ABOUT THE DICE ROLLS

Take one of the most modern, fun and engaging board games created in recent years. Match it with cutting-edge technology. Add a ridiculously well-crafted art style and a ton of extra content. This is the recipe to one of the most thrilling, charming and captivating strategy games you will ever find!
VPaulus
Slitherine
Slitherine
Posts: 7671
Joined: Mon Dec 27, 2010 8:33 pm
Location: Portugal

ABOUT THE DICE ROLLS

Post by VPaulus » Sun Oct 04, 2015 1:18 pm

** UPDATE **

The 1.06 update has a system to prevent (well greatly reduce really) any runs of bad luck for the player. Try it out and see what you think. If you prefer the real random rolls then turn on Iron Man mode in options!

** END UPDATE **

Hi guys,

we are receiving a lot of feedback from the community: much of it is positive and we're happy about it. Some of it is negative, and we are trying to listen and respond to whatever we can, in order to improve the game.

One point that is generating negative feedback concerns the dice. This is a game with a strong presence of randomness, and it requires a lot of risk management strategy, in order to minimize the luck factor. That's the nature of the board game, and it's understandably not everyone's cup of tea.

But many negative reviews (and some of the discussions here) are mainly saying a different thing: "THE DICE ARE LOADED". Usually with a bit more of color.

We talked a lot about this in this thread, but let me clarify it again here, in bold. The. Dice. Are. Not. Rigged. The A.I uses exactly the same random generator as the player, and it's programmed to not even use any veterans (that would legitimately allow reroll) and ammos (that would legitimately allow a +1 on the dice).

This post is not to convince anyone about this. We know it's true, and it's easy to prove it: it's sufficient to log carefully the first 100 dice rolls for you and the AI, and looking at the results. Not SOME of the rolls, not A FEW rolls. 100 consecutive rolls for each side. If someone wants to spend a weekend doing this job, he's welcome to find out that the A.I. does not cheat.

The point is that many players have the impression that the A.I. cheats, due a mix of the human tendency to find patterns everywhere, to only look at the dice when things go bad, and to mistrust the machine. This is affecting negatively the player experience, so we'd like to solve this thing immediately.

This is the strategy we came up with:

1) We will create different difficulty levels. The hardest level will be the current one. The lower settings will grant you additional gear (ammos and veterans), that will allow you to reroll the dice N times, and/or to give a +1 to your roll N times. These are normal elements of the game, we're just going to give the player some of them in every single-player scenario.

We hope that making things easier for newcomers will make them less paranoid about the dice. When they'll get a better understanding of the game mechanics, they'll raise the bar by switching to a different difficulty level, without frustration kicking in.

We will, of course, ensure that completing a mission with a higher difficulty level grants you more points.

2) We will insert a logger that you can consult during the dice rolls. The logger will keep track of the last 1000 dice throws in the game for both the player and the A.I., and give appropriate statistics about it. This will show that the game is indeed fair, especially if you help us showing to the public your statistics once they get significant.

3) We will introduce an "anti-streak randomness" option, turned off by default. This will ensure that there will be no streaks and other things that would naturally occur with true randomness.

But we'd really like to hear the opinion of the community on this.

Would you welcome this kind of "anti-streak randomness" in the game? Do you have any other suggestions on how to handle the dice rolls? The debate is open.

We love this game: we put a lot of time and energy into it, and we plan to support it for a long time, adding new features and content, fixing the existing issues, and trying to create a thriving community around it. Help us to make it succeed.



http://steamcommunity.com/app/365890/di ... 353969058/

IronFist00
Sergeant - Panzer IIC
Sergeant - Panzer IIC
Posts: 181
Joined: Mon Jul 18, 2011 1:21 am

Re: ABOUT THE DICE ROLLS

Post by IronFist00 » Sun Oct 04, 2015 3:28 pm

I think all three of your options are fair and I look forward to the logger the most. As long as the "anti-streak randomness" option toggles and is off by default I am good with that.

Personally I have seen the dice rolls favor me as much as the AI. And as I stated in the other thread I use tactics and cover to gain the best chance of success when I can.

Thanks for being so responsive to community feedback.

