Splinterlands suggestion, changing up gameplay, hampering bots.

in #splinterlands5 years ago (edited)

hobgoblin.PNG

I've been pondering the issue of bots and semi-automated play.


A lot of our top players keep some pretty impeccable records of results and use those to reflexively select what they know to be an effective lineup for the ruleset and mana cap.

Anyone with those records and the requisite cards then has a huge advantage, potentially with only minimal experience, or despite not even being human.

The recent addition of a potential second ruleset is a huge step in the right direction; and I'd like to suggest we build on that momentum with even more proof of brain.

I'd like to see Neutral cards sometimes pop up as mercenaries.

The card would cost zero mana to play, be for that match only, and present the same offer to both players.

The card, level, playing position and DEC cost would all be randomised, within certain ranges per card and per league, eg; Goblin Mechs demand 75-95% of DEC and are levels 6-10 in Diamond League.

If hired, the mercenary would then be locked into their assigned position with only the owned cards attracting a mana cost.

Any match could go in four directions.

  • Both players accept the offer.
  • Neither player accepts the offer.
  • A accepts; B declines.
  • A declines: B accepts.

Both players know the deal on offer, and will then need to formulate a team based on whether or not they've accepted the offer and whether or not they believe their opponent has accepted.

While the card itself is free in mana terms, it's something your opponent can be fairly confident they'll be facing, so you're paying with information.

In this example, Thorns and magic would be their likely response.

If so, you could decline the offer and play Lord Arianthus instead, or accept the offer and play Arianthus in second position; of course that only leaves you with four more slots.

In a high mana match, that 5 mana saving might not be worthwhile.

All of this deliberation needs to go on inside a human head.

It has to be incredibly hard to code a bot to take all of the different possibilities into account and decide whether/not to accept the offer, then pick a viable team.

Either the bot will turn down good opportunities, or accept bad ones.

Let's look at another example.

image.png

The mana cap is low, so there's no competition for slots; but the rulesets mean that a single blast from an Exploding Dwarf or Elemental Phoenix, followed by earthquake damage; are going to kill him off fairly quickly. He's also vulnerable to sniper fire.

His speed is decent, but he's unlikely to have too big an impact on the outcome of the match, unless perhaps summoned with + to life, armour, speed or ranged.

@pioneer888777 isn't sticking to one splinter, so its unlikely he's pushing hard to do his quest; if he's farming for DEC he probably won't want to burn almost all of it securing the services of that Highland Archer.

Will he accept that offer? Should I?


Even an AI bot which looks at my past matches doesn't have enough information to know how I'll react to a specific set of circumstances. Maybe in Earthquake I usually play Earth.

An Elven Cutthroat mercenary pops up; I don't own one and I've always wanted to try double melee sneak, the bot can't know this and goes into the match assuming I'll play Earth again.

I choose Fire and hit hard and fast with the Cutthroat and my Kobold miner; while the bot assumes I'll be playing a long game with double heal and some armour from Spirit of the Forest.

The mercenary card has broken my mental pattern and made my past irrelevant, but there's no way for the bot to know if this is the case; the match has also reduced the value of universal data regarding past matches with those rulesets in that league with that mana cap.


Potential misgivings.

This will hurt my earnings. If I have to choose between losing most of my DEC and losing the match, I may as well flee.

Your opponent will share your hesitation. If you use the card and your opponent doesn't, you have an advantage; as much as you will earn substantially fewer crystals; a win will drive you up the leaderboard, helping with daily quests and season rewards.


Its already too confusing to play; this just makes it harder for humans.

Try to see this as two separate processes. The first decision is important, and complex, but it's only one decision. Yes or No.

Once that decision is made, then you're just choosing a team, keeping in mind the possibility that you might already know one of your opponent's cards.


Why should people who didn't buy the cards get to play with them?

Buying a card gives you exclusive rights to use it in any match with a suitable ruleset, any time.

Having them occasionally pop up as options will give those who don't have that particular card a chance to use it, and may inspire them to buy it; driving up the value of the ones you hold.


Is there a chance the track could bend?

Not on your life, my Hindu friend.


Why only neutrals?

Many players are only building up one or two strong splinters.

An offer from a mercenary from a splinter in which they're weak would give too great an advantage to their opponent.


Why lock them in position?

Because it's much easier for a human to adapt to an obstacle/opportunity like that, than a bot.


Surely people will just make arrangements; agreeing to never accept the offers when matched with each other?

Certainly; but if bot squads are already collaborating internally; I see no problem with people doing the same.

I imagine some guilds would adopt it as policy, for example.

People will be keen to get an agreement from those below them in the rankings; while reluctant to give up the chance to gain an advantage over those above them.


We're in an arms race, so I want the devs to have as many levers to pull and twist as possible, and this could be a big one.

It might be every second match, or one in a hundred.

The offers might offer huge advantage for only a small expense; or they might offer the tiniest head-start at tremendous cost.

The important point is that the concept presents a lot of opportunities for the development team to react to bot activity.

That is only going to become more crucial as card and DEC values increase.


