1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. Welcome to Smashboards, the world's largest Super Smash Brothers community! Over 250,000 Smash Bros. fans from around the world have come to discuss these great games in over 19 million posts!

    You are currently viewing our boards as a visitor. Click here to sign up right now and start on your path in the Smash community!

  3. Use the Smashboards Store to get awesome Smash stuff and support the site, like a Nintendo Controller or the Wii U - Gamecube adaptor ! Check out the inventory in our store and support Smashboards with your purchase today!

Leveling up Nana's 'Reference AI'

Discussion in 'Ice Climbers' started by Myougi, Aug 1, 2014.

  1. Squid Kid

    Squid Kid
    Expand Collapse
    Smash Rookie

    Joined:
    Apr 7, 2015
    Messages:
    18
    Honestly, using 20XX for testing makes me skeptical on this. 20XX modified CPU properties from regular Melee. That, plus the posts here show evidence though CPU states in debug mode that Nana has to be angered first. Didn't Fly already disprove percent being a factor?


    Edit: Reading back through the thread, I noticed this from achilles:
    https://www.reddit.com/r/smashbros/...bers_trick_allows_for_you_to_manually/cjvt6ur
    Which happens to fall in line with the video showing Nana's address changing posted on page 1.
     
    #121 Squid Kid, Apr 28, 2015
    Last edited: Apr 28, 2015
  2. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    You don't need to use 20XX, I suggested it so you can shield for 8 minutes to prove lvl1 doesn't grab you out of shield.

    Nope, I don't remember any post where Fly disproves percent being a factor.
    When popo is in teeter, nana will act without popo's input, and she will do exactly what I said. At popo 0%, she will never grab an opponent out of shield (even if you do the lvl9 ai 'glitch'), at popo percent above 120%, she will grab you out of shield because lvl9 ai's grab opponents out of shield. She will do it often, almost every other move if you just stand there in shield. She will never do it (or hasn't done it for the multiple 8 minute periods I've tested) when popo is below a certain percent.

    You should try these tests for yourself on vanilla melee with and without the lvl9 glitch, and tell me if there are any differences with what I've posted. If there are, than that will be gladly taken as new information, but as of right now, I was a skeptic, but I'm being convinced the more I test this. There is a definite "she does this/she doesn't do this" thing you can test instead of DI, and 'aggressivenes' which are often hard to measure.
     
  3. Squid Kid

    Squid Kid
    Expand Collapse
    Smash Rookie

    Joined:
    Apr 7, 2015
    Messages:
    18

    I'm testing this right now, and here's something that happened. I got Popo to walk to a ledge and go into teeter and Nana decided to run off, then jump back on. I noticed that if you input anything they instantly resync. Are the two still synced when Nana is close to Popo like it says in the video?

    Either way, I'm going to keep messing with this. I put my trust in the icies gods though.
    [​IMG]
    and
    https://www.youtube.com/watch?v=CVMrHBX19To&feature=youtu.be&t=5m49s
     
    #123 Squid Kid, Apr 28, 2015
    Last edited: Apr 28, 2015
  4. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    When you input something, it takes popo out of teeter which would resync them. You then have to walk slightly towards the edge again to put popo back into teeter. Try to get it so popo is teetering and nana is just standing next to him, staring down the opponent like the killer she is.

    Everything you've posted and that I've read from people like Fly, Achilles, Wobbles, etc is that they still don't know, are "skeptical", "RAM doesn't lie" (in reference to the AI not changing), I'll have to run more tests, not proven in any statistically significant way, could easily just be confirmation bias, etc. I've been waiting for someone to do some tests to see if I can recreate them.
     
    #124 shuall, Apr 28, 2015
    Last edited: Apr 28, 2015
  5. Squid Kid

    Squid Kid
    Expand Collapse
    Smash Rookie

    Joined:
    Apr 7, 2015
    Messages:
    18
    Same thing seems to be happening, regardless of how many times I put Popo back into teeter. Haven't got Nana to stare down the other yet. Maybe I need to pivot or something.

