On the Game "Picking it's Winners"

Personofsecrets · January 8

One thing that I do tend to notice in TOT is that the same card tends to show up in the tavern more often than it should. For example, in the last month of play, I can recall 3 games where the opening position gave a player multiple of the Kwama Mine card.

Although there are some biases in noticing patterns, we also have pattern recognition for a reason. And I don't think that I'm remembering only the key moments of games in a bias way. This phenomenon just seems to always happen and has obnoxious effects with a number of cards.

If the RNG seeding in the game is not quite right and tends toward dupicate RNGs, then combo based play will tend to have unbalanced advantages compared to just buying the best card in the tavern. Indeed, some average cards become quite good when you start getting multiples of them.

And this issue goes beyond just TOT RNG. I feel like I find it in many places of the games RNG systems. Loot drops being repeated down to the same trait, crate legendary and Apex drops being repeated with crates opened close in time to eachother, and, as I noted in another thread, trophy drops used for the TOT achievements.

https://forums.elderscrollsonline.com/en/discussion/comment/7942804/#Comment_7942804

https://forums.elderscrollsonline.com/en/discussion/comment/7937484#Comment_7937484

And although I feel like I've played enough games to see that the trend of cards repeating more than they should is real, maybe someone good with numbers can give the statistics?

What are the chances of any of the 5 cards in the tavern either being alike or being replaced with their copy within a pick or two? I don't feel like the chances are that high, but the phenomenon happens in a large number of games.

Look for yourself. Maybe TOT doesn't pick it's winners, but there do seem to be some RNG seeding issues.

Note, there has also been some evidence that the first player vs second player seeding is broken as to cause repeats as well. There was also some debate about how beneficial such broken seeding could really be, but we don't need to say if it helps players or not for this debate. It's sufficient just to say that it isn't working well.

https://forums.elderscrollsonline.com/en/discussion/638738/proven-rng-of-starting-position-is-broken

https://forums.elderscrollsonline.com/en/discussion/comment/7929856#Comment_7929856

Pixiepumpkin · January 8

I have seen the same patterns, and this is why some people say that "TOT chooses its winners". Whatever the game uses to seed for "random" is clearly not random and certain playstyles, times of day, or locations might prove a better chance of "luck". We don't know what creates the seed, it could be as simple as log in time.

I opened a few hundred dollars worth of all maker crates. ZERO radiant apex rewards, ZERO apes rewards, 7 of 20 Legendary rewards and in each crate sitting I would get the same superior/ epic rewards over and over. In 15 crate round 1 it might be the same chicken. In 15 crate round 2 a few days later it might be the same dog, in round 3 just moments after round 2 it might be the same tattoo. This means that certain "trash" rewards in given tiers are not given a higher % to drop, but rather the seed has determined it. This is part of what makes it frustratring for players. We can see its not truly random which means its fixed be it intentional or not. Its not a matter of luck per crate open, but luck per crate pack.

This same formula seems to be used throughout the game, including TOT.

El_Borracho · January 8

I tend to focus on NPC games for RNG patterns as PVP games are inconsistent and based on how good your opponent is. Plus, sometimes you just want to relax and play a mindless game of ToT. The one trend I have noticed is when an NPC is going with an agent-heavy strategy (if that's what it can be called), there is a marked drop off in the amount of black hand/flaming wheelbarrow cards that populate but an unusually high number of agent cards, proportionally speaking in relation to a average game. Its typically when the NPC chooses decks like Pelin, Druid, and Ansei. When its with Pelin, there is also a noticeable increase in Shield Bearer cards that populate, where it is almost guaranteed to pop up when a agent is knocked out. Its to the point that if the NPC chooses one or two of those decks, you have to go with Rahjin or Celarus to be able to eliminate agents.

We have all experienced the game where you seem to get second-rate cards forced on you where Midnight Raid, Pounce, or Siege Weapon is hiding behind that card and you are just short of the coin to get it. Or the opposite, where you can't fail and get everything you want. But that is different than this. There is definitely a weighted RNG going on, but I think it has more to do with decks and time of play than who the players are.

Largomets · January 8

Putting on my conspiracy theory hat, I noticed that there is a certain player in the end game who always picks the same two decks every game, with one of them being druid, had on unusually good RNG for their draw combo. They tend to buy a "poor" deck, as in, always buying the "good" cards and end up with a very large "bad" deck from a strategy perspective, and yet the "correct" cards come into their hand almost every time. I actually was going back and forth with another player on discord where they noted a similar thing, and if this person had 20-30 cards in their hand of every color mixed and matched (plus bewilderments), the druid cards came into their hand at the same time, even if they only had 2 or 3 cards of that color, every game, even if the probably was really really low.

