Minor bug fixes don't count, the change must affect incentives/betting strategy to resolve this market. Possibly resolves to PROB if multiple answers are deployed in one change.
Related questions
Loans will now be on-demand, you'll have to click a button rather than getting a loan automatically each day. My reading is that this is not relevant to this market, since:
the change must affect incentives/betting strategy to resolve this market
and "always click the button" is the obvious strategy that leaves things the same as they were when the loans were automatic.
I'm guessing the fact that it incentivises you to open the app every day even if you weren't otherwise going to use it is a change in incentives for using manifold as a whole, but I am guessing this is not the kind of change in incentive the creator is including in this market.
Edit: I don't think this change is actually implemented yet.
@chrisjbillington Fair enough, thanks for weighing in, I was wondering about this while reading those updates. I think there is still a chance they sneak in an additional change in the near future.
Technically, long term loan recovery under the new scheme requires committing to checking the site every day for, say, 2 months, while previously you could bet your entire stack on blue and then come back once a month to re-invest the loans. So there might be a slight disincentive to long term plays under the new scheme. I agree it isn't consequential enough for this market.
@chrisjbillington Hm, maybe it does somewhat. Like, it means that infrequent Manifold users have less loans, so will not get as leveraged.
It's being done mostly for performance reasons I believe:
But is also educational: having to click a button means you're more likely to realise it's a loan that you have to pay back and not free money:
It's also a "fun daily reward" - that's about incentives for using the site. But doesn't change what optimal betting strategy is (always click the button, use leverage the same way as was optimal prior to this change).
Perhaps @brubsby you can share your thoughts on whether this change is relevant.
I think the total lack of people buying Other Yes speaks volumes.....this has been discussed and Obviously Happening for at least a week by now
@chrisjbillington The #suggestions post for it got upthumbs from James, stevo, and ian over a week ago, and then it came up in Standup this past Monday with additional support from David and Sinclair -- I've had it at 95% likely to happen soonish in my head for a while at least.
@SimonGrayson Wait, I’m an idiot.
I’ve just realised you mean the way we get Mana back from the markets we’re invested in…
@SimonGrayson "scheme" could maybe be "structure" here basically you get 4% of your invested mana back every day yeah. just that and the exact algorithm used.
Can you help clarify some stuff about how you plan to resolve if two options happen on the same day. Like:
Resolve based on which one was added to GitHub first
Resolve based on which one's effects are visible first to users of the site
Split resolution evenly between all options that happen in the same [commit, day, whatever]
I think I'll say here if two options happen between loan payouts it counts as happening at the same time, because I don't really care about the status of the code on the github, just the reality of the incentives. I realize this could technically be in conflict if the change was like, purely in the payback mechanism, but multiple changes happening between the 3am (or whenever) loan payments seems appropriate to be labeled as one change.
this would mean that if a change is made and then rolled back before the next loan payout, it doesn't close this market (even if it's not a change to the payout mechanism). also, if a NO LOAN DAY happens accidentally due to a bug again, then this market doesn't resolve, because that's just transient noise, when this market is about informing future strategy.
@brubsby Thanks for the follow-up, those were the exact two next points I was going to ask about. So:
1. Has to be an intentional change
2. Has to be left in place long enough for the effect to be seen
3. If more than one option happens in a very short time window, they'll probably get equal resolution weight. Even if one of the changes is visible in some fashion before the other one ("not exactly simultaneous")
pretty much yeah.
but for 1. if someone managed to accidentally (and significantly) change the loan percentage or calculation when trying to do a bug fix, but then decided to leave the new calculation in place either through permanent inaction or post hoc rationalization, I'd still count it as resolving this question. so I don't think intention is strictly required. I just wanted to protect against someone going "there was a no loan day (due to a bug), this could be construed as a change", when it is something I do not fundamentally care about here. this market is mostly for predicting the future of the loan system's mechanics.
ideally this would be a free response, but i didn't want to have to close it early and/or deal with people trying to specificity snipe exact changes after they're made. I'm not really sure what else they could do that would actually trigger "Other" here, so feel free to suggest things and I can add them or direct you to the bucket to bet in.