Igorputski
Staff Sergeant - Kavallerie
Staff Sergeant - Kavallerie
Posts: 300
Joined: Sat Nov 14, 2009 1:08 pm

Re: ABOUT THE DICE ROLLS

Post by Igorputski » Mon Oct 05, 2015 4:20 am

While it may be beneficial for newcomers now it certainly doesn't do anything for those of us who saw the problem immediately.

If you're so certain you're game doesn't cheat with die rolls why the change? Whatcha covering up? I been around long enough to know when things ain't right in Sing Sing. I smell a Watergate. Is Nixon working for you guys now?

It wasn't enough the rolls had to be in the fire phase but I also saw them in the assault phase. Three times in a row I assaulted and three times in a row I got 1 and 2. That to me just ain't random.

twilightling
Lance Corporal - Panzer IA
Lance Corporal - Panzer IA
Posts: 10
Joined: Wed Dec 28, 2011 9:20 pm

Re: ABOUT THE DICE ROLLS

Post by twilightling » Mon Oct 05, 2015 6:14 am

I like the second option the best.
I am ok with the first, although I do not intend to use it.
I really dislike the 3rd - streaks are naturally occurring in real life and removing them sounds less than dice.
A slightly different option would be to use a "deck" for the dice rolls, it is simple to do and gives you almost perfect although slightly predictable randomness.

BiteNibbleChomp
Lieutenant-General - Do 217E
Lieutenant-General - Do 217E
Posts: 3213
Joined: Mon Jul 01, 2013 6:35 am

Re: ABOUT THE DICE ROLLS

Post by BiteNibbleChomp » Mon Oct 05, 2015 7:03 am

Igorputski wrote:It wasn't enough the rolls had to be in the fire phase but I also saw them in the assault phase. Three times in a row I assaulted and three times in a row I got 1 and 2. That to me just ain't random.
There is a 1/3 chance of getting either a 1 or a 2
There is 1/3 of 1/3 of a chance of getting this result twice (1/9)
There is 1/3 of 1/9 of a chance of getting this result three times (1/27)

1 in 27 is often enough that someone who has played every mission in the game at least once is (nearly) guaranteed to see.

- BNC
The American Civil War is coming soon to Strategic Command WWI!
Discuss here: http://www.matrixgames.com/forums/tm.asp?m=4504986

IainMcNeil
Site Admin
Site Admin
Posts: 13360
Joined: Fri Apr 01, 2005 10:19 am

Re: ABOUT THE DICE ROLLS

Post by IainMcNeil » Mon Oct 05, 2015 9:25 am

The game does not cheat but clearly people dont understand probability so we have to do something about it.

Every game involves ~50-100 die rolls which means that every game is going to have than 1 in a 100 chance happen. When you're talking about d6 than means 4 1/2 rolls in a row for you and four 5/6 rolls in a row for your opponent.

That's all in one game. Across the 36 campaign missions you're going to throw thousands of dice meaning you will see streaks of 5 1's in a row for you and 5 6's in a row for the AI.

You will also see all those results reversed in your favour but likely not remember them. This is how human nature works. I've been playing tabletop wargames since I was 6 and rolled hundreds of thousands of dice. I know how people deal with this as I've played in competitions all over the world and won the world championship multiple times. People see good luck as skill and bad luck as fate. Its just the way we are. Me too!

As we can't fix people we have to change the game if people dont like it. This does not imply there is a bug in the game.....

Falke_MatrixForum
Senior Corporal - Ju 87G
Senior Corporal - Ju 87G
Posts: 80
Joined: Sun Nov 30, 2014 6:17 pm

Re: ABOUT THE DICE ROLLS

Post by Falke_MatrixForum » Mon Oct 05, 2015 5:12 pm

1) Good idea to ease new Players into the game.

2) Good idea to silence the doubters.

