|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
This was an original idea by Eroschilles, i believe. The purpose is to show people how to find out the statistics for making saves in the game and even pulls from a case or booster. To start out, I am only going to do the simple math involved in statistics. This will involve simple multiplication, division, addition, subtraction, and fractions. Later on, I will add the more difficult nCr probabilities. Those simply make it faster to figure out very large probabilities, such as the chances of getting your most wanted pieces in a case.
To Start: The d20 has, of course, 20 sides. This means that each number is essentially 5%, or 1/20. A save of 11 or higher basically means above 10, so the chance of success is 10/20, or 50%. That is the pretty much all you need to know about the d20. Single Saves: Since you know how the d20 works, you also know how to figure out your chances of making saves.
Here they are: save 6: 75% success, 25% failure save 11: 50% success, 50% failure save 16: 25% success, 50% failure
Multiple Saves: But what about multiple rolls for one save? Simple multiplication works good for this.
Here is an example of a roll where you must make two saves: save 11: 50% 2nd save 11: 50% .5*.5=.25 that means you have a 25% chance of succeeding both rolls, and a 75% chance of failing one or both of the rolls.
Here is an example in which you have two chances to make a save: save 11: 50% 2nd save 11: 50% This is trickier, since you must account for all outcomes. The simplest way to do this is to write them all out and then add up the different chances of success and failure. 1. succeed: .5=.5 2. fail, succeed: .5*.5=.25 3. fail, fail: .5*.5=.25 Success: 50%+25%=75% Fail: 25% That means you have a 75% chance of succeeding overall, and a 25% chance of failing.
The same logic can easily be applied to more than two saves.
Mettle Reroll: This essentially works the same as multiple saves, but you add 4/20 to the roll, or 20%.
Here is an example anyway:
Force Point reroll with Mettle save 11: 50% 2nd save 11: 50%+20%=70% 1. succeed: .5=.5 2. fail, succeed: .5*.7=.35 3. fail, fail: .5*.3 =.15 Success: 50%+35%=85% Fail: 15% That means you have a 85% chance of succeeding overall, and a 15% chance of failing.
In Practice: Lets combine them all in a demonstration of ****'s infamous soresu style mastery rolls:
****'s Force Point rerolls with mettle save 11: 50% 2nd save 11: 50%+20%=70% 3rd save 11: 50%+20%+20%=90% 1. succeed: .5=.5 2. fail, succeed: .5*.7=.35 3. fail, fail, succeed: .5*.3*.9=.135 4. fail, fail, fail: .5*.3*.1=.015 Success: 50%+.35%+13.5%=98.5% Fail: 1.5% That means you have a 98.5% chance of succeeding overall, and a 1.5% chance of failing.
I will do the probability on cases later. If you find any mistakes, please notify me. P.S.: Please do not turn this into a **** argument, it was simply a demonstration since he had all of the different abilities necessary for me to show the scenario.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
Cases: This is a little more complicated. I am going to assume there are no doubles of rares or very rares in the cases. Since the forty minature sets, you get all but four very rares, so figuring these probabilities out is much easier than before. You can do this the long way and find out every possible case, then find out how many of those have what you want. This will take a while. There are 70 different combinations if you discount order, 1680 if you do not. Also, I guess you could just record your pulls with the the "Open Boosters" function. That would also take a while. I have a much shorter method that follows.
To start, you must know the mathematical terms.
Factorial: Factorial of x. The factorial of a natural number x is the product of all positive integers less than and equal to x. Its symbol is an exclamation point. example: 5!=5*4*3*2*1=120.
nPr: The number of possibilities for choosing an ordered set of r objects (a permutation) from a total of n objects. How you get it: nPr=n!/(n-r)!
nCr: The number of different, unordered combinations of r objects from a set of n objects. How you get it: nCr=nPr(n,r)/r!
To Start: We will be using nCr since the order of the very rares does not matter in a case.
You start out with a simple fraction: [(what you want)(what you do not want)]/(total possibilities)
The hard part is figuring out what to put in the fraction. We can start with the bottom because finding the total number of possibilities is easy. It is simply 8C4.
Now our equation looks like this: [(what you want)(what you do not want)]/(8C4)
Now for the top. For "what you want" Out of the eight, you put in what you want for n. Leave r alone for now. For "what you do not want", put in the rest of the very rares in n (example: if you want 3, put in 5 for "what you do not want"). r is all up to you, you can manipulate it to find out your chances for getting various figures. If you are confused about this, I have examples below.