While pattern recognition is somewhat suspect and this is anecdotal, the odds of any particular hand coming up when you have 30+ cards should be an astronomical number due to factorials of a truly random set. So, when the "unlikely" combo happens with a strange amount of regularity, you do start to wonder if there's a seeding component to tavern/draw RNG vs. true and pure RNG.

This was actually a large part of why I started moving to Rajhin for every game, as adding bewilderments is sometimes the only way to break up a players snowballs on decks like crow/druid/mora etc.

And going into the game at large, in all the years and hundreds and thousands of times I've cleared vAS+2, to this day, I've never gotten a poly drop. Yet I've been in runs where the same person gets a poly in back to back runs, one time even 3 consecutive runs. While it could just be totally random luck, that sort of pattern is... dubious... I've been in farm groups where we clear 20+ times in an evening, and I've been doing that trial for YEARS... so it's not exactly a small sample pool I'm drawing from.

Personofsecrets · January 8

Largomets wrote: »

Putting on my conspiracy theory hat, I noticed that there is a certain player in the end game who always picks the same two decks every game, with one of them being druid, had on unusually good RNG for their draw combo. They tend to buy a "poor" deck, as in, always buying the "good" cards and end up with a very large "bad" deck from a strategy perspective, and yet the "correct" cards come into their hand almost every time. I actually was going back and forth with another player on discord where they noted a similar thing, and if this person had 20-30 cards in their hand of every color mixed and matched (plus bewilderments), the druid cards came into their hand at the same time, even if they only had 2 or 3 cards of that color, every game, even if the probably was really really low.

While pattern recognition is somewhat suspect and this is anecdotal, the odds of any particular hand coming up when you have 30+ cards should be an astronomical number due to factorials of a truly random set. So, when the "unlikely" combo happens with a strange amount of regularity, you do start to wonder if there's a seeding component to tavern/draw RNG vs. true and pure RNG.

This was actually a large part of why I started moving to Rajhin for every game, as adding bewilderments is sometimes the only way to break up a players snowballs on decks like crow/druid/mora etc.

And going into the game at large, in all the years and hundreds and thousands of times I've cleared vAS+2, to this day, I've never gotten a poly drop. Yet I've been in runs where the same person gets a poly in back to back runs, one time even 3 consecutive runs. While it could just be totally random luck, that sort of pattern is... dubious... I've been in farm groups where we clear 20+ times in an evening, and I've been doing that trial for YEARS... so it's not exactly a small sample pool I'm drawing from.

Don't you do some work with computers? There is some method behind how games determine what the RNG is.

A long time ago, there was a player that claimed to know about good "loot servers." So they would send a ping to see what server they were connected to and then reconnect until on the good loot server. I forget how they thought that they could tell which were the good and bad servers, but it makes me wonder how RNG is figured out and if it could be server related.

Largomets · January 8

I don't work in an area that gives me expertise towards this, but yeah, I am aware that truly random number generation is rare. Also, to be very clear, I'm not claiming this person did anything untoward, and they might not even be aware of what is happening. But you can definitely pick up on some patterns that "should not" exist, and if you start playing thinking that "this might happen," it's wild how frequently it happens. Like "wow, wouldn't it be crazy if I could single hand proc kwarma mines?" and then you go for it and it happens. Your example is valid, I've seen it a disproportionate number of times relative to what a factorial set of of variables should allow.

Personofsecrets · January 8

Largomets wrote: »

I don't work in an area that gives me expertise towards this, but yeah, I am aware that truly random number generation is rare. Also, to be very clear, I'm not claiming this person did anything untoward, and they might not even be aware of what is happening. But you can definitely pick up on some patterns that "should not" exist, and if you start playing thinking that "this might happen," it's wild how frequently it happens. Like "wow, wouldn't it be crazy if I could single hand proc kwarma mines?" and then you go for it and it happens. Your example is valid, I've seen it a disproportionate number of times relative to what a factorial set of of variables should allow.

Worse now is the Mora contracts. Even if I try to remove one by a tavern replacement effect, here comes another. Happens VERY often to the point of incorporating it into strategy.

spartaxoxo · January 9

The only way to know is to keep track of these things across a large sample size. Bias affects all of our observations.

El_Borracho · January 9

spartaxoxo wrote: »

The only way to know is to keep track of these things across a large sample size. Bias affects all of our observations.

