather than pretending absences do not happen. The general question of comparing a projection to what a market believes is covered in the model, line, and market piece.

Alternate lines with modified payouts

Several platforms sell alternate versions of a line alongside the standard one: an easier number at a reduced payout, or a harder number at a raised payout. These offers make probability alone insufficient as an evaluation tool, because the payout has been adjusted precisely to compensate for the easier or harder threshold. A very likely over at a discounted multiplier can be a worse purchase than a coin flip at standard terms. That subject deserves its own article and gets one later in this library; the short version is that expected value per pick, probability times payout, is the only honest lens for them.

What this means for a research process

Pull the threads together and a few working rules fall out.

• Research the offer, not the stat. The platform, the line, the sold sides, and the settlement rules are all part of what you are buying.
• Never transfer a fantasy points conclusion across platforms. Different charts make it a different product.
• Treat cross platform line spread as a prompt: find out whether the outlier is stale, sharp, or structurally different.
• Confirm a side is actually sold before you fall in love with it. An under that does not exist is not a lean; it is a pass.
• Reread void rules whenever a platform updates its terms, because they silently reprice every conditional situation.

One more rule sits above the others: none of these platforms is the best one. Each structure trades something for something, and which trade favors you depends on the offer in front of you that day. Slateline is an independent research product with no affiliation with any platform named here, and nothing we publish is a recommendation of where to play. Our job is to show what the platforms are selling and what our projections think of it; deciding whether and where to participate stays entirely with you.

If you participate, keep it inside limits set in advance. Support is available at 1 800 GAMBLER and through the National Council on Problem Gambling, and our commitments are documented on the responsible gaming page. For the research itself, the Signal Board shows where our projections currently stand against the offers these platforms are actually posting.