Introduction
Currently anyone wishing to rent GU cards has a good solution in Lootrush. With the recent changes to the Daily Play & Earn rules, renting Gold and Diamond cards has become an attractive financial proposition, since the earning potential at ranks 10 and above with a shiny deck can be comfortably higher than the renting fees.
But what about lending? What is the situation for those owners who would like to get some return on their idle cards?
Since Immutable X layer 2 solution (L2) where the cards are minted and traded is an implementation of Starkware's StarkEx engine, it is not possible to run smart contracts in this network. This precludes trustless, temporary transfers of assets between accounts. There are no plans to move assets to a smart contract-enabled engine in the foreseeable future.
https://discord.com/channels/457152051239976961/457152051239976963/1238361870516097076
Any third-party implementation of a renting service thus requires the lender to surrender ownership of the assets to the third-party service, which becomes the effective owner of the asset from that moment on until the lender makes a recall request.
This is evidently not a trustless system; the lender / asset owner transfers the cards (in this case) to the renting service and assumes they'll honor the lending agreement and keep the assets safe. I find this compromise less than ideal. A lot of things can happen:
outright fraud (rug)
the company may go bankrupt and the assets frozen until the courts decide on a resolution
hacking by exploiting vulnerabilities in the system / server security
theft by a disgruntled employee with privileged access or any other circumstance
All in all, it seems like a poor solution in an ecosystem that has built its credentials on cutting-edge technology, trustlessness and ownership.
The 'internal' alternative
But what if the developers handled renting themselves? Here I make the case why this would be a better, safer solution, with upside to the game studio and perhaps to Immutable X as well.
I would like to propose a system where renting is handled internally via a traditional database and the cards never leave the owner's wallet, but become temporarily unavailable to play as long as they are being lent.
The game client is already able to filter the cards available to the player: even if you own all the cards, there are restrictions that remove cards from Formats; filters hide domain cards that do not belong to the God you selected to build a deck, etc.
Conversely, the game is able to make certain cards available to you, even if you don't own them - think about Sealed pools. Dev accounts and testing accounts for Superpower members have access to restricted cards. So there are already mechanisms in place to make cards enabled or disabled in a player's account.
With this solution, the GU backend would disable cards from the lender's account and enable it at the renter's, while also handling payments. No cards change wallets and the risks disappear. Renting becomes just a transaction on a standard backend database.
On the other hand, I don't know if such a system would bring any legal or taxation issues that would be difficult to overcome.
This is how I think the system could work:
The GU team provides the renting infrastructure including a webpage and in return keeps a percentage of the renting fee.
An algorithm could optionally take care of the renting fee automatically, based on a percentage of the current card price, plus or minus an adjustment based on demand: cards in high demand without supply would get algorithmically more expensive. This system would allow a 'set and forget' setup for the lender.
An example sometimes mentioned in Discord is Spider Tanks. In this game, the player parks the tanks in the UI (a garage) and they become available for others to use. Any gains with the shared tank are split between the owner and the renter. This system could serve as an alternative to paying an upfront rental fee, with the automatic splitting of any earnings from rented cards made during DP&E calculations, while GU still collects a fee.
I would prefer a selection approach where all cards are initially whitelisted for renting by clicking a button. Then the owner would pick either individual cards to exclude or entire decks (the decks have information about card quality). The game would keep unlisted just enough copies to play while making any surplus available for lending.
A percentage of income could go towards a fund for future tournaments, creating a sustainable source of income to foster a regular tournament calendar with attractive prizes.
A potentially huge improvement would be to set up the renting system based on a public API, similar to the system used for buying and selling cards in the marketplaces. This way not only renting could be embedded in the client but also 3rd parties could also continue to provide renting services, perhaps with different payment methods, like subscriptions for example. Marketplaces could even expand their offers to include renting alongside trading.
Another impactful improvement would be to allow the player to set up default renting limits for each card quality. This would allow for a seamless renting system where rented cards would be automatically added to a deck:
rent cards with the highest possible quality without confirmation, if the fee is under the threshold;
only rent if the card is of higher quality than what the player owns;
only rent if the player doesn't own the card already at the same maximum available quality (from the renting market);
a toggle during deck creation or when importing a deck code would allow the player to choose whether to automatically include rented cards or not.
Handling renting internally would also reduce the selling pressure on the token, since any external renting service most likely sells $GODS tokens regularly as part of their business.
Potential upside for other games in the IMX ecosystem
The renting fees collected by GU could probably justify the additional engineering costs to code the system, but I see a larger upside. This internal renting system could become part of the package Immutable X offers to every game they onboard (like the integration with a marketplace, technical support, or any other services).
Regardless of the blockchain network the games use, a turnkey renting solution with a standard database could be provided, ready to be integrated from the start. This approach would likely be simpler to implement than coding smart contracts. And safer, too.
The rewards earned on this comment will go directly to the people( @agrante ) sharing the post on Reddit as long as they are registered with @poshtoken. Sign up at https://hiveposh.com. Otherwise, rewards go to the author of the blog post.
Just wanted to mention that there is a rental service out there. Its called lootrush.com. I rent my higher quality cards for GU. I also rent out the cards im not using for passive daily income.
Yea, if you read the post, I mention LootRush in the first line... ;)
The first part of the post is why LR is not an ideal solution!
oh sorry! Yeah there is some risk potential involved but is still a solution.
I agree with everything! Immutable X is a very smart and interesting technology, but the tide has shifted toward EVMs. Immutable has rightly shifted its focus to their zkEVM, which is more scalable and universal.
That said, Gods Unchained will never migrate to zkEVM or undertake any heavy lifting. The main purpose of the game was always to promote the Immutable X chain and IMX token.
me gusta mas tener mis propias cartas,aunque es bien que se pueda rentar para el jugador que lo desee.
I didn't know we could rent cards similar to Splinterlands, that was very helpful, thank you.
Thank you for your witness vote!
Have a !BEER on me!
To Opt-Out of my witness beer program just comment STOP below
View or trade
BEER
.BEER
Hey @ifhy, here is a little bit of from @isnochys for you. Enjoy it!Learn how to earn FREE BEER each day by staking your
BEER
.