A different take on the Requiem guessing game

[Steel_Rook]

When I first heard of system for killing Kuva Liches, I was concerned about having to guess at the right combination of mods. Having actually tried it first hand, however, that's rapidly becoming my favourite aspect of that entire system - bizarre as that might sound. However, I find that the current implementation of the "guessing game" to be far too simplistic, to the point that it isn't really gamified to any significant degree. It got me thinking, though. Does anyone remember the old Mastermind board game? I use to have a very simple version of that when I was younger, and spent quite a bit of time playing it. That's precisely what the Requiem guessing game reminds me of, and kind of the direction I wish it could go. Join me on our Ship of the Imagination and let's see if we can make that a reality.

The core mechanics

I would alter the core mechanics of how Liches are fought. Rather than downing the Lich three times and testing each Requiem mod in sequence, the Lich would be "fatally downed" only a single time and all three mods would be tested simultaneously. Note that a Lich may still need to be downed several times, but only the final time would actually "count" for the purposes of killing or converting them. I'd argue that Level 1 and 2 Liches would have 1 "down," Level 3 and 4 Liches would have 2 "downs" and Level 5 Liches would have all 3 "downs." Once a Lich is "fatally downed" and the player attempts to use the Parazon on them, the Lich is defeated in all cases. If the correct mods were used, the player has the option to kill or convert them. If the wrong mods were used, the Lich simply fades away and then comes back later one level higher. After a Lich has come back to life, the player is left with a permanent record of the attempt, which I will describe below,

Note - if the player elects to not execute the Lich, it will bleed out and fade away after 60 seconds, offering no information whatsoever. The Lich will still come back one level higher than before.

The information

Once a player has attacked a Lich with their Parazon but failed to kill it, a permanent record of the attempt can be found on the Lich's page. The record would list the sequence of mods used and then a score comprised of three circles. A red circle signifies a correct Requiem mod equipped in the correct place, a white circle represents a correct Requiem mod in the wrong place, an empty circle represents an incorrect Requiem mod. These are always listed in the order of Red -> White -> Empty and don't actually correspond to any particular Requiem mod slot. Basically, after each attempt, the player is told that "You had this many mods in the right place, this many mods in the wrong place and this many entirely wrong mods."

Additionally, Lich Thralls can still be attacked and killed for information. Their number is drastically reduced, but so is the value of their information. The player will get a new piece of information for every 10 Thralls defeated. The information given will be in the form of "what not to do," however. Every piece of information the player gains will inform the player that a specific Requiem mod DOES NOT go in a specific slot. In addition to the above-mentioned information gained when killing a Lich, players would also get the same information on an unsuccessful attempt as they would have from defeating 10 Thralls.

The core gameplay loop

In order to defeat a Kuva Lich, players would want to engage the Lich's Thralls in order to reduce pool of combinations for Requiem mods while also trying to engage the Lich directly in order to test various combinations of their mods. Crucially, the information provided by the Lich and the Thralls would be of different "kinds." Executing the Lich should be the primary means of finding a way to kill it as that gives the most information available, while fighting his Thralls would offer supplementary information with which to narrow the guessing as much as possible. Because the Lich doesn't kill the player and levels up with every encounter regardless of the player's actions (obviously aside from killing it permanently), players should have no reason to ignore their Lich. In case players choose to do so anyway, the Lich will go away on its own, rather than lingering forever.

Essentially, my goal here is to shift the "gamification" of the system away from just a really tedious, boring grind for information and more towards an actual guessing game focused on clues, deduction and educated guesses. Progress through this guessing game should reward the player not with information about what TO do, but rather information about what NOT TO do, laving the player to reason out the correct sequence of Requiem mods by a process of elimination. Crucially, by standardising the consequences of all actions which don't lead to killing the Lich permanently, players should find themselves encouraged to actually engage with their Lich, rather than avoiding it and grinding more.

Thank you for your time.

