Has SG 100% nerfed color stacking?

Anecdotal evidence that stacking has been nerfed is no more or less convincing as anecdotal evidence that they are not. When people complain that SG is tampering with boards, without statistical proof, it is very valid to refute that with anecdotal evidence to the contrary.

2 Likes

Hence the useful thread:

Anecdotes are not data, so it’s useful when people take the time to collect data systematically.

Another gem from @Garanwyn:

FWIW, these are contributing factors to my strong preference for 3-2 stacks, even though I can mono in any color with maxed 5*. The variance of a 3-2 is much lower than a 5-0 with still-substantial benefits from stacked tile damage. Put a healer in that 2-stack, and you’re good to go.

7 Likes

If they were going to nerf color stacking they’d do it by capping the total attack value you can get by having multiple heroes of the same color, not by messing with the tile appearance rates.

4 Likes

Except the null hypothesis is that there is no bias against stacked tiles. As the ones making the accusation the impetus is on the accusers to provide evidence.

Example: If I claim that unicorns are real and that there is a conspiracy to cover up that fact, it should be readily apparent that it would be asinine for me to then declare that those who disagree must prove their point otherwise our claims are equally valid.

3 Likes

Interestingly enough, there does appear to be some sort of soft cap on the damage you get from tiles when stacking. Several of us are collecting data to try to understand how it works.

So I completely agree: if they wanted to nerf color stacking, they’d just increase the effects of the soft cap. Messing with tile appearance rates would be more complicated to implement and unnecessary to boot.

3 Likes

I agree.

Or they could again rebalance the damage equation ( linky, linky)

3 Likes

@Garanwyn are you sure it’s the cap and not the emblems? I don’t see much of a difference when they are not high but I have come across multiple tanks with defence and health upped quite high with emblmes and this in fact seems to make a slight difference.

xx

What difference would you be expecting to see? How often do you calculate expected damage versus measured damage on your tiles at different color stack depths (3, 4 and 5 heroes)?

2 Likes

Hi @Garanwyn
I’m nowhere near as experienced as you are😀
Just asking a question as a lot of ppl in game chat were asking this question and we were thinking if it’s the emblems that make folks feel like stacking has been softened as opponents have higher defence than what we are used to. For example if I punch Guin twice with 3 purple stack and she has no emblems she puffs in the air when she is emblems up high this is not enough as her defence is much higher.

Just saying I don’t think SG would not make this public of they softened the stacking power as this would mean they have lied to the whole community. Yes ppl complaint about odds being crap and materials not being widely available BUT they have always made the odds clear. With stacking they have not announced that using whatever stack will deal less damage than it says on the tin. As for tile appearance, for me, it’s same old same old, good spells and bad spells bit in the end it all events out.

Oh, and I don’t think I ever calculated the damage, this is probably why I get my butt handed to me sometimes😆 it’s just the feeling I get sometimes when I realise I did not pay enough attention to emblems and troops lol

xx

Gryphonkit, my wife, just told me this morning the Devs had set Boards to anti team colors :man_facepalming::woman_facepalming:

I resisted the temptation to tell her she was right. I am a good spouse.

3 Likes

Sorry–just reread my post and ot comes off a lot gruffer than I intended.

The way the damage equation is supposed to work is, for tiles:

Average damage = 33 * ((Sum of Attack)/(Defense * 1.2))^1.35 ( and then the total times 2 if it’s strong color tiles)

So if you up defense, it really cuts down on the damage that comes in, no matter what. But the thing is, total damage for deep stacks seems to lag behind what the damage equation predicts. This is what we believe is the result of a soft cap. It looks like they scale down the exponential as the total stack damage ramps up.

3 Likes

@Gryphonknight lol good one. But anti colour boards have always been a reason for heated discussions😀
I guess everyone going mono knows that this is going to happen, what would be the point otherwise? It would turn the game into is making sure we have a set in each colour and just going in for the kill. Where would be the fun in this?

x

Please don’t misunderstand me as I know it is random and thus not complaining but sharing my experiences if colour stacking. I will actually post vids online to these effects are they have been more and more frequent since the last hero upgrades.

When I colour stack I use mostly 3/1/1 and on the rare occasion 3/2. Also I should point out that Win or lose I will fight to the end rematching every loss as I do it for fun not cup amounts.

My findings!

Doesn’t matter which colour is 3 stacked that colour is greatly limited to the point extent of none at times. Yes sadly enough I have come across many occasions where only 1 or 2 sets of 3 tiles appear of any stacked colour to none at all.

Other times at start up you will get a good decent amount of SC ( stacked colour) and then none to follow.
Another confusing occurrence is sets of stacked tiles which will let off around 15 to 200 tiles in continuous sets loading up the opponents mana so they can all fire away BUT yet out of that 150 plus tiles NOT A SINGLE one of your stacked colours where included meaning the opponents point loss was so minimal it made no impactat all on them what so ever. This happens quite a bit.

There are many and plenty other examples I could point out that happen when colour stacking but I will leave these to when I upload the vids.

