Executive Summary:
- Players who want to use a bot in wild must buy a low cost, recurring Automation License from the DAO
- Bot providers add their info to each battle they submit.
- Botting without a Licence becomes a violation of the Terms of Service.
Lore From the Tome of Chaos:
*Praetoria has a problem.
The first few animated robotic battle mages were useful - they scared away the screeching vultures and harpies, and the Harvesters and Sunkai could finally work the fields. Plus, they made useful battle practice for aspiring young battle mages, on their road to glory.*
But over time, they started to increase. We now see swarms of robotic battle mages. These automatons are everywhere. They utter metallic cries of "VONAK!" at all hours of the day or night - caring not if they win or lose, but only to harvest the power of the Splintershards for their hidden masters.
The League of Vericho has determined that these mechanical mages have become too numerous to abolish - the only way to control these automatons is to start a registry.
Introduction
Botting in Splinterlands is a messy topic. However, most players and investors agree that there are too many active bots, and their ability to operate in perpetuity, with minimal capital, allows an outsize impact on the economy and the playerbase.
REALITY CHECK:
Most web games don't allow botting at all. Here in Splinterlands, we allow it. This is a privilege that any account owner has, and this privilege has value - to earn "playing rewards" without committing time.So if I bot, it EARNS something for ME, but I acknowledge it also TAKES from the COMMUNITY.
This proposal intends to give the community better transparency into automated battle activity, and also to give the DAO more control on the ability for large scale "bot farming" to create negative impacts on the ecosystem, at what is expected to be a very minimal development cost.
There are two goals to this proposal:
- Better Data for Better Decisions: Give Splinterlands and the community better data on botted activity, to make better decisions
- Alignment of Rewards vs Participation: Create an economic hurdle, specifically applicable to accounts using automated battle bots, to reduce the ability to operate in perpetuity with minimal capital or player engagement.
Phase 1:
- Non-transferrable Battle Automation Licenses are to be sold by the DAO for players that want to use a bot to play their account. Cost should be low but recurring.
- Battle Automation services (public or private) are required to add an Automation field to each battle submitted, True or false and optionally the name of the service used recorded on the battle chain.
- Update to Terms of Service - no changes to where you are allowed to bot, but botting > needs a licence, and needs to identify their battles where they are submitted using a bot.
After this is running, we will start to measure and manage, and fine tune:
Phase 2:
- Splinterlands or Third Parties could use battle chain results to obtain reliable data on automated battle volume and reward share since each battle will be labelled with bot activity
- Cost for Battle Automation Licenses can be periodically adjusted for these licenses to manage supply/demand in line with desired match liquidity parameters.
- Better data will potentially help Splinterlands in their ToS enforcement efforts
- Look for opportunities to fine tune and reward real players with bragging rights, badges or real rewards.
Discussion: What are we trying to achieve?
Look, we always want REAL PLAYERS - regardless of their budget. So anyone can manually play in any league and we don't want to discourage that in any way.
But some players have invested a lot, play when they have time, and want to have the bot run sometimes when life gets busy. I am not going to debate the merits of that - it's the way it is.
But I think we can all agree that the number of low quality bots is too many. The sheer volume of these accounts has dwarfed the rest of the playerbase, and has many unfortunate side effects, such as low quality matches in high leagues and excessively diluted reward pools.
For example, at the time of writing, having reviewed a subset of botted accounts in Diamond 1 receiving around $0.12 per day of SPS, and found single BCX rented decks: spending less than $0.02 in DEC for cards and $0.09 for a modest SPS rental, using a multiplier of 3x including guild bonus. This is very scalable profit margin, and so the incentive to continue scaling, and adding accounts is clearly obvious. Second hand accounts are widely available from former botters and disillusioned players for a fraction of the cost of a new spellbook. So all signs point to this situation getting worse, not better. (NOTE: edited to correct error in SPS rental price)
I would propose that the community can help to limit the number of bots by requiring bots to declare themselves, and to pay a small registration fee to the DAO. This is to create an economic equalizer, to prevent unlimited and perpetual expansion of low-quality, high-volume bot farms to unsustainable levels. As we know, that is just a race to the bottom.
Looking at the Numbers
I'd suggest, based on these "real world" bot profitability numbers, that an automated account that earns 5-10 SPS per day, so $0.10 and pays $0.04 each day is not the kind of purely-profit-driven player we are looking for. If there are 25,000 of these, that means 150,000 SPS per day are going to these bots...
So let's set the cost of botting a bit higher than that to start with, while we get some data.
I'm proposing this as a starting point for a license to bot your account.
- $0.10 per day per account, in DEC
- Minimum licence period is 14 days, but you can prepay up to 26 weeks in advance.
- Licence proceeds go to the DAO (since it's the DAO paying the reward pool)
- DAO could decide to burn a portion, or send a portion to the team
- Cooldown period is something to consider.
Potential Outcome A: Unprofitable Bots quit, reward pool share per account increases, some rental income will decrease
I expect that many of the "least-desired" low-value botted accounts will be immediately unprofitable if they pay for a license.
As a result, these accounts may reduce their scale of operations to their higher value accounts (good), or they may try to operate in the shadows, in violation of the ToS. At that point, the team can point to the community will, enforcing the ToS accordingly, and they will have a clear mandate and better data to focus their enforcement activity.
Low end rental income also decreases - but remember that these are accounts that are spending less on rent than they are earning in SPS, so I think this is net positive for higher value assets.
Potential Outcome B: No one quits. Tens of thousands of Bots Farms are somehow still profitable, DAO gets Millions of DEC per Day for bot licenses
I don't think this is the likely scenario. But it's possible. If this happens, we use our data and decide if the price of the Licenses is commensurate with the value received.
