Current market prices imply that Trump is more likely to win the Nobel Peace Prize if there is not peace in Ukraine:

@EvanDaniel Manifold is unsurprisingly bad at calibrating very low probabilities.

@ProjectVictory Oh, I'm extremely aware, and have a long series of markets and comments about the problem :)

Here, I was curious whether a market trading not especially near 0% or 100% might exhibit the same problem, and whether posting math and tools about it would help. I was also curious whether having an arb bot occasionally trading would help. I assume it did a tiny bit based on trade volume and such, but I'm guessing it mattered more that people had time to look at math and place relevant limit orders. I'm guessing the big holders on this market know what they're doing. I saw several limit orders that looked like they might be arb attempts, given they were outside the arb limits by a fair bit.

I guess I feel like neither of these things will happen.

Spreadsheet is now updated, you can view some calculations here:
https://docs.google.com/spreadsheets/d/1_AnZwqGO8MsWs4ycTZEim41uuaiiWnVmHrMmPjdrzEY/edit?usp=sharing

This shows the correct price for this market assuming no correlation, alongside the min and max possible values from the underlying prices. (Those min/max values are what my bot is trading.)

I also added calculations for the implied probabilities for A AND B, A|B, etc.

Note that it doesn't refresh on its own. It refreshed when one of the cell values changes. You may want to make your own copy of the sheet, change a value, and change it back if you need to get it to refresh.

@creator This market can never be accurate or well-calibrated because most people would bet on "no ceasefire and no prize" but people are scared to bet on "ceasefire and prize"

@realDonaldTrump in that case, if the market is predictably and reliably mispriced, I suggest taking their money :)

And trying to find more ways to create popular and predictably mispriced markets.

it's interesting to think through the dynamics of this market — to me it seems more helpful to only ask "if there is a ceasefire, will he win the prize?" @EvanDaniel why did it seem helpful to also add "if there isn't and he doesn't" — what are you trying to gauge here?

@paw either version, along with "a and b" or "a or b", technically fills in the remaining degree of freedom in the existing markets. You can compute whichever implied probability you want.

I think markets that resolve n/a are, in general, bad design when there are alternatives; they remove the ability to make profit from predicting movements instead of results, and are generally difficult to hedge, so I don't like that answer. You would probably also want two conditional markets if going that route.

me: futarchy doesn’t have that fundamental flaw

also me: i hope nobel prize doesn’t give in to xor blackmail lol

So how does it resolve if there is a ceasefire, but he doesn't win; or if there is no ceasefire, but he does win? N/A?

@Lilemont in those cases the biconditional is false so I suppose No.

@EvanDaniel am I right?

@MachiNi correct, this market resolves yes if the underlying markets resolve the same way, no if they resolve differently.

This will only resolve n/a if one of the underlyings does.

0.07 × 0.18 + 0.93 × 0.82 = 78%

And that's assuming no correlation. With positive correlation, I think the number is even higher

@Fion I'd like to hear the reasoning of people betting this down. I arrived at 84.2% but I'm sure this kind of calculation misses some key aspects.
The only way for this to be lower is if the correlation is negative no?

@patrik or the other markets are mispriced. I didn't try very hard to arbitrage them.

My guess is that the question feels a bit like "if there is a ceasefire, will Trump win the prize" and people are thinking "he's not likely to get the prize even if there's a ceasefire, so bet NO". Either that or people read the probability of the market as if it's a correlation and think "these two events might be a little bit connected, but not 70% connected". Maybe the intuition is that if they were totally uncorrelated the percentage should be 0% or 50% or something.

I have to admit, the probability "feels" too high to me. I had to double check the calculation to convince myself.

@Fion I suspect you're right, the numbers do feel weird and the reasonable trading range is tight. I could have written the "and" version instead, but that has and even narrower and less centered trading range. Either one fills in the missing degree of freedom, both have problems.

I still don't feel like I know how to write good markets in this genre.

@Fion I added it to my spreadsheet. Also includes the min and max range that are consistent with the underlyings (which is what my bot is trading on).

https://docs.google.com/spreadsheets/d/1_AnZwqGO8MsWs4ycTZEim41uuaiiWnVmHrMmPjdrzEY/edit?usp=sharing

@EvanDaniel yo how can I get the getMarketProbability function?

@diracdeltafunk from the menu: extensions -> apps script, should be visible there. Here's the code:

function getMarketProbability(marketSlug) {
  var url = "https://manifold.markets/api/v0/slug/" + marketSlug;
  //var url = "https://manifold.markets/api/v0/slug/will-the-lk99-room-temp-ambient-pre";
  var response = UrlFetchApp.fetch(url);
  var json = JSON.parse(response.getContentText());
  var probability = json.probability;

  return probability;
}

@EvanDaniel wonderful, thank you!

@diracdeltafunk That will only work on binary markets, you'll have to add some extensions to get multiple choice working.