[Avienas]

53 minutes ago, Steel_Rook said:

-snip-

The main issue of the Kuva liches are not the executing part of the lich, but more along the lines of the pre-&post-grind involved with the things. Though D.E. is starting to take a few `SMALL` steps in the right direction by now stating for the next few fixes they intend to do next:

• You will not get the exact same weapon in a row(yet apparently they conveniently left out the part that the duplicates could easily rotate between eachother, meaning its no real change on getting the weapons, except maybe at the begining, which also creates a hilarious issue, where what if a person eventually WANTS duplicates to upgrade their existing kuva weapon?).
• 4 Depleted mods can be transmuted into a random recharge mod. Dear Clem is D.E. going to learn to QoL this system? Make it 3 mods instead at least, the fact you can only transmute depleted mods for a single random one, basically means you need to take down 6 liches before you can scrap 4 out of 6 of them for 1, with a total of 12 liches to make 3 of them. Just D.E. tried to wedge this system into riven mods which are already scarce as heck and its only because of many people do sorties, that a supply exist. It may be possible to farm loads of requiem mods from requiem relics, but i do not recall the mods being tradable, meaning the system barely acts as a recycler, since we all know how long it takes just to take a single lich down.
• Tradable `converted liches`, so it sounds like this system basically has a player able to trade(sic) their lich on someone else in a sense, so they become feral again except for a different player. If this system does not have a huge number of checks in place, then people are going to be spamming market chat `WTB X Kuva lich weapon, 40% or higher X damage type, 10 plat.` There is a reason why that if d.e. had to improve the system, i would rather they keep away from a trading system, since the one they already have is in a poor, exploitable mess. Since if they lock it to clans only or even just a once a week limit, its still going to be abusable in some form, plus if it uses a different trading post then the normal one, Then that pretty much opens it up straight for being exploited by scammers en-masse.

So thats basically me shooting down the recent `fixes` they decided to start working towards, but i still rather see them, if said system is more oriented to a single player experience focus, with next to no worth on multi-player interactions, they should get rid of the grind elements that DEMAND a group of people or take the time to fix said elements, before cramming in new things, based on that broken system (namely the relic sharing gag and paultry amount of void trace stocking systems).

Plus try to remember, the Murmur portion of the Kuva Lich system is entirely optional, it literally serves no purpose except for you to get certain info, which could probably be abused by exiting the mission early, after learning a symbol, so one can skip out the lich tax & your lich killing you and leveling up. Especially if their is no specific order on how the symbols are revealed by the murmurs.

Edited November 12, 2019 by Avienas

[FireSegment]

53 minutes ago, Steel_Rook said:

Essentially, my goal here is to shift the "gamification" of the system away from just a really tedious, boring grind for information and more towards an actual guessing game focused on clues, deduction and educated guesses. Progress through this guessing game should reward the player not with information about what TO do, but rather information about what NOT TO do

One way of doing that in my mind since the day they introduce the 3 requiem system is that you kill the lich to get a "grammar rule" on the order of which requiem go after, or before which requiem, and the thrall would get you the actual correct requiem. Then those would be design so you can at least narrow in on either the specify requiem, or order, down to 3 - 5 options after you got the 2nd one reveal. It would make the requiem have more of a language feel and spell-feel, if you get what i mean, than the system we got now.

[Steel_Rook]

2 hours ago, Avienas said:

The main issue of the Kuva liches are not the executing part of the lich, but more along the lines of the pre-&post-grind involved with the things. Though D.E. is starting to take a few `SMALL` steps in the right direction by now stating for the next few fixes they intend to do next:

Right, but that's not what I'm addressing here, though. I'm specifically addressing the "guessing game" and the mechanics DE chose to implement it with. As it stands right now, the system is all but entirely luck-based, with each new attempt and each new Murmur only narrowing the opportunity space without really improving your ability to guess correctly. My aim with this system is to give players more clues as to the right combination on every attempt and narrow the opportunity space only on Murmurs, so that the "guessing game" is an actual game, rather than just trial and error.

 

2 hours ago, FireSegment said:

One way of doing that in my mind since the day they introduce the 3 requiem system is that you kill the lich to get a "grammar rule" on the order of which requiem go after, or before which requiem, and the thrall would get you the actual correct requiem. Then those would be design so you can at least narrow in on either the specify requiem, or order, down to 3 - 5 options after you got the 2nd one reveal. It would make the requiem have more of a language feel and spell-feel, if you get what i mean, than the system we got now.