3) So so on this, but it is optional

I think part of the perception is that the Scenarios are often very short and with the limited unit numbers, so a bad roll / streak can have a large effect. eg taking out a tank on turn 1 vs spending 3 turns to kill
I found the last german Scenario easy but because a) I destroyed the North tank on Turn 1(Double 6s) blocking that route and i could then dedicate the other to Holding the infanty assault off.
b) the other AI tank twice missed when needing 2+ to hit

Will565
Private First Class - Wehrmacht Inf
Private First Class - Wehrmacht Inf
Posts: 6
Joined: Tue Sep 29, 2015 3:44 am

Re: ABOUT THE DICE ROLLS

Post by Will565 » Mon Oct 05, 2015 8:45 pm

I like the idea of a dice logging. My experience has been it seems the RND slightly and I mean slightly towards 3's

I logged my last 564 rolls that happened in the game. posted here.
DieTracking.jpg
Dice roll log
DieTracking.jpg (111.7 KiB) Viewed 7914 times

genesteal
Private First Class - Opel Blitz
Private First Class - Opel Blitz
Posts: 4
Joined: Wed Nov 28, 2012 1:58 pm

Re: ABOUT THE DICE ROLLS

Post by genesteal » Mon Oct 05, 2015 9:01 pm

Thanks to slitherine to be so close for supporting this game.
BiteNibbleChomp wrote: There is a 1/3 chance of getting either a 1 or a 2
There is 1/3 of 1/3 of a chance of getting this result twice (1/9)
There is 1/3 of 1/9 of a chance of getting this result three times (1/27)
1 in 27 is often enough that someone who has played every mission in the game at least once is (nearly) guaranteed to see.
- BNC
Don't know your source about your stat. ;)
When you roll a 1D6 you have , each time, 1/6 chance of getting any number ...so :
There is a 1/3 chance of getting either a 1 or a 2 > true
There is 1/3 of 1/3 of a chance of getting this result twice (1/9) > false . On each roll you have the same chance. I can roll 9 time a D6 whithout 1 or 2.
Same for the third sentence... It's quite possible, to have neither 1 or 2 rolling 27 ...
I don't know how translate it in English : combinatorial analysis ?
On each roll, each time , you have as many chance to get 1 , 2, 3, 4, 5, 6.
the second roll is independent of the first...
Only hundred our thousands rolls can smooth this stats (theory of large numbers)... ;)

the post above is a perfect illustration of this....(thanks!)
on 564 roll the stats for each number is between 15 and 19 %...so...
the effect of the randomize function is visible on all rolls : the "3" is quite more present... (20%) ...but this is a very difficult task to have a neutral randomize feature in any video games...

VR_Nocturne
Corporal - Strongpoint
Corporal - Strongpoint
Posts: 62
Joined: Thu Aug 04, 2011 12:51 am

Re: ABOUT THE DICE ROLLS

Post by VR_Nocturne » Mon Oct 05, 2015 10:27 pm

What the heck is "anti-streak randomness" ?

You mean if I happen to roll 6's three times in a row, the fourth will not be 6 and therefore if I needed a 6 to hit (and only a 6 will do), it's a guaranteed miss?

In addition since I can only get 1-5, that means now I have a 20% chance of a 1 (say a 1 is the only failing roll), instead of the usual 16.7% chance?

I wouldn't use such a thing - and if it's in, I would HOPE there's some kind of indicator. I'd hate for someone to turn this on and then get screwed because they didn't realize "they had too much of a number recently" so it's axed and no sign of that on the interface. They gleefully roll only to have failing rolls augmented in chance.

I mean, if 1-2 is failing but 6 is axed because "it's on a streak"...that just cut your odds to 60% (3-4-5) instead of 67% (3-4-5-6). Why would anyone even want to do that? Folks that want this do realize any number can "go on a streak", not just the ones and twos?

I mean, it's a 1D roll - so it's flat chances. Not even a curve like with 2D6 or such.

VR_Nocturne
Corporal - Strongpoint
Corporal - Strongpoint
Posts: 62
Joined: Thu Aug 04, 2011 12:51 am

Re: ABOUT THE DICE ROLLS

Post by VR_Nocturne » Mon Oct 05, 2015 10:33 pm

Igorputski wrote: It wasn't enough the rolls had to be in the fire phase but I also saw them in the assault phase. Three times in a row I assaulted and three times in a row I got 1 and 2. That to me just ain't random.
What's not right in Sing Sing is this idea that "I rolled low numbers three times in a row, therefore the game isn't using actual randomness".

Seriously? Three results is enough to call an RNG biased for flawed? That's your case? Three bad rolls?

Get a real die, roll a one or two 3 times - does that mean it's automatically loaded?

