Will an up/down arrow on the market probability be the default way to bet from the homepage?
16
Ṁ105Ṁ444resolved Jun 25
Resolved
YES1H
6H
1D
1W
1M
ALL
Currently the homepage has up and down arrows on the market probability as the way to bet on a market, and defaults to a M$10 bet.
It's possible that this increases engagement by reducing friction. It also possible that this leads to confusion since interactions feel less like a betting, and more like a voting. Will Manifold change the current implementation?
This question is managed and resolved by Manifold.
Market context
Get 
1,000 to start trading!
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ44 | |
| 2 | Ṁ20 | |
| 3 | Ṁ11 | |
| 4 | Ṁ10 | |
| 5 | Ṁ9 |
People are also trading
Sort by:
@AkhilWable Have you seen the new preview on hover? You can see how your bet will change the prob now when you hover over one of the arrows
@ian Not on mobile unfortunately. In general not a big fan of on-hover states. I have a few different nitpicks about how it currently works, but I still think this is generally the right direction with some changes/iterations I'm sure you guys will try.
UI which best captures both intention and outcome of what users want to do is going to be an interesting problem. This other question I'd asked also shows all the different ways in which people think about what it means to "bet": https://manifold.markets/AkhilWable/how-do-you-decide-how-much-to-bet-o
I doubt most users start with the intention of wanting to change the probability from say 49% -> 54%, yet the current implementation makes that the visual focus and message: "Click this if you want to change the probablity from X to Y"; and that feels a little odd since Y feels like some arbitrary fixed number I have no particular attachment to.
That said, it's a significant and interesting change from how it used to work, and it's quite possible that I'll warm to it the more I use it.
@AkhilWable My guess is that most users have a sense of "this probability is too low/high" and want to do something about it. Thus, this design.
I agree that it's rarely exactly "54%" aka that M$10 is rarely the right amount, but I think having a small but notable increment that you can then spam-press to get to the correct final value, is actually a very quick/intuitive system - because you get the feedback on it super quickly.
My analogy is that on a microwave, the ONLY button that needs to/should exist is "+30sec". Not because you warm most food for exactly 30s, but because it reduces having to think "okay, so I want 2:30, let me find 2 => 3 => 0 => Enter". Even though 2:30 takes 5x 30s clicks, it has the advantage of being in a single location, so you can fairly mindlessly press up until the display shows you the right result. Super fast feedback.
As I'm writing this, I wonder if the right thing to do is to have the up/down arrows, but also add a "confirm" button somewhere once you've moved it to the right probability - eg by pressing on the number itself? This is the compromise I'm most in favor of, but it may also just be unnecessary...
There's also some chance this changes once markets get more deep with liquidity, and M$10 starts failing to change the market at all. I'll actually be kind of sad - to me, it's cool that you see your bet change "the state of the world", and high-liquidity markets are almost a bug (rather than a feature).
But maybe we could fix this by eg using the kelly criteria to help you decide what the correct bet size ought to be, increasing your bet defaults (and showing you more liquid markets) as your net wealth increases. My original proposal was to bet min(M$10, 5% delta), but now I'm wondering if it should be min(1% of portfolio, 5% delta)...
@Austin I like the microwave analogy, and from that perspective the up/down click is not a fixed 30s; rather an arbitrary increment between 1s and 30s. Also the increment is based on a concept like liquidity; a concept that isn't easy to grasp or factor into each bet. That is what makes it odd.
Further, I'm not sure most people *want* to change the market probability as their primary intention; most see markets as YES/NO polls. And then want to bet some amount *against* the market to make some manna.
I think not being able to budge highly liquid markets feels bad because that is the number the UI focuses on as a visual dopamine hit, and makes it the target of your interaction. Facebook spent a ton of time on the animation and sound effect of the like button for that very reason :)
Also, It would be a shame to hide highly liquid markets from regular users - those markets would likely represent the most interesting questions.
All that is to say that it's not an easy problem to capture all the ways in which people think about betting, calibration and profits. The current implementation comes close by making it a 1-click YES/NO poll. The fact that it changes the underlying market prob is almost redundant information, and maybe even a little confusing - most peoples experience with markets doesn't involve the odds/price changing when they make a bet.
And to finish some rambling thoughts - I'm guessing the initial set of Manifold users are people who have thought deeply about "probability" and what it might mean. So early Manifold users are more comfortable being probablity centric in their thinking and betting.
Though my guess is most users make essentially YES/NO predictions - and market probability is what emerges from that. Those jumping ahead to thinking about emergent consequences (i.e. in terms of adjusting market probabilities) would represent a minority.