Let me know what you think in the comments.

Sort:  

Overall you're spot on that we need more "proof of brain" (not the Steem kind). We have a number of cool ideas that would add that but they are mostly pretty large projects. I've never really considered something like you're proposing so that's pretty interesting!

Thanks :)
Its been brewing for a while, but didn't make sense until DEC/ECR and Neutrals arrived.
I think it'd also go some way to addressing mismatch concerns, eg, between a player who just stumbled out of silver, and another who just slipped from Diamond.
A higher ranked/better equipped player is less likely to take the offer; while the lower ranked player can choose to sacrifice potential DEC to get an advantage.

That said, a free card isn't always an advantage.

A Grumpy Dwarf as tank, for example; could be a disadvantage, even at max level.
If you take the offer and your opponent declines and fields an Exploding Dwarf (ensuring sufficient speed to go first), between Blast and Trample, you're going to lose the Grumpy Dwarf and almost definitely the next card too, in the very first move of the first round.

How do you code a bot for that? Tell it to always decline particular cards? Always accept?
Tell it to only accept that particular card when Red splinter is unavailable? Only in Reverse speed? Only in matches with Mana cap below 22?

Wait what there's potential second rulesets?! O_O Maybe I will just stay in the low tiers forever, I'm already getting multiple rulesets I hate in a row with this thing I'll probably start getting grouped multiple rulesets I hate in a row XD

The idea is otherwise pretty good, would make things more interesting for sure (and more wtf panic for me XD).

And

Why should people who didn't buy the cards get to play with them?

I guess these people have never loaned a friend a tcg deck before XD

Lend me a what the what - tcg deck? I just play daily prize thing when can..low tiers forever for me i guess - still learning with casual play. I guess one day i try a tournament thingy, ever done one of those?

TCG is trading card game 🙃 my partner taught me how to play Magic shortly after we got together. I’ve played heaps of games and built a few decks and never owned any cards 😆

Pfft you know what a coward I am do you think I’d play a tournament 🤣

Posted using Partiko iOS

Tournaments are good fun. I usually get wrecked in the first round (cough, 20 mins ago, cough @samarek69)
Doesn't matter. Had fun :)

They probably are fun but I'd have to actually pay attention then XD

Currently I play Splinterlands while doing Memrise and Duolingo

Also I'd have to stop being a coward

That's an idea I can get behind.

Another similar idea would be to allow users to choose to have 3* extra MANA for X amount of DEC in random matches.

The problem with that option is that extra mana is always an advantage; and since DEC can be bought in the marketplace, you're giving more scope for people to win by throwing money at the game.

I was thinking more of pay X DEC from your won DEC from that specific battle.

For example: Sacrifice 50% of your potential won DEC for the extra 3 MANA. Lost the match, no DEC needs to be paid.

I like the idea, would certainly annoy a lot of bot users. 👍

Hard to strike a balance. Its easy to make the game too hard for people, but still easy for bots. That'd be an own-goal :)

Nice idea, as time moves on this becomes a larger investment for all of us, in time and a lot of money.

I hope the devs and community keep that in mind when they implement new game features.

Posted using Partiko iOS

I don’t know much about Splinterlands yet. In fact, I have no idea what I’m doing! I guess I just have to play it more to figure out how it all works. Can you recommend any beginners guides?

Posted using Partiko iOS

Playing it more is definitely the way to go. A few guys have made guides, but the pace of development has been phenomenal, which has left most guides outdated reasonably quickly.
The "How to play" section is probably best. Learn about the different abilities; and put your monsters to work. Try and predict the outcome of each match before you hit 'Rumble'. Then watch the match.
If it doesn't end the way you thought it would, watch it again to work out why. Best way to learn.
image.png

Thanks. Are there rewards for playing? Just like potential rewards with making posts on Steem?

Yes, you earn DEC with each battle won, and there are reward cards at the end of each season depending on your position in the global league.

Thanks mate! Like I need another distraction in my life haha!

I’ll have to get into a match with you 😅

Posted using Partiko iOS

To listen to the audio version of this article click on the play image.

Brought to you by @tts. If you find it useful please consider upvoting this reply.

Yeah, pretty cool idea. I would certainly take up the offer to get a merc every now and then. Sounds like a fun addition and with the benefit of making botting more difficult. You should pitch that to the devs.

The pitch is in the posting :D

When that is the pitch, then I pretty much guarantee that it was a big waste of time because the devs are busy around the clock AND PARTICULAR ask for people to directly give them feedback.

I see, i think. Yabap has seen the post at least so I assume mission accomplished?

Which happened while I was reading this post as well, so I never knew that. :P

Anything to confuse the Bots :)

An enjoyable read and I see yabap in the comments so I guess he'll be considering options.

This is more likely to affect the top ranks? I'm not sure how many Bots I face in the gold league - perhaps more than I think.

I'm sure the devs have a good idea of what sort of percentage of players in each league are bots. They could bump up the mercenary drop rate in leagues where bots are the biggest problem.

Sounds like a plan 🤖🤖🤛🏽