Alternate Lines Versus Modified Payouts

A reader asked us why our board treats two offers on the same player differently when both show a number lower than the main line. It is a fair question, and the answer is that only one of them is a different number for the same product. The other is a different product wearing a number.

The distinction is worth being pedantic about, because the two things require different research. One asks where you think a distribution sits. The other asks whether a price is worth paying. Conflating them produces confident conclusions about the wrong question, which is the most expensive kind of confidence.

Two definitions, stated precisely

An alternate line is a different threshold for the same stat, offered alongside the main one. If a strikeout market is posted at 6.5, an alternate ladder might also offer 4.5, 5.5, 7.5, and 8.5. Each rung is the same product, the same stat, the same settlement, at a different number. On a book, each rung carries its own price, and the prices move in the obvious direction: an easier number costs more to buy, a harder number pays more.

A modified payout is a change to the payout structure attached to a line, usually on a pick style platform where the base product has a fixed schedule. Instead of quoting a price per rung, the platform moves the number and attaches a multiplier to the pick: a discounted offer at an easier number that returns less, or a boosted offer at a harder number that returns more. The mechanics and the arithmetic for judging these are covered in the modified payout article.

Stated that way the two sound almost identical, and structurally they are cousins: both are ways of selling probability at a price. The differences that matter are in the shape of the product and in what you are allowed to do with it.

• An alternate ladder typically sells both directions at every rung. A modified payout offer is typically sold over only.
• A ladder is priced per rung, so the price is visible next to each number. A modified payout attaches a multiplier to a pick inside a schedule, so the effective price depends on the schedule too.
• A ladder gives you a continuous picture of the market across numbers. A modified payout gives you one point with a price on it.
• A ladder rung and a boosted offer can land on the exact same number and still be different products with different settlement terms and different available sides.

Reading a ladder requires a distribution, not a projection

A single number cannot answer a ladder. If your projection for a hitter is 1.4 total bases, that tells you almost nothing about whether the 0.5 rung, the 1.5 rung, or the 2.5 rung is mispriced, because those three questions are about three different regions of the same distribution. A point estimate has no regions.

What a ladder actually asks is: what is the probability of clearing each of these thresholds? That is a question about shape. Two players with the same mean can have completely different answers at the 2.5 rung if one of them produces in bursts and the other produces steadily. The projections article covers why simulation based approaches produce a distribution rather than a single number, and this is the clearest case for why that matters: a ladder is a distribution being sold in slices.

This is also why a ladder is informative even when you do not intend to buy any rung. The prices across rungs describe the market's implied distribution: how much probability it assigns to each region. Where your simulated distribution and the market's implied one disagree in shape, rather than just in location, that disagreement is more interesting than a single number gap, and it is usually more diagnostic of a model error too.

What a ladder tells you about the market

Work down a ladder and you can recover roughly what the market thinks the distribution looks like. The probability of clearing each rung falls as the number rises, and the rate at which it falls describes the spread. A ladder whose implied probabilities drop steeply from rung to rung is a market saying the outcome is tightly concentrated. A ladder that drops gently is a market pricing in a long tail.

That gives you two distinct kinds of disagreement to look for. Location disagreement means your whole distribution sits above or below the market's; that is the ordinary case, and it is what a headline gap on the main line reflects. Shape disagreement means the two distributions are centered similarly but differ in spread, so you agree at the main line and disagree at the extremes. Shape disagreements tend to be where naive models fail, because a model that produces a good mean can still produce badly wrong tails, and the rungs far from the middle are exactly where that shows up.

Treat a shape disagreement as a warning before you treat it as an opportunity. If your model clears a far rung far more often than the market implies, the most likely explanation is that your simulated distribution is too wide, not that the market forgot about the tail. Calibration evidence across many graded outcomes is the only thing that settles which it is, and the honest default until then is that the market's shape is the better one.

The trap of the identical number

Here is the failure this article exists to prevent. You see a rung on a ladder at 5.5 and a boosted offer at 5.5, and you conclude they are the same bet in two places, so you take whichever pays more. They are not the same bet, and the comparison can be wrong in several independent ways at once.

• The stat definitions may differ. Two platforms can both call something total bases or fantasy points and settle them differently.
• The available sides may differ. One offer may sell both directions and the other only the over.
• The payout structures may differ in kind, not just in size, because a per rung price and a multiplier inside an entry schedule are not directly comparable.
• The void and settlement rules may differ, which changes what happens when a player does not appear or a game is shortened.
• The effective cost of a modified payout depends on the schedule it sits inside, so the same multiplier is not the same deal in every structure.

The platform comparison in the platform differences article covers how far these divergences go in practice, and the glossary pins down the vocabulary when a platform uses a familiar word for an unfamiliar thing. The working rule is simple: two numbers being equal is not evidence that two offers are equal. Verify the stat, the sides, the payout, and the settlement rules on the platform itself, because platform terms change and no article can be current about them.

Two products, two research questions

So the two products ask you for different work. For an alternate ladder, the question is where the distribution sits and how wide it is, and the answer requires a distribution you trust at the specific rung in question. Probability comparison does most of the work, because the price is quoted per rung and you can convert it directly into an implied probability to compare against your own.

For a modified payout, probability alone cannot answer the question at all. A discounted offer will nearly always show a higher probability than the standard line, and that fact carries no information, because it is what you paid for. Only expected value, probability multiplied by the actual payout under the platform's current rules, distinguishes a good discounted offer from a bad one. And when the offer sits inside a multiple leg entry, the schedule for that entry enters the calculation as well, which is the arithmetic in the entry structure article.

Slateline treats these as distinct product classes rather than as variations on one. Ladder rungs are evaluated as separate offers against the same simulated distribution, so a model can prefer one rung and pass on another for the same player without contradicting itself. Modified payout offers are graded on expected value per pick with the payout class included, and offers sold over only are never surfaced as unders. Keeping the classes separate is also what stops a board from reporting a book's own ladder as though it were a disagreement between two books.

Precision in vocabulary is not pedantry here; it is the thing that keeps the analysis attached to a product somebody actually sells. Before doing any work on an offer, name which of the two it is, then ask the question that class of product deserves. Set a budget before the slate, and if this stops being recreational, help is available at 1 800 GAMBLER and through the National Council on Problem Gambling. To see ladders and modified payout classes labeled side by side against our projections, Prop Grid is where that view lives.