Very true. And players tend to remember the negative more than the positive. But there is something to be said for repetitive incidents that should be unusual. For instance, I played an NPC this morning who picked Pelin and Rahjin. He lucked into Grand Larceny on his first turn. At round 3, the bucket consistently populated on his turn. He picked the bucket, burned a coin card, and the bucket was replaced by Siege Weapon. 3 turns in a row. When I picked the bucket, another bucket appeared. By round 5, I was getting wrecked by 3 Siege Weapons in a row, and wouldn't you know it, he had enough coin every time to buy that timely Shield Bearer that kept populating because he also had Grand Larceny. Game over by round 7.

My cards? Oddly, the only Pelin card that populated after I took a card was Knight Commander. Crow and Red Eagle were the other 2 patrons. No Crows. All Red Eagle contract agents and actions and tavern cards. Not one Midnight Raid. Tried to keep up with Rahjins to attack his prestige. Fail. I also make it a habit to not take a cheap tavern card at the end of my turn to avoid the NPC getting a good card, so I wasn't setting the NPC with his bucket-Siege Weapon draws.

Normally, when the NPC uses Red Eagle to burn his deck away as you win. But something has been going on when the NPC picks Pelin. Its turned into a predictable onslaught from the jump, as it has been doing recently. Its also more than bad luck, as I don't see this NPC behavior with other decks. But with Pelin, its a constant destroying a coin card with bucket or a cheap Rahjin/Red Eagle card then reveal Siege Weapon or Midnight Raid that I have yet to duplicate on my end in that situation.

I don't really mind, as its an NPC game, but I do find it interesting and unusual

Personofsecrets · January 9

I think that a good goal moving forward is to identify the type of data that would be needed in order to prove or disprove the speculation here. There is no point in gathering data if the result of the data gathering is still some non-verifiable claim.

AnduinTryggva · January 9

tbh I long came to the conclusion that we have a fake random number generator in that there is quite some portion of non-randomness in it. Specifically with ToT.

Cards that come into your hand should be much more mixed unless you have cleaned your deck with black eagle or the rhajin or stock contracts. However, too often than not they are sort of pre-sorted. For instance I've just played two matches, one with alma, druid, psijic and rhajin with rhajin patron spam against me and another one with crow, druid, black eagle and psijic. Both matches were lost by me because in my last turn that could potentially have pulled my through the opponents threshold win I had ALL useless treasure or bewilderment cards in my draw pile. That is I had ZERO chance to skip the threshold to force the match into more turns.

Pixiepumpkin · January 10

El_Borracho wrote: »

spartaxoxo wrote: »

The only way to know is to keep track of these things across a large sample size. Bias affects all of our observations.

Very true. And players tend to remember the negative more than the positive. But there is something to be said for repetitive incidents that should be unusual. For instance, I played an NPC this morning who picked Pelin and Rahjin. He lucked into Grand Larceny on his first turn. At round 3, the bucket consistently populated on his turn. He picked the bucket, burned a coin card, and the bucket was replaced by Siege Weapon. 3 turns in a row. When I picked the bucket, another bucket appeared. By round 5, I was getting wrecked by 3 Siege Weapons in a row, and wouldn't you know it, he had enough coin every time to buy that timely Shield Bearer that kept populating because he also had Grand Larceny. Game over by round 7.

My cards? Oddly, the only Pelin card that populated after I took a card was Knight Commander. Crow and Red Eagle were the other 2 patrons. No Crows. All Red Eagle contract agents and actions and tavern cards. Not one Midnight Raid. Tried to keep up with Rahjins to attack his prestige. Fail. I also make it a habit to not take a cheap tavern card at the end of my turn to avoid the NPC getting a good card, so I wasn't setting the NPC with his bucket-Siege Weapon draws.

Normally, when the NPC uses Red Eagle to burn his deck away as you win. But something has been going on when the NPC picks Pelin. Its turned into a predictable onslaught from the jump, as it has been doing recently. Its also more than bad luck, as I don't see this NPC behavior with other decks. But with Pelin, its a constant destroying a coin card with bucket or a cheap Rahjin/Red Eagle card then reveal Siege Weapon or Midnight Raid that I have yet to duplicate on my end in that situation.

I don't really mind, as its an NPC game, but I do find it interesting and unusual

You just described my games...but this happens to me against players as well.

I have also noticed NPC's getting their power cards back to back. They always come in 3 cards. Its happened so many times, it simply can not be RNG, and oddly its always at a point where I have a lot of agents on deck.

In fact, my wife even made commentary that its as if the NPC deck is being shuffled based on what you have played. Obviously no one can prove this, but it happens often enough that we have taken notice.

I also RARELYget the "good pick" at the start of the match. I also find it funny that when I do get the lead on an NPC or a player and I have 6 agents played, they get 2 ambush back to back and 2 black sacraments.