Right, that's a different approach to it. "A before B, but not after C." I can't really argue which is better necessarily, but we seem to be thinking along the same lines. Give us additional clues with which we can do our own deductions and our own exclusions, then let us figure it out from there. I personally went with the Mastermind approach simply because that's what I played as a child, but this isn't bad, either. In fact, it reminds me of a puzzle in Resident Evil 2 remake, where six chess-piece-shaped fuses needed to be arranged into six fuse boxes, three on each wall facing each other. The clues provided were to the tune of "The Queen and the Rook were opposite each other, I know the Queen and the King were not on the same wall, I remember the pawn was on the left." That's not an exact quote, but that and a few labels was enough to properly arrange six pieces.

If I may go off-topic for a moment... DE have an unfortunate tendency to replace game design with grind. Rather than "gamifying" systems to where dealing with them is a combination of skill and reason, they're all too often reduced to just repeating a simple activity until the "puzzle" solves itself for you. The Requiem guessing game has the potential to be an actual game with actual gameplay, puzzle-solving and logic elements, and it has the potential to accomplish this with a LOT less grind. Giving players clues to the solution rather than... the solution means we can get far more than three clues before we have enough information to solve the puzzle, which means each clue can take less grind AND the result is a compelling puzzle worth thinking about a little bit.

[Steel_Rook]

I got to thinking last night... It's all well and good that I make these suggestions in theory, but how would my proposed system work out in practice? So, I decided to do a few theoretical runs to see how many attempts I'd need to correctly guess a Requiem sequence with no additional information, using only the rules of Mastermind. For a refresher for people who don't feel like reading the OP, the rules go like this: When attempting to kill a Lich, all three of your equipped Requiem mods are tested at once and evaluated via non-specific dots. A green dot means that one of your Requiem mods is part of the sequence and in the correct place, a grey dot means that one of your Requiem mods is part of the sequence but in the wrong place, and no dot (or a blank) is awarded when one of your Requiem mods isn't part of the sequence. The dots are always awarded in the order of Blank to Grey to Green and don't correspond to the order you've equipped your mods in. So let's do a few guessing runs. Each will start with a spoiler-tagged true sequence and simulate my attempt to reason it out.

For reference, the full set of Requiem mods is: Fass, Jahu, Khra, Lohk, Netra, Ris, Vome, Xata.

 

Spoiler

Fass Vome Khra

1. Fass Khra Vome: @@@
2. Khra Fass Vome: @@@ (either Fass or Khra is green)
3. Vome Khra Fass: @@@ (either Fass or Vome is green, so Fass is green)
4. Fass Vome Khra: @@@ - SUCCESS

 

Spoiler

Jahu Xata Netra

1. Lohk Ris Xata: @
2. Xata Lohk Ris: @ (shift right, look for green)
3. Ris Xata Lohk: @ (shift right, found one green)
4. Fass Xata Lohk: @ (green is not in slot 1, neither Ris nor Fass are in the sequence)
5. Fass Xata Vome: @ (green is not in slot 3 so it's Xata in slot 2, neither Lohk nor Vome are in the sequence)
6. Netra Xata Khra: @@ (either Netra or Khra don't belong, the other is in the wrong spot)
7. Khra Xata Jahu: @@ (Jahu is definitely in the sequence but in the wrong spot, so Khra doesn't belong)
8. Jahu Xata Netra @@@ - SUCCESS

 

Spoiler

Ris Khra Fass

1. Jahu Netra Xata: nothing (neither Jahu nor Netra nor Xata are in the sequence)
2. Khra Ris Lohk: @@
3. Khra Lohk Ris: @@ (either Ris or Lohk don't belong)
4. Lohk Ris Khra: @@ (either Khra or Lohk don't belong, so Lohk doesn't belon, Ris and Khra are in the wrong spots)
5. Vome Ris Khra: @@ (Vome does not belong)
6. Fass Ris Khra: @@@ (Fass, Riss and Khra all belong, sequence is correct but offset)
7. Khra Fass Ris: @@@ (shift mods right, nothing - guessed wrong)
8. Ris Khra Fass: @@@ - SUCCESS

