Pool, fixed odds or exchange: the difference is where the cut sits

§1Where the cut sits

The three structures are often compared on price, which is the hardest thing to compare and the least structural. It is more informative to compare where each one takes its money, because that determines what can be shopped, what is fixed, and what your own stake does to your own outcome.

§2The fee on one winning bet, with the price held equal

Take an illustrative 100 staked at an illustrative 4.00, returning 400 if it wins, of which 300 is profit. Holding the underlying price identical across all three — which the three structures would not in reality do, so this compares the fee shapes and not the markets — the three cuts cost very different amounts.

The take is charged on the amount returned, not on the amount won, which is why it is relatively heavy on a winning ticket and invisible on a losing one. Commission on net winnings works the other way round: it is charged only where there is a profit, and it is charged on the profit rather than on the return. Both are legitimate ways to be paid for running a market. They are not equivalent, and their difference grows with the profitability of the bettor who pays them.

§3What a pool does structurally well

There are real virtues in the structure, and stating them is part of stating the whole mechanism honestly.

Nobody has to price anything. A pool has no price-setter, so it cannot be wrong in the way a book can be wrong. A book that misjudges a market and takes a large position against the event pays for it. A pool pays whatever its census produced, and the dividend can be a number no sensible pricer would have published.

It cannot refuse your bet for price reasons. A book limits or closes a market when a customer is winning, because the operator carries the risk. A pool does not carry that risk in the same way: your money is matched against other money in the pot, so the reason a book restricts a customer largely does not exist here, subject to the operator's own limits and rules.

It needs no order book. There is no liquidity to fill, no unmatched order to be left holding, and no reliance on another specific user completing their side of a trade. The pooling is statistical, and it happens at settlement.

Everybody gets the same cut. However uncomfortable that is, it is a kind of fairness that the other two structures do not offer: the same percentage is removed from the winning dividend of every ticket in the pool.

§4What a pool does structurally badly

The costs are the mirror image of those virtues, and they are worth listing with the same bluntness.

§5Where each structure is at its worst

None of the three is uniformly better, and the comparison is most honest at the level of conditions rather than of averages.

Read across the rows rather than down the columns: each structure has a row where it is the only sensible answer and a row where it behaves worst of the three. The pool's worst row is staking large relative to the money already on your selection, and its best row is the case in which no price-setter would be trusted and the crowd's own census is the only available estimate.

§6The honest comparison

Stripped of the vocabulary of each market — value, edge, expected value, efficiency — the three structures are three ways of paying for the same service. One charges a margin inside a price, one charges a commission on profit, one removes a percentage from a pot. Each of them takes money proportionally to how much a customer bets or wins, and none of them is a strategy.

What a reader can reasonably take from the comparison is a short list of structural facts rather than a preference.

• If the price being known matters to you, a pool is the wrong structure. That is the defining property of a pool, not a fault that can be worked around.
• If you want to compare offers, a pool offers nothing to compare. One pool, one take, one dividend, set by rules you read rather than prices you choose.
• If your stake is large relative to the pool, your own money is part of your own price. This has no equivalent in the other two structures and cannot be avoided by staking later.
• If you are going to win, the shape of the cut matters more than the size of the headline. A take charged on the amount returned is a heavier fee on winners than commission on profit.
• If you do not know the take, you do not know the price. Everything else on the dividend is decided by the crowd; the take is the part that is decided by the operator.