In Practice: I will do a series of examples below.
Say I want 4 very rares out of the set.
Here are my chances of getting all four out of one case: [(4C4)(4C0)]/(8C4) [(1)(1)]/(70) 1/70 1.4286%
Three of the four: [(4C3)(4C1)]/(8C4) [(4)(4)]/(70) 16/70 22.8571%
Two of the four: [(4C2)(4C2)]/(8C4) [(6)(6)]/(70) 36/70 51.4285%
One of the four: [(4C1)(4C3)]/(8C4) [(4)(4)]/(70) 16/70 22.8571%
None of the four: [(4C0)(4C4)]/(8C4) [(1)(1)]/(70) 1/70 1.4286%
Chances for Pulls in a Second Case No missing pieces: 1.4286% At least 1 missing piece: 98.5713% At least 2 missing pieces: 75.7142% At least 3 Missing pieces: 24.2857% All 4 missing pieces: 1.4286%
Chances for Pulling the Entire Set
1 case: 0% 2 cases: 1.4285% 3 cases: 25.5102% 4 cases: 55.7141%
I will add more upon request.
I realize that this can be confusing. Please ask for clarification if needed.
Also, if you find mistakes I would appreciate it if you notify me.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 7/30/2008 Posts: 1,290 Location: Stow Ohio, just north of Dantooine (vacay on Ando)
|
did you know that 86.356 percent of stats are made up on the fly?
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
This is some analysis that I find interesting. Let's look at GOWK, a boosted NR Ossus Guardian (Wedge+Leia), and a boosted NR Jensaarai Defender (Wedge+Leia+LukeFS). Let's assume they are all being fired on while in cover by an opportunistic HK-50 for 40 damage at +14 attack.
=== First GOWK:
The chance of a hit: Defense 26, +14 atk so a 12 to hit. Chance of a hit: 45%
If it hits, assuming 2 force points available for re-rolls, the chance of SSM blocking it is 98.5% (as described above).
So the expected damage from this attack is 40*.45*.015 = 0.27 damage.
=== Now NR Ossus Guardian with evade and free re-rolls:
The chance of a hit: Defense 22, +14 atk so an 8 to hit. Chance of a hit: 65%
If it hits, it will be evaded 75% of the time. If it is not evaded, the full damage will only get through 6.25% of the time. .25*.25 = .0625.
So the expected damage from this attack is .35*0 + (.65*.75)*0 + (.65*.25*.0625*40) + (.65*.25*.9375*30) = 4.98 damage.
=== Now NR Jensaarai Defender:
The chance of a hit: Defense 25, +14 atk so an 11 to hit. Chance of a hit: 50%
If it hits, it will be evaded 75% of the time. If it is not evaded, assuming a force point available to use Lightsaber Deflect, it will be deflected 75% of the time.
If it is not deflected, the armor will reduce the damage by 10 75% of the time.
So the expected damage from this attack is 0*.96875 + (30*.0234375) + (40*.0078125) = 1.016
===
These guys are all tough to take out. Actually, the two non-uniques are tougher to take out than GOWK if they face multiple shots per round and if the shots cause less damage, since GOWK is likely to run out of force points and the Ossus and Jensaarai's armors would become more important. Feel free to run your own numbers for other scenarios if you're curious.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 4/2/2008 Posts: 538 Location: GC, Missouri
|
I am thinking that you may have way too much time on your hands.
Were you listening to Styxx while compiling this info?
|
|
Rank: Acklay Groups: Member
Joined: 11/6/2008 Posts: 49
|
Math I made a spread sheet for this and often use it to solve who would win question like Kol vs kyle JBM.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
Those are interesting numbers, but I think those commanders, especially Leia, are easily dispatched. **** needs no commanders to get his numbers. I think I would give the edge to **** just because of that. Very interesting, nonetheless.
How about this:
HK-50 vs. Fully Boosted Shock Trooper 34 Defense .05 chance to hit Expected damage: 2
How about my method for cases?
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 5/26/2009 Posts: 8,428
|
defender390 wrote:Those are interesting numbers, but I think those commanders, especially Leia, are easily dispatched. **** needs no commanders to get his numbers. Very interesting, nonetheless.