 

This was probably a bit confusing since I skipped a few steps in my reasoning, so let me explain. In the first instance we were lucky. We got all three of the Requiem mods correct, but two of them were in the wrong places. I knew two things: First of all, it's impossible to have two greens and a grey. If two mods belong to the sequence and are in the correct place, then there's no place for the third mod in the sequence to be BUT also in its correct place - there are only three places. Secondly, if I swap the positions of two grey mods, I'm NEVER going to get two greys again. If neither mod fits into either position, then both of them must fit in the third position, which is impossible. Swapping two greys always gives a green and a grey or two greens. Finally, swapping a green and a grey always gives two greys. I'm moving a green mod away from its position thus turning it grey, and I'm moving a grey mod to the position of a green mod where I know it doesn't belong. Using the above reasoning, I test mod combinations. I swap two at random, I lose a green. This tells me I swapped a grey and a green, thus I know one of the two was green. I randomly swap two mods again (not the same ones), I lose a green again. This tells me the green was the common one between the two swaps. If I know which the green was, then there are only two ways the greys could be arranged, so simply swapping them gives me the correct sequence.

The second one is more complex, because I only got one mod correctly and it's not in its right place. I make two further attempts, shifting the entire sequence right. I don't care about the other two mods yet, I just want to shift the grey mod to its proper place. I was unlucky so this took two tries. Once I have a green, I swap one mod at random. If I lose the green, it means I swapped the green mod, so I know which and where it is. No other mod goes in this place so I can never keep the green in this case. If I gain a green or a grey, it means I swapped out one of the wrong mods with a correct mod. If nothing changes, then I swapped a mod that's not in the sequence with another one which isn't. I know the green isn't in that spot and that neither of those two mods is in the sequence. I do this one more time, and in my case swap another blank with another blank. By process of elimination, I now know where and what the green mod is, and I know of at least four other mods which aren't in the sequence at all. That leaves me with three mods, two of which are in the sequence. I randomly swap the two blanks and in my case I got unlicky - I got a blank and a grey. So one of these mods isn't in the sequence, one is but in the wrong spot. The easiest thing to do here is to swap the locations of the two mods and randomly swap one of them for the remaining mod which I KNOW will be in the sequence. In my case I guessed wrong and swapped out a grey mod, but that still gave me all the information I needed. I knew what one mod was and where it was placed, I knew what two other mods were and where they WEREN'T placed. By process of elimination, only one possibility remained.

The third one I still got a bit lucky. My first guess was completely wrong, which immediately eliminated three of the eight mods. My second guess was guaranteed to get at least one right, but I was lucky and got two, either in their right place. Here I used similar logic as in the first run. If I swap two grey mods, I'm ALWAYS going to get a green and a grey. I also knew that if I swapped a grey mod and a blank mod, either nothing would change or I'd get a green. So, if I swapped two mods and got a green, that meant the third was the blank that I needed to get rid of, the other two were simply flipped in their positions. I wasn't that lucky, though, so I swapped a grey and a blank. Knowing this, I swapped another two mods at random and ended up swapping a grey and a blank again. I then knew which the blank was - the mod which doesn't belong in the sequence - by process of eliminations. I could have tried to save a few steps my swapping around the two greys to try and guess their sequence, but I decided against it. Instead, I kept swapping the blank to try and find the third mod in the sequence. There were only two left, so that was 50/50, so I figured why not. That mod would have told me two things. If it had come up green, then it would mean just swapping the other two would lead to success. I was unlucky so the mod ended up grey. Now, when you get three grey mods, you know you have the sequence correct, but it's offset by a position on either side. I gambled and tried shifting the mods right, but I was wrong - should have shifted left. So I shifted them right again and got the correct sequence.