The other interesting thing that happens more often than not is when doing a rematch you will lose the first 2 battles in such a way that you have no hope what so ever and then the last rematch will give you such a great board that the opponents have mo chance of winning to the extent that even if your left with a single hero that hanging off a thread at being killed and your opponent has 1 or 2 heros left the tiles will keep dropping your way and the hits from your opponents heros who in the first 2 games killed you in a single shot now can hardly make any damage and this goes on until you win the game. I have proven that happening way to many times and will prove it on the vids to come.

NOTE!
I can understand why players believe colour stacking is in someway (for lack of a better word right now) rigged but you can’t help but forgive that line of thinking when you get something like 20 bad boards and then 2 good ones before the next 25 bad ones and so on. More like getting a 5 of a kind on a poker machine, few and far between.

But lets not forget that although random is exactly that random with winning percentages set at low rates there is more to the settings of the game into how it is played than just the rng which is only the engine blick of a car.

Settings would involve certian affects heros/teams have against other heros/teams SS and this carries through to how heros react against stacked heros etc etc, these settings are what manipulate the game much like your carbi, exhaust, electronics etc etc manipulate your engine block on a car.

Like most RNG’s the outcome is already determined as soon as you start and the graphics are nothing more than an appearance factor designed to make you think your doing to the work based on your hero/team selection. I know, hard to believe but very very true.

So yes RNG’s extremely unpredictable and over time will to some extent even out.


So yes frustrations can be forgiven as it can be for any RNG based game whether in real life or online as long as you understand/learn that it will never change thus leaving you with really only 2 options,

  1. Accept it and play for the fun of it. As I myself do.
  2. leave and don’t play any RNG based games as they all always frustrate you.

Watch out for my posts as I will add them to this thread.

PS

Above is a link explaining a little for those who are interested. Plenty more out there go into much more detail.

This link is from an Engineering School of Graduates.

1 Like

I updated thread to a more suitable, discussion provoking title.

1 Like

Here is an interesting link on randomness which involves gameplay and loot drops.

A must read!

http://jaysaidit.com/2016/03/understanding-randomness-in-video-games-and-related-apps/

Internet articles aren’t a great way to learn about how random number generation works

Here’s the bio of the author:

“Jay has been a freelance IT specialist for 20 years with a focus on user support and troubleshooting for PCs, laptops, and mobile devices running Windows and Android operating systems…[he] graduated with my BS in Software Engineering in 2011.”

Does that scream “expert in random number generation” to you?

For an article from a lay-person’s perspective, at least it’s not wildly inaccurate. But he clearly doesn’t understand how various tests for randomness work. And he’s not really clear on seeds, periods, and the statistics of partial samples from distributions.

I’m glad that you’re trying to read up and learn more about how RNG work, but absent some coursework in statistics, probability and stochastic processes, it’s likely going to be challenging for you to get more than a superficial understanding of the subject.

If you actually want to understand how RNGs work

If you want to understand random number generation from a technical perspective, Donald Knuth’s excellent Art of Computer Programming, Volume 2: Seminumerical Algorithms is a good technical work covering some key concepts and issues in the field. Chapter 3 is the relevant chapter.

If you work your way through that, I’m happy to point you to some more advanced materials on the subject.

3 Likes

Without delving into the rest of your write-up, I feel compelled to comment on this statement. It’s simply not accurate.

Assuming a client-side RNG, it would be true that the sequence of random numbers returned during a match would be deterministic.

But what those random numbers are used for is completely dependent on your decisions. To help you understand this, here’s a simple example:

It’s your turn to act.

  1. Alasie, on your team, has her special hot
  2. There are 5 enemies on the board still, each with less than 500 health: GM, Kage, Delilah, Sartana, Lianna
  3. Delilah and GM both have their specials hot

So there are three very different possible scenarios going forward, and only your choice determines which one you will be in:

  • If you shoot GM, Delilah will heal the rest of the enemy team and generate minions.

  • If you shoot Delilah, the enemy will receive no more healing for the rest of the fight. But GM will do damage to your team.

  • If you shoot one of the other three enemy heroes, Delilah will heal the rest of the enemy team and generate minions, AND GM will do damage to your team.

The RNG has no influence whatsoever on which path you choose, and the sequence of numbers it generates following your decision will be put to very different uses in the game.

3 Likes

Everyone gets bad boards, everyone gets good boards, some complain, some don’t, some see evil plots and conspiracies hatched especially in order to take their favorite tiles away from their boards, bottom line is stacking three strong color heroes against the tank works just as well as it worked last month and the month before that or six month ago… if you expect to go mono all the time and get 15 strong tiles on every starting board, you’re in for a great disappointment…

10 Likes

Can we just delete the thread and pin this post :point_up:

7 Likes

150 tile is over 4 complete changes of the board. If this were accurate, it would be evidence of severe issues with the algorithm.

Please upload a video of this happening as soon as possible, so we can take a look at it. Since you say it happens “quite a bit,” I assume you have multiple such videos in hand?

1 Like