How about this:
HK-50 vs. Fully Boosted Shock Trooper 34 Defense .05 chance to hit Expected damage: 2
How about my method for cases? Plus GOWK is almost unhittable in melee, too. (Which is why he was banned.) Not at all trying to argue the other guys are better than GOWK. But if you want to force your opponent to come close enough for melee, a Jensaarai Defender could probably do it. He can effectively get a couple thousand hit points if all you ever do is shoot at him, especially if it's always for 10-20hp per shot.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 5/8/2009 Posts: 585 Location: New York City
|
With JA Leia, Wedge and Luke force spirt Kol almost has the soresu style and mettel.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 4/19/2009 Posts: 487
|
enough math can destroy the fun of swm Remember, its just a game!
|
|
Rank: Grand Master Yoda Groups: Member
Joined: 9/13/2008 Posts: 508
|
saeseetiin wrote:With JA Leia, Wedge and Luke force spirt Kol almost has the soresu style and mettel. Main problem there: That's 96 points. The *ahem* Soresu Style Mastery figure was 55 points. By himself. No boosting. And wasn't owned by Shooters adjacent to him.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
FlyingArrow wrote:Plus GOWK is almost unhittable in melee, too. (Which is why he was banned.) Not at all trying to argue the other guys are better than GOWK. But if you want to force your opponent to come close enough for melee, a Jensaarai Defender could probably do it. He can effectively get a couple thousand hit points if all you ever do is shoot at him, especially if it's always for 10-20hp per shot. True, very true. I still wonder if the Jensaarai Defender is worth its points, though. I just cannot get over the lack of hit points for that cost. I know his armor helps, but it is still tough to get over. By the way, I am still working on how many cases it would take to reliably get all of the very rares. It will never hit 100%, but I will stop at about 50% and post my findings. 2 cases give you a 1.4286% chance of finishing the collection.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 8/24/2008 Posts: 812 Location: Parkville, MD
|
Thanks for doing this thread. It's good to get math practice.
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 7/30/2008 Posts: 1,290 Location: Stow Ohio, just north of Dantooine (vacay on Ando)
|
defender390 wrote:
I realize that this can be confusing. Please ask for clarification if needed.
I will, when I care
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 8/24/2008 Posts: 812 Location: Parkville, MD
|
I remember seeing over at the WOTC boards a few months back the probability of Boba, BH disintegrating some one being around 9%. Is that right? I thought each time he attacked it was treated as an independent occurrence, putting the probability of getting a disintegration at 5%. Or were they treating the twin as a single occurence?
(I know, random question, but at least it has to do with probabilities).
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 2/9/2009 Posts: 936 Location: Southern Illinois
|
Since it happens on a 20 that would be 5% right? How did they get 9%?
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 8/24/2008 Posts: 812 Location: Parkville, MD
|
I think they got something like 9.75%, which I think they would have used the formal addition rule to do. But that doesn't make sense as each attack is a separate occurence or whatnot. I think they were confused, or I'm confused now. Maybe both...
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
Since he has twin attack, there are three possibilities for his attacks.
1/20=.05
1. disintigration: .05=.05 2. normal, disintigration: .95*.05=.0475 3. normal, normal: .95*.95=.9025
Chance of disintigration on his turn: 9.75% _________________________________________________________________________
I am not so sure about this one, but here is what I have for now as far as the chances of completeing a set.
1 case: 0% 2 cases: 1.4285% 3 cases: 25.5102% 4 cases: 55.7141%
I may add more later. Again, if you find any mistakes, please notify me.
|
|
Rank: Moderator Groups: Member
, Moderator
Joined: 8/24/2008 Posts: 812 Location: Parkville, MD
|
Ok, I see how you came up with that for disintegration. Only problem is that doesn't account for if a bodyguard takes a disintegration from a first attack by Boba. Then he gets he second attack from the twin, even though the first was a disintegration. But, either way the math comes out the same with or without a body guard for his chance of getting a disintegration in a given turn. I always counted each roll independently at 5%. If Boba was given extra attack, would that significantly increase his chances of getting the big D?
|
|
Rank: Advanced Bloo Milk Member Groups: Member
Joined: 12/7/2008 Posts: 396
|
Yes, with four attacks he would then have an 18.5494% chance of a disintigration. The bodyguard secenario lessens his chances greatly. The chances of getting a disintigration twice in a row is .25%.
|
|
Guest |