I wonder how much impact there would be on SPS burning if we would start locking away that much DEC every day?
Wild players: Just some Food for Thought...
If 20,000 low end bots shut down, do you think that your battle earnings would increase by more than $0.10 per day?
Or what's the impact to your daily SPS if 20,000 bots start to burn 100 DEC each day?
Technical Discussion
All the bot services are perfectly capable (and prefer) to work directly with the game's APIs. So I suggest that essentially no frontend work is needed from Splinterlands except to trigger another ToS update. Adding a page to the Shop is unnecessary. The only changes would be to APIs, defining a couple of new custom block operations and backend battle processing.
In fact, most bot providers have an attractive web interface and already have Keychain integration to authorize blockchain posts directly. So they could be the interface for buying Licenses.
| Step | Responsibility | Comment |
|---|---|---|
| Define: Global License Price | Splinterlands | add a field to game settings API "automation_license_daily_price" |
| Define: Player License Expiration | Splinterlands | add a field to player settings API "automation_license_expiration" |
| Create custom operation: sm_automation_license_update | Splinterlands | Moves the DEC to DAO, update players "automation_license_expiration" |
| Interface to sell Automation License | Bot Service Providers | Can set up a button or field in their interface to authorize the License purchase |
| Add field to battle submissions | Bot Service Providers | add a "automation: True" field to the json data to indicate this is an automated match |
| Add field to battle submissions | Splinterlands | add a "automation: False" field to the json data for matches submitted through the website |
| Validation | Splinterlands | Reject battles without an Automation field or if "automation:true" but "automation_license_expiration" has passed |
Crowdfunding a proposal...
So I think significant community interest means this post gets upvoted to at least 30 Hive and 200+ upvotes, that means there's a legitimate level of community interest. (I'm going to exclude downvotes for the calculation)...
If we hit those two targets, I'd suggest we consider crowdfunding the 100,000 DEC for the proposal.
So if you like it, upvote, reblog, share, get your guildmates to upvote, etc. If you don't like it, downvote if you feel it's necessary, but know that this phase is ignoring downvotes. The threshold for gauging community interest right now is only UPVOTES. Anyone that likes the idea so much they want to help crowdfund the proposal, please make a DEC commitment in the comments.
So then if it gets to 30 Hive, I'll kick in the first 20,000 DEC... and I'd encourage supporters of the idea to offer to commit some DEC for the proposal fee. If we get more than 100K in commitments, then we can do prorata or something.
(Please don't send anything at this time, unless we get enough community support to move forward)
Feedback and Community Voice
I expect others will have other suggestions, so please add your opinionsto the comments below, and if you like the ideas of others, please vote for them so that we have more data.
Please vote for as many as you think you would support if this were a real SPS proposal...
NOTES: made some updates to improve the flow and added Technical Discussion based on feedback
Hey OAA, I like part of it - the registration aspect. I think the cost is too high and you are way overestimating the earnings, so I think that should be majorly reduced.
I also think there are 2 things that will do a lot towards the objectives you seek coming up:
I would like to see that impact before we focus on this aspect. I do think a registration process is good, but when it comes to pricing there is way too much uncertainty right now and I believe it makes sense to put a price tag on it when we have seen the impact from these 2 things.
ps... personally I think a lot of bot farmers are going to quit, I don't think they will be able to survive with the lack of use of free starter cards combined with the requirement that they have both SPS and CARDS too.
thanks Dave for your input - much appreciated!
The cost is the easiest thing to negotiate on! So yeah, I picked a nice round number but it could well be lower if that's better for our community.
I agree the earnings are overstated - I was trying to show that if this change somehow doesn't eliminate any bots, that would be the impact on the tokenomics... but I can understand that looks unrealistic since something like this would definitely take out SOME bots...
But sure, I agree there's lots of change afoot, and taking a wait and see approach is certainly prudent.
I like this idea! The numbers may need some work, but overall I think this would cause bot farms to consolidate their army of low level bot accounts into much fewer higher level accounts. While this may have a neg effect on the 1bcx card rental market, it may have a positive effect on the higher level cards for rent. Lots to consider here thank you for posting this idea it is certainly worth a discussion.
Appreciate you taking the time!. I'd love to get any feedback on the numbers. Too expensive? Not expensive enough?
I agree with Dave that there is nessisary first step before using a " tax". Liscences would disecentivise but if we have seen anything about the nature of bot farms its the low friction scaling that causes the mass drains, ratings inflation and match quality issues,Liscensing requires upkeep and fidelity were as eliminating the fuel of frictionless scale is more effceint. It comes at a cost to new players but the game is slowly dying so perhaps wroth it risk reward.
I love supporting the DAO, Love better data and Love the lore/spirit and shared sentiment we all feel but I am also feeling the crypto ethos might be challenged with centralized bot gatekeeping with Liscences. I keep thinking about how goverments think about issues and this feel verry much like a fishing liscence here in the states.
That's a valid concern, appreciate the input winddrake.
Good analogy, it kind of IS a fishing licence... and if there were no licenses then the rivers would have barely any fish left. And that's kind of where we are at.
So maybe the fishing license is too expensive - that's negotiable :)
But I'll push back, that if it's the DAO that is determining the "fishing licensing terms" then it's not centralized, it's decentralized
Yeah the whole "D" in DAO good point. I guess i was seconding the Dev's view but the DAO is " Decentralized" by definition so fair point. To emphasise I love your idea as a " phase" two of Bot control after/if 1BCX cards stop being given out.
I am starting to wonder if bots can be dealt with in a free market way as all the known prevention measure I know of in other games revolve around " ban " which in my mind isnt verry free market.