Will my calibration improve if I only use bets of size 10 mana?
3
53
140
resolved Mar 29
Resolved
N/A

Current state:

I treated the whole website as a social platform and was not concentrating on calibration.

The rest of the spring i will limit myself to increase diversification:

  • Not allowed to use limit orders.

  • Can bet maximum 10 mana per market per day.

  • Can sell shares worth of maximum 10 mana per market per day.

  • I treat unlinked multimarkets as separate markets, but linked multimarkets as a singular.

The rule about selling only 10 mana also works on already standing bets

If in June I have higher score than -1.25, resolves YES. Otherwise resolves NO.

If I forget about this idea and buy/sell big values, then resolves NO.

Get Ṁ200 play money
Sort by:
opened a Ṁ100 YES at 60% order

The way Manifold calculates calibration takes bet size into account, so smaller bets will have a smaller impact on your calibration. I initially bet NO before realising that your calibration would still probably improve a bit. Still, if you want to maximise your calibration, I’d recommend not limiting yourself to such small bets.

If i wanted to maximise by any means, i would just follow the news and make bets between the information reveal and resolution. I actually don't like that about manifold.

Bigger bet, if wrong, is also a bigger damage for calibr.

I think your calibration may improve but I don’t think that’ll be due to betting only 10 M per market

But the degradation of my calibration from A to B+ was definitely a result of my biggest bets, which turned out wrong.

@KongoLandwalker Hum okay

I was thinking something like your calibration will improve but the main factor is that you started optimizing for calibration. Actually I didn’t think of it first but small bets can help because they don’t move the initial probability a lot. So small bets on well calibrated market are in some sense well calibrated