[Search] Search   [Recent Topics] Recent   [Hottest Topics] Hottest   [Login] Login
Seemingly Non-Uniform Random Number Rolls
Forum Index » Guides & Tips
The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline

The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline
Hello everybody!

I'm not normally one to talk on forums, but I have noticed something odd and I wanted to know if anyone else has noticed this, or knows why it is happening.

The Situation:

I have 1 marauder and 1 marine, and they are each going to attack a different opposing marine (so no gangup bonus). As everyone knows, the exact damage they do is based upon the "rolls" of a RNG (random number generator), and you can choose which units get the better rolls by having them attack first or second, depending on which is the better roll.

The Problem:

I've noticed in a very small number of cases, reversing the order of attack can increase BOTH units damage output. Using the previous example:

option 1: marine attacks for 3, marauder attacks for 5.
option 2: marauder attacks for 6, marine attacks for 4.

It seems to me that under some specific situations, the rolls do not act the same when different units are used, almost preferring different units to be used in certain orders during a turn. I've noticed this happens most often with varying attacks between air/ground/ranged units, as if each of the rolls have a different value for each of them. I'm not really sure how to add pictures to forums, but if people need screenshot evidence I'm sure I can provide it.

I feel that if roll #1 makes a marines attack lower than roll #2, then the same should be expected from any other unit. Am I missing something?

Thank you for your time!


Member of the balance team
Moderator
Highest Score: 2685
Highest Rank: #4
Current Rank: N/A
Highest Championship Ranking: 4th
Highest Tournament Ranking: 4th
Current Record: 131W, 14D, 4L
Winner of Duaneski's mapmaker challenge
Winner of Angkor's 2v2 Grand mapmaking challenge
Winner of Lkasr's 3-Minute Tournament
[WWW]
[Avatar]
Duaneski

Messages: 1021,
Joined: Nov 27, 2015,
Offline

[Avatar]
Duaneski

Messages: 1021,
Joined: Nov 27, 2015,
Offline
Interesting isn't it?

I imagine it as this:

I am going to roll 2 six sided dice. I need to score a 7 or better to "hit"

I roll a 6. I miss. I roll a 8. Hit. I roll another 6. I miss. I roll a 2, I miss.

So I hit 1/4.

Now, I realize that 1/4 isn't a very good outcome, and I already know my rolls.

So I check my marine attacking instead. He only needs a score or 6 or better to "hit".

Same rolls, he hits 3/4 instead of 1/4.

I don't have the specific numbers or knowledge of how the formula specifically works, but this is how I imagine it, and to me it makes sense
The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline

The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline
You gave a good suggestion, but I feel that only explains half of the situation.

If a marine needs 6 and a marauder needs 7, then wouldnt it always be the case that a roll that increases a marauder attack would also increase the marine attack?

My problem is only when trading places increases BOTH of their attacks, I feel that if the second rolls is better that the first for one unit, it should always be better regardless of the different minimums different units have, because it's numbers must be higher. But that did get me to think about it in a bit different way, maybe next time I come across it, it will make more sense. I know stats can sometimes have. tricky non-intuitive solutions, so I'll think about it more.

Thanks!

Member of the balance team
Moderator
Highest Score: 2685
Highest Rank: #4
Current Rank: N/A
Highest Championship Ranking: 4th
Highest Tournament Ranking: 4th
Current Record: 131W, 14D, 4L
Winner of Duaneski's mapmaker challenge
Winner of Angkor's 2v2 Grand mapmaking challenge
Winner of Lkasr's 3-Minute Tournament
[WWW]
[Avatar]
Duaneski

Messages: 1021,
Joined: Nov 27, 2015,
Offline

[Avatar]
Duaneski

Messages: 1021,
Joined: Nov 27, 2015,
Offline
OH

Yeah. So you're probably correct in your assesssment that both shouldn't increase if that was the only thing at play. Here's another thing I've been made aware of:

Say the roll sequence is:

1,2,7,7,6,6,6,7,7,7,6,7....

That's 12 rolls. Well, depending on the health of the attacking AND defending units, you use up a certain number of rolls. Let's say a 10 health unit uses 4 rolls (I'm sure that isn't it, but for this example..) and a 5 health uses only 2 rolls.

So if you attack with a 10 health marauder first you'll use:
1,2,7,7... then the marine would get 6,6 if it was at 5 health.

So if you flip the order, now your marine gets 1,2 and your marauder gets 7,7,6,6

My example doesn't have them necessarily increasing damage, but hopefully shows how it would be possible.

I'll send a thread link that maybe helps to illustrate this:

http://forum.uniwar.com/posts/list/2349.page

This message was edited 1 time. Last update was at Jun 07, 2017 16:45

The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline

The Impaler

Messages: 147,
Joined: Jun 07, 2017,
Location: 221B Baker St.
Offline
Ahh yes, that's essentially what I was looking for, that makes sense. Thanks a lot for the help! I think this pretty much clears it up Also thanks for being so quick to reply, that's dedication!

If I hadn't already added you in game, I'd add you again!

Member of the balance team
Moderator
Highest Score: 2685
Highest Rank: #4
Current Rank: N/A
Highest Championship Ranking: 4th
Highest Tournament Ranking: 4th
Current Record: 131W, 14D, 4L
Winner of Duaneski's mapmaker challenge
Winner of Angkor's 2v2 Grand mapmaking challenge
Winner of Lkasr's 3-Minute Tournament
[WWW]
Forum Index » Guides & Tips
Powered by JForum 2.1.9 © - 2020-04-14 v124 - UniWar website