VR_Nocturne
Corporal - Strongpoint
Corporal - Strongpoint
Posts: 62
Joined: Thu Aug 04, 2011 12:51 am

Re: ABOUT THE DICE ROLLS

Post by VR_Nocturne » Mon Oct 05, 2015 10:37 pm

IainMcNeil wrote:The game does not cheat but clearly people dont understand probability so we have to do something about it.
Why?

Unless you're moving to a model with NO randomness, there will always be RNG and people mentioning how the RNG is "cheating".

So what if they keep complaining because ANY randomness opens the chance for scenarios to go against the player, by definition?

Are those of us who deal with it going to have to keep re-learning the game to learn the new combat systems?

twilightling
Lance Corporal - Panzer IA
Lance Corporal - Panzer IA
Posts: 10
Joined: Wed Dec 28, 2011 9:20 pm

Re: ABOUT THE DICE ROLLS

Post by twilightling » Tue Oct 06, 2015 4:14 am

genesteal wrote: On each roll, each time , you have as many chance to get 1 , 2, 3, 4, 5, 6.
the second roll is independent of the first...
Each roll is independent on the previous, but when you make a "prediction" they add up: If I had thrown 10 '1' and 10 '2', the chance for my next die to be less than 3 is 1/3, but the chance of my next three throws to be all under 3 is 1/27.

Igorputski
Staff Sergeant - Kavallerie
Staff Sergeant - Kavallerie
Posts: 300
Joined: Sat Nov 14, 2009 1:08 pm

Re: ABOUT THE DICE ROLLS

Post by Igorputski » Tue Oct 06, 2015 12:20 pm

VR_Nocturne wrote:
Igorputski wrote: It wasn't enough the rolls had to be in the fire phase but I also saw them in the assault phase. Three times in a row I assaulted and three times in a row I got 1 and 2. That to me just ain't random.
What's not right in Sing Sing is this idea that "I rolled low numbers three times in a row, therefore the game isn't using actual randomness".

Seriously? Three results is enough to call an RNG biased for flawed? That's your case? Three bad rolls?

Get a real die, roll a one or two 3 times - does that mean it's automatically loaded?
It wasn't just those rolls, it was every roll in every game I played and I played several. I can't remember the last time I rolled a 6. But, I can tell you how many times the AI rolled 5's and 6's. Multiply the games by everytime.

Igorputski
Staff Sergeant - Kavallerie
Staff Sergeant - Kavallerie
Posts: 300
Joined: Sat Nov 14, 2009 1:08 pm

Re: ABOUT THE DICE ROLLS

Post by Igorputski » Tue Oct 06, 2015 12:21 pm

VR_Nocturne wrote:
IainMcNeil wrote:The game does not cheat but clearly people dont understand probability so we have to do something about it.
Why?

Unless you're moving to a model with NO randomness, there will always be RNG and people mentioning how the RNG is "cheating".

So what if they keep complaining because ANY randomness opens the chance for scenarios to go against the player, by definition?

Are those of us who deal with it going to have to keep re-learning the game to learn the new combat systems?
Yes, sucks to be you doesn't it.

Igorputski
Staff Sergeant - Kavallerie
Staff Sergeant - Kavallerie
Posts: 300
Joined: Sat Nov 14, 2009 1:08 pm

Re: ABOUT THE DICE ROLLS

Post by Igorputski » Tue Oct 06, 2015 12:25 pm

Will565 wrote:I like the idea of a dice logging. My experience has been it seems the RND slightly and I mean slightly towards 3's

I logged my last 564 rolls that happened in the game. posted here.
DieTracking.jpg
The problem with that and a new game is if the die rolls go against the new player (entry players) he's going to quit before he ever sees those other rolls. He'll mark it down as AI cheating and give bad reviews to the game. If it happens enough it will basically kill the game like that new game "Hell". Players had a hard time with it during the first few missions and reported it and the forum on both here and Steam are dead and so is the game. Ask Zakblood he'll tell yah.

