There are several ways in which our market choices are very restrictive.

1. Numeric markets are just binary markets with a mapping from [min, max] into [0%, 100%]. For normal numeric markets the mapping is (X - min) / (max - min). For log scale numeric markets, the mapping is scaled logarithmically to be more granular near 0% and 100%. There's no reason these need be the only two options; Manifold could let the market creator specify any mapping function rather than only letting us pick between two specific options. (I think this would primarily be useful for numeric markets on outcomes that have a weird distribution where you want more granularity somewhere that isn't either extreme of the scale.)

2. For some reason, numeric markets don't work properly with small ranges. Trying to create a market with a min of 1 and a max of 2 for fine-grained betting results in very low liquidity. I'm not sure why this is, since such a market should be completly isomorphic to a market with a larger range, just scaled down.

3. Given the equivalence between binary and numeric markets, Manifold could let us pick whether there's a percentage sign displayed or not. This could be used for markets on events that are known to have a near-0 or near-1 probability; you could create a 0%-100% market on a log scale. Allowing us to display other units, such as currency symbols would be nice. So would a "date" unit, where the underlying market is just a numeric market on unix time or whatever, but it displays as a human readible date.

4. Many binary markets have two roughly equivalent primary outcomes. For example, whether a democrat or a republican president wins the election. Right now, we have to arbitrarily pick one option to highlight in the market, such as "Will a democrat win the election?". This is inelegant and skews the market, since it could bias how traders think about the result, and an unexpected result (like an independent candidate winning) will favor one side or the other. This can be solved by using a multiple-choice market instead and creating only two choices. However, multiple-choice markets have a much worse market-making algorithm (you're not guaranteed your payouts when you place a bet) and worse UI (you can't see shares held, can't place limit orders, can't sell any number of shares, etc). This could be fixed by improving multiple-choice markets and/or giving us the option to label binary markets outcomes with labels other than "YES" and "NO".

5. Often there are colors associated with certain market outcomes, such as blue for democrats and red for republicans. But market graphs don't let us pick their colors; on binary markets we're stuck with green and white, and on multiple-choice markets we're stuck with a predetermined selection. Letting the market creators choose what colors represent which options would be a nice aesthetic improvement, and also make it easier for people to hold in their mind what option means what.

In short, there's a lot of overlap between different market types and use-cases, and a lot of things that it seems like we should able to do that we can't. This market resovles YES if we get a significant number of these options available to us.

For example, if two-option multiple-choice markets are given the same market mechanism and usability features as binary markets, and we're able to change the graph colors, and low-range numeric markets start working, that would be sufficient to resolve this to YES, even if we don't get the ability to use custom mapping functions. (Unless we lose other options that cancel out those improvements.)