 

So what's the point of all this?
I went through this mental exercise for one simple reason - I wanted to show that this proposed system actually constitutes a "game" unto itself. It's not simply procedurally going through all the available combinations and hoping to stumble onto the right one through sheer strength of statistics. I mean, you could, but the way clues are presented to you allows you to concoct your own experiments, test various combinations and create your own hypothesis of what the combination could possibly be. There are 336 possible permuations of the 8 Requiem mods in the 3 Parazon slots, but I was consistently able to guess right within ~5-8 moves with no additional information from Murmurs. Now imagine also being able to gain Murmurs which tell you information you didn't already know. In the second example, for instance, I might have known that Khra is not in the first slot, which would have immediately told me that WHY attempt 6 failed and saved me attempt 7. After all, if Netra Xata Khra gives me a grey and a green, I know Xata is the green and Khra can't be in the first or third slot, I know it doesn't belong without further testing, no?

The Requiem guessing game as it exists right now is not much of a game and too much of a blind guess which produces little useful information. It is my hope that it can be turned into an actual puzzle for players to figure out or - if worse comes to worst - corwdosurce here on the forums. I mean, the Orokin Cipher in the Sacrifice is kind of the same general idea, no?

Edited November 25, 2019 by Steel_Rook

[Steel_Rook]

Now, I realise that nobody cares about this any more, but I was curious if my theory could work in actual, non-predetermined practice... So I wrote a small Java application to basically let me play Mastermind. I couldn't do formatted text so I went with a ? for a rune in the sequence but out of position and ! for a rune in the sequence and in position. Also used numbers because I'm typing these by hand in a terminal window and don't feel like typing out long sequences. I gave it a limit of 15 attempts after which time it will fail me, but I've consistently been able to succeed in up to 8 attempts. And this time, I'm not fudging the results to match my intended sequence of steps. This is actually me guessing. I wanted to share a few runs directly out of my terminal window, just so you guys can see that the guessing game isn't too too easy but is simultaneously not too difficult if you apply some lateral thinking:

Quote

1: 1 2 3
2: 4 5 6 ??
3: 7 5 6 ??
4: 8 5 6 ??!
5: 8 6 5 !!!
Congratulations. You guessed right.
The sequence was: 8 6 5

 

Quote

1: 1 2 3 ?
2: 4 5 6 ?!
3: 1 5 6 ?!
4: 2 5 6 ?!
5: 5 3 6 !!!
Congratulations. You guessed right.
The sequence was: 5 3 6

 

Quote

1: 1 2 3 ?
2: 4 5 6 ?
3: 4 2 3 ?
4: 5 2 3
5: 6 1 4 !!
6: 7 1 4 !!!
Congratulations. You guessed right.
The sequence was: 7 1 4

 

Quote

1: 1 2 3
2: 4 5 6 ?
3: 4 7 8 ?!
4: 5 7 8 ?!
5: 7 6 8 !!!
Congratulations. You guessed right.
The sequence was: 7 6 8

 

Quote

1: 1 2 3 ?!
2: 4 5 6 ?
3: 4 2 3 ??!
4: 2 4 3 ???
5: 3 2 4 !!!
Congratulations. You guessed right.
The sequence was: 3 2 4

 

Sadly, as I said this is a Java application so I don't know of a good way to share it without making it complicated plus I'm not sure what the file-sharing rules are on this forum. Suffice it to say, though, that this has been a ton of fun to play around with even as basic as the minigame is and as limited as the implementation can be. And yes, granted - I did get lucky on at least one occasion in every run - it IS a guessing game. That's likely what's made my real test runs somewhat shorter than my previous attempts. However, just thinking about fighting Liches using this system for guessing at their Requiem combination alone has me excited, because I'm actually involved in making the final guess.

Sure, I got most of them down in as few as 5 attempts but that's after I'd been racking my brain about it for over a week and coming up with all manner of "clever tricks." You'll note that in every run, my second attempt was with all-different numbers even if I got a lot of hits. That limits opportunity space and lets me draw some more reliable conclusions with fewer attempts than if I were working with the full remaining set. For instance, on one occasion I got REALLY lucky when a swap gave me three wrong numbers WHILE also identifying two correct ones and restricting the third to one of two. I realise this is just repeating things I've said before, but having had the chance to play around with this for real has been an entirely new experience. Working in theory was entertaining but very "artificial." Playing around with an actual game version of this where I don't already know the answer ahead of time is a legit and fun minigame 🙂

Edited November 30, 2019 by Steel_Rook