(2-VT) Proposal: Change Deal Price for Human-AI Trades

Status
Not open for further replies.
It's a raw nerf to trade deals if you have absolutely no intention of staying friendly. That's why I said it favor peaceful play (you trade some gpt for diplomatic boost) over aggressive play (you simply lose out gpt). Players have zero choice in the matter.
If you want to fix the huge gold influx from trade deal, fix the calculation, not player's choice. As long as we still have ridiculous deals, even when turning half of them into diplomatic boost it would eventually become a big swinging contribution.
At the day this isn’t really any different than just making the AI stingier with their deals. There has never been “choice” with AI deal valuation. Either they value deals in a reasonable way, or they don’t. If people think the deals are too high, then the AI should be made stingier…which is the meat of this proposal.
 
At the day this isn’t really any different than just making the AI stingier with their deals. There has never been “choice” with AI deal valuation. Either they value deals in a reasonable way, or they don’t. If people think the deals are too high, then the AI should be made stingier…which is the meat of this proposal.
The thing is the AI would be more stingier only if you're gonna be aggressive later on (which, at that point of the deal, the AIs have no idea yet). If you're playing peacefully and can make use of the diplomatic boost then it simply exchange gpt for friendship, not a raw nerf.
Dupes of luxury are literally worthless. No amount of calculation change is going to change that, unless you want AI to give away dupes for free.
I just made a suggestion a few posts before this. Applying a soft cap/diminishing value into the calculation would be enough.
We only care about the top cap because it causes inflation, nobody cares about bottom value.
 
we want a raw nerf.
Then do it to both peaceful play and aggressive play all together, and not give one an advantage over the other. Change the calculation, not how it's applied to different playstyles.
 
@azum4roll I do agree that it seems very easy to max out the capped +60 trade opinion modifier with this model.

Why isn't the averaged amount the baseline +0 opinion, and the player lowering their sale price or the buyer increasing their buy price the source of opinion? Does the AI not do this with each other?
 
Both sides actually gain something (gold, or saving gold) from a deal at the average price, so this should give the opinion.
 
I'd second making the even point +0. It's true that both sides are getting a better deal from their point of view, but you can just as easily say they're getting a bad deal from the trade partner's point of view. ("They're charging me more than what they think this is worth.")

If you really wanted to get fancy you could try and apply some kind of AI trait to the diplomatic outcome. Is this list accurate for VP?
These two seem like they could work.
  • Diplobalance indicates how likely the leader is to propose and accept fair trades. It may also indicate how hard the leader will work to stop other players from gaining insurmountable leads.
  • Gold Emphasis indicates how strongly the leader focuses on increasing the :c5gold: Gold output of his or her cities.
 
I'd second making the even point +0. It's true that both sides are getting a better deal from their point of view, but you can just as easily say they're getting a bad deal from the trade partner's point of view. ("They're charging me more than what they think this is worth.")

If you really wanted to get fancy you could try and apply some kind of AI trait to the diplomatic outcome. Is this list accurate for VP?
These two seem like they could work.
Trade at the event point is not neutral, though. It's still better for the AI to take the deal than not too. The more trade deals you have with a civilization, the worse a war would be that would block such deals, hence positive modifier.
 
The true cost would have to work like an auction - it's the opportunity cost of selling to the highest offer, i.e. the second highest offer. But we don't know what buy price everyone would offer, without actually making it an auction.
 
Not sure why we're trying to dig more into this but from my perspective, if the goal is to eliminate the ridiculous trade deal we're having, just slap a scaling soft cap onto it.
Nothing wrong with the current AI offering system, since if we nail down the trade deal issue, the difference between best price and worst price would be small enough to ignore along with the small diplomatic boost, and players keep the options to be nice or to be greedy.
 
Not sure why we're trying to dig more into this but from my perspective, if the goal is to eliminate the ridiculous trade deal we're having, just slap a scaling soft cap onto it.

That actually exists now, there is an era scaling cap on how much the AI can offer. It is possible to just lower that number further
 
I'll reiterate. The AI treats humans differently than it treats other AIs when it considers trade deals. This must change. It is the primary reason for this proposal.
 
And the primary reason why we have this system in the first place is because of AI programing limitation.
The AI can only say yes or no to a trade, thus you need a compromise for AI to AI trade to make them not permanently not accepting trade deal because of a few gpt difference (and they got compensated with diplomatic boost). It's not purposefully done to create advantage to human (more like giving AI more of an edge because it's easy for them to get more gpt, but diplomatic boost is hard to go by since they don't know how to manipulate global politic).
Don't fix what's not broken just because it's not "equal". Fair doesn't always mean equal.
 
AI is already throwing out gifts without any effort on my part -- I thought this was something they did when you were on good terms. So as-is, Human players are benefiting on all sides: price gouging for AI, best deals from AI, all while maintaining positive opinion. To make accepting a lower price actually meaningful, diplomacy would need to be made harder. But then you're tweaking not only trade deals, but also diplomacy AI. All to avoid removing "choice" from the player, which really isn't a choice at all, it's just a bunch of upsides that you can opt out of.

At the very least you'd need the AI to hate you for forcing them to pay full price for an item, and give some way for the player to know where the middle point is so the AI can see the deal as fair. The only way diplomacy works right now is that the AI only takes fair deals; it doesn't use trade opinion as a resource. So maybe a more robust solution would be in that direction? Have some kind of diplomacy currency that can help offset bad deals one way or another. You gain it be giving good deals, and lose it by giving bad deals. That actually sounds really cool, but way out of scope for this change.
 
That actually exists now, there is an era scaling cap on how much the AI can offer. It is possible to just lower that number further
That shouldn't exist and the valuation formula should change to account for the new WLTKD changes (which haven't even been implemented), but that's a whole other proposal.

The problem is that the luxury really is worth that much, and it's a gain to pay that much GPT for it. Meanwhile there's no difference for the seller to have that duped luxury or not.
 
At the very least you'd need the AI to hate you for forcing them to pay full price for an item, and give some way for the player to know where the middle point is so the AI can see the deal as fair. The only way diplomacy works right now is that the AI only takes fair deals; it doesn't use trade opinion as a resource. So maybe a more robust solution would be in that direction? Have some kind of diplomacy currency that can help offset bad deals one way or another. You gain it be giving good deals, and lose it by giving bad deals. That actually sounds really cool, but way out of scope for this change.
To make the AI hate you but still accepting a bad deal sounds like an ok idea (rather than the old binary yes/no demand option) that can go to another proposal regarding AI interaction.
As far as I can see, the main goal of this proposal is to nerf down the amount of gpt players can get from trade deal, and the proposed method doesn't work (and also have nothing to do with above AI interaction, as overvalued trade deal would still be the same, just got split into gpt and diplomatic boost and become an easy way to play peaceful as it's very easy to cap out trading bonuses this way)
 
Proposal vetoed.

Reason:
Discretionary veto because the proposal is based on a bugged mechanic. Can be re-proposed once fixed.
 
Status
Not open for further replies.
Top Bottom