Ooo, the new League page is pretty.

@Quroe Yes, but it is visual overload compared to the previous.

@Quroe Looks messy and disconnected from the rest of the site.

If the entire site ends up looking like a cartoon I'm not sure what I will do. Use an alternative frontend?

@Eliza yeah it looks like the weird scammy PM clones that pop up. having a unified, consistent UI is very nice

i have a status quo heuristic bias but i think this new ui takes too much space and has unnecessary colors

@Bayesian Poll: /SG/which-manifold-leagues-aesthetic-do

@Quroe top 3 worst takes ever 🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀🥀

@JeromeHPowell I reserve the right to be wrong.

@Quroe very much so 🥀🥀🥀🥀 :)

It would be really cool if we could set dependent MC markets to resolve to more than one answer. Relevant for markets like this:

/Tumbles/who-will-be-top-4-in-melee-singles

@Tumbles but then it becomes impossible to bet them to the correct odds since betting up one bets down another, that's why you have to resolve to split percent

@TheAllMemeingEye I don't think it creates an issue there at all. I don't think the math would even need to change. If you bet one option up, the others need to go down by the same amount, exact same as it is now. If you want to keep multiple options right at 99% you could do it with limit orders.

@Tumbles but by definition dependent/linked multi-choice sum to 100%, surely setting 99% limit orders on multiple options will just cause them to cancel out the previous one, right?

@TheAllMemeingEye my suggestion is that they allow for dependent markets that sum to a different multiple of 100

This would be a very useful feature for my Bilt Rent Day markets where we know that only 3 out of 12 answers can resolve YES. If all 12 answers summed to 300%, then I think there would be less slippage (if I'm using that term correctly).

@Tumbles Yeah the math isn't hard here. You've got two share identities in the normal MC markets:

• a set of 1 YES in each outcome can be redeemed for M1

• a set of Y+N in a single outcome can be redeemed for M1

And everything else follows from that plus the per-outcome AMM. The auto-arb mechanism just applies those share identities in user-favorable ways as part of the share buy/sell process.

In the new version, with n outcomes that resolve with m yes:

• a set of Y+N in a single outcome is still worth M1

• a set of 1 YES in each outcome can be redeemed for m

And so instead of buying/selling all outcomes to return the total price sum to 1, the auto-arb mechanism instead returns it to m. Needs some code changes, but the math isn't hard.

It's much better than resolving equally, because it is far more capital efficient for anyone buying No or taking an equivalent position. Also far better UI/UX. It's also superior to independent MC for that application, because no one likes trying to run the arb calculations manually every time they want to trade.

I'd like a better comment structure where replies appear directly below the comment their replying to, rather than a high-level comment just being like a "bucket" and the comments below are sorted newest --> oldest

Wait, this market is ranked?

@EvanDaniel decimal limit orders would be lowkey goated on markets above a certain liquidity/volume threshold!

@bens Yeah having a threshold like that seems fine; it's not so helpful if we're talking about M2.7. Toggleable by market owner also seems reasonable.

expressing gratitude to the team for the pace of great new features recently :)

@brod 🙏

@brod seconded!!

@ian elaborate?

@realDonaldTrump He has comment notifications turned off so he never sees them. He wants to create a new version of notifications that bursts through (new setting) only showing comments the AI thinks is interesting.

(High confidence, not 100% sure.)

@Eliza i would just really love there to be comment notifications that were truthful when they said 'so and so mentioned you in their comment!' and they absolutely did not do such a thing