I have also seen the game give players and NPC's 3x Harvest Seasons back to back to back. This happens regularly with NPC's.

There is a reason few play TOT and its because most folks can see that there is something wrong with how the system works. Anecdotal does not mean something is not true, it just means it hasn't been proven true or untrue.

As it was said in Wargames. The only winning move is not to play.

Pixiepumpkin · January 10

AnduinTryggva wrote: »

tbh I long came to the conclusion that we have a fake random number generator in that there is quite some portion of non-randomness in it. Specifically with ToT.

Its used throughout the game. You can see the pattern when opening crown crates, you can see the pattern when doing crafting dalies or farming patterns (time of day and player count is a factor).

El_Borracho · January 10

@Pixiepumpkin the NPC deck shuffling (without Almalexia or Celarus) is very interesting to say the least. As in when you see the NPC pick up Armory or Midnight Raid, or even Grand Larceny, and you say to yourself "well I'll be seeing that next turn." While you say to yourself, "I know I picked up Siege Weapon, when do I get to play it?"

I like your point about the Harvest Season cards. I thought it was just me as well, but I have see a lot of those pop up back to back to back with the NPC. Most of the time I'm happy to see it, like when they use the scales to eliminate a card I don't want. But there has been a recurring situation I've noticed when Pelin or Ansei is one patron and Hlaalu is the other patron. The NPC will play 2-4 Harvest/scales cards in a row while they have a Hlaalu agent or card in their hand, unplayed, that has the ability to scoop up a 6 coin value card. "How fortuitous," I say to myself as they shuffle the tavern with Harvest Season or scales until the card they really want appears.

Again, I have waited to see what is underneath a card in the tavern when I have one of those Hlaalu cards ready because sometimes you get lucky. But I have yet to spend 9-11 coins to shuffle the tavern until Armory appears. I don't believe the RNG is written to make those specific incidents occur, but there has to be something to explain this unusual NPC behavior.

Pixiepumpkin · January 11

El_Borracho wrote: »

@Pixiepumpkin the NPC deck shuffling (without Almalexia or Celarus) is very interesting to say the least. As in when you see the NPC pick up Armory or Midnight Raid, or even Grand Larceny, and you say to yourself "well I'll be seeing that next turn." While you say to yourself, "I know I picked up Siege Weapon, when do I get to play it?"

I like your point about the Harvest Season cards. I thought it was just me as well, but I have see a lot of those pop up back to back to back with the NPC. Most of the time I'm happy to see it, like when they use the scales to eliminate a card I don't want. But there has been a recurring situation I've noticed when Pelin or Ansei is one patron and Hlaalu is the other patron. The NPC will play 2-4 Harvest/scales cards in a row while they have a Hlaalu agent or card in their hand, unplayed, that has the ability to scoop up a 6 coin value card. "How fortuitous," I say to myself as they shuffle the tavern with Harvest Season or scales until the card they really want appears.

Again, I have waited to see what is underneath a card in the tavern when I have one of those Hlaalu cards ready because sometimes you get lucky. But I have yet to spend 9-11 coins to shuffle the tavern until Armory appears. I don't believe the RNG is written to make those specific incidents occur, but there has to be something to explain this unusual NPC behavior.

It almost feels to me that the "AI" does not play from a random draw, but rather a draw that is desinged to counter your currently played agents and your upcoming deck.

I have seen the same playstyle in the NPC's favor too many times for this to be RNG.

I get anxiety playing against players because I rarely get a good pick in the first 4 rounds while my opponent gets all the good stuff. I literally feel like the deck is stacked against me.

So, I play against NPC's a LOT. I do my cards across the continent on 4 to 5 toons a day and sometimes I will just play the same NPC over and over and over for hours (just so I don't have to ride around). My initial curiosity here was not about how the NPC's play, but rather the loot pool from the white/greenTribute containers you get. But during this process I noticed the NPC plays certain patterns in the decks they choose, but not just that but in how the treasury favors them all too often with harvest, ambush, black sacrament, blackmail and imprisonment cards. I have seen the NPC get 3 imprisonments back to back (along with ambush/black sac and harvest season.

My wife noticed the same stuff before I even mentioned it. I saw the patterns and then one day while she is sitting behind me playing (our PC's are in the same room) she mentions things I have noticed. Now, its a running gag because once you see it, you can't unsee it.

EDIT: I wanted to add that often I will play really long matches against NPC's. Using cats to whittle down their prestige while I amass a god mode deck so I can achieve dumb stuff like the screenshot. These matches are useful for seeing the same patterns over and over, outside of individual matches....and doing this kind of stuff seems to be getting more difficult. I wonder if they are feeding player games into an AI. Who knows.