Tac2i
Administrative Corporal - SdKfz 251/1
Administrative Corporal - SdKfz  251/1
Posts: 147
Joined: Sat Mar 22, 2008 8:57 pm
Location: WV USA

Re: ABOUT THE DICE ROLLS

Post by Tac2i » Tue Oct 06, 2015 12:39 pm

Rolling dice is rolling dice, sometimes good luck is with you and sometimes not. As has been stated by Slitherine, the dice are working in the same way as physical dice. Yea, it is frustrating to have a streak of bad rolls but that is the nature of probability. I remember a board game I played many years ago. One player got so angry at his die roll that he threw the dice against the wall. I've also seen players line up physical dice in their hands and just barely roll them on the table in an effort to improve the result. Just saying...
Igorputski wrote:
VR_Nocturne wrote:
Igorputski wrote: It wasn't enough the rolls had to be in the fire phase but I also saw them in the assault phase. Three times in a row I assaulted and three times in a row I got 1 and 2. That to me just ain't random.
What's not right in Sing Sing is this idea that "I rolled low numbers three times in a row, therefore the game isn't using actual randomness".

Seriously? Three results is enough to call an RNG biased for flawed? That's your case? Three bad rolls?

Get a real die, roll a one or two 3 times - does that mean it's automatically loaded?
It wasn't just those rolls, it was every roll in every game I played and I played several. I can't remember the last time I rolled a 6. But, I can tell you how many times the AI rolled 5's and 6's. Multiply the games by everytime.

genesteal
Private First Class - Opel Blitz
Private First Class - Opel Blitz
Posts: 4
Joined: Wed Nov 28, 2012 1:58 pm

Re: ABOUT THE DICE ROLLS

Post by genesteal » Tue Oct 06, 2015 2:54 pm

twilightling wrote: Each roll is independent on the previous, but when you make a "prediction" they add up: If I had thrown 10 '1' and 10 '2', the chance for my next die to be less than 3 is 1/3, but the chance of my next three throws to be all under 3 is 1/27.
A "prediction" is not a statistic*, nor a mathematical behaviour. You hope to make a result doesn't make the result.
As I say, the smoothing (eg chance to obtain different results each other in rolls) is made by very large numbers of rolls.
In combinatiry , the best way to master rolling dice (their results) is to roll 2 or 3 dice together : then you have better chances to get median result (middle total of their number) then on a only dice roll. This is the better way to master chance and this is why many games use the roll of severals dice for featuring many actions.
The same effect is found in the board game : HoN gives a large part on chance... and this is quite mastered by using cards. Infortunately, cards are missing here.
It seems the purpose of this post is not to get any consideration on statistic of a real dice but to control if the randomization of the soft is OK and doesn't make "artefacts" to produce fake results on following rolls...

* You cann't make a "trend" (statistical meaning) on the roll of an only dice builded on previous rolls. I've played wargames and boradgames during many years (almost 30, I'm 60) and some players were able to roll 6 on 1d6 , on 6 or 8 consecutive rolls and others with same "chance" on different results so...
Last edited by genesteal on Tue Oct 06, 2015 3:08 pm, edited 1 time in total.

IainMcNeil
Site Admin
Site Admin
Posts: 13360
Joined: Fri Apr 01, 2005 10:19 am

Re: ABOUT THE DICE ROLLS

Post by IainMcNeil » Tue Oct 06, 2015 3:05 pm

We feel the issue is more one of difficulty than pure random number generation. People only rage at the random numbers if they lose. You never see someone getting upset about 2 6's in a row when they roll them, or the AI throwing a 5 when it needed 6 to hit. We're looking for options to make it easier for new players. Hoping to have something to test tomorrow.

WAGinWNY
Private First Class - Wehrmacht Inf
Private First Class - Wehrmacht Inf
Posts: 8
Joined: Sun Dec 14, 2014 7:43 pm
Location: Western NY

Re: ABOUT THE DICE ROLLS

Post by WAGinWNY » Tue Oct 06, 2015 3:09 pm

Rolling dice is what makes it a true board "game." If you look at 'Panzer Corps,' you can switch between three modes of chance for resolving combat, one of which is "chess," meaning that attacks are resolved by the units' base statistical values only.

Instead of offering the anti-streak option offered above, which breaks the very mathematical meaning of *RND (theory of large numbers included), it might be better to offer an option of increasing the base statistical values of the units so that the dice rolls become less important. For me, however, what makes board games enjoyable is the pucker factor you experience when you roll the dice, because you know in advance that you will have to beat the odds to be successful. (Kind of like in real combat, or so I have read.)

Post Reply

Return to “Heroes of Normandie”