    Edit: Debug mode data from the thread:
    http://smashboards.com/threads/leveling-up-nanas-reference-ai.363085/page-2#post-17375393
     
    #125 Squid Kid, Apr 28, 2015
    Last edited: Apr 28, 2015
  6. DerfMidWest

    DerfMidWest
    Expand Collapse
    Fresh Eskimo

    Joined:
    Mar 31, 2011
    Messages:
    4,063
    Location:
    Cleveland, OH
    On his stream a while back wobbles was saying that there isn't really conclusive evidence yet and that the best way to confirm this would be through a series of a bunch of double blind tests.

    I'm very skeptical that it actually does anything, or that it's significant if it does, because nana on her own is capable of doing some really weird things, they just aren't super likely to happen.

    Regardless, no one has really proved or disproved that the level 9 glitch effects gameplay yet.
    Messes with people's heads though.

    Also generates good luck.
     
    #126 DerfMidWest, Apr 28, 2015
    Last edited: Apr 28, 2015
  7. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    Did you read my post on the previous page? lvl 9 AI will grab you out of shield. Nana AI will only grab someone out of shield when popo is above a certain percent, regardless of lvl 9 'glitch'.
    Try it out for yourself.
    There's a list of things cpu's will start doing at different levels, and percents where nana will start doing those things, which @BigVegeta found out, and now I'm convinced as well, that nana levels up according to popo's percent.
    Things like offensive aerials (lvl 1 and 2 will not do these), offensive b moves (only lvl4 lvl3 and up I think), and grabbing someone out of their shield.
    I don't have them on hand, but there are specific percents where nana will start doing things, like 39% she won't do it, no matter how long you wait, and 40% she starts doing it.
     
    #127 shuall, Apr 28, 2015
    Last edited: May 2, 2015
  8. DerfMidWest

    DerfMidWest
    Expand Collapse
    Fresh Eskimo

    Joined:
    Mar 31, 2011
    Messages:
    4,063
    Location:
    Cleveland, OH
    to be completely fair, I don't check this thread very often.
    I did see the nana leveling up thing, which is very interesting, a little of which was already known (i.e. she generally won't smash attack under 40%)
    But to expand on that, nana can do very weird things whenever she wants. It's just way less likely.
    she's a dumb ho and sometimes breaks the rules.
     
    Squid Kid likes this.
  9. Fly_Amanita

    Fly_Amanita
    Expand Collapse
    Master of Caribou

    Joined:
    Aug 24, 2007
    Messages:
    4,224
    Location:
    Claremont, CA

    The only thing I recall ever doing with percentages was verifying that Popo's percentage does indeed have an effect on Nana's behavior. For an easily quantifiable way of observing this, count the number of frames it takes a Nana at 0% to break out of Yoshi's egg with Popo at varying percentages. I don't remember any actual numbers for that off the top of my head since I haven't done it in forever and don't plan on digging through any old data I have, but the difference is easy to observe.

    The only concern I have with what Shuall posted is that, with the previous claims of Nana's AI having multiple distinct states (only one of which reads the AI level visible at the CSS), I don't see any intrinsic reason to believe that the one she is in this test is the one that actually reads that level data. If BVB knows something that could lessen that concern, I'd be happy to hear it.
     
    DerfMidWest likes this.
  10. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    I'll have to make up an asm code that counts nana's moves over a large time period, and compare the results with popo at different percents.
    I'd also like to get a collection of videos of nana doing blizzard, aerials, or grabbing, and check what percent popo was at when she did these things to see if it matches up with the theory.

    EDIT: also, relevant video added to original post, but here as well https://youtu.be/mkRJ2_RYQf8
     
    #130 shuall, Apr 28, 2015
    Last edited: Apr 28, 2015
    DerfMidWest likes this.
  11. Bones0

    Bones0
    Expand Collapse
    Smash Legend

    Joined:
    Aug 31, 2005
    Messages:
    11,053
    Location:
    Philadelphia, PA
    This is probably already the case, but you should keep track of Nana's % as well (i.e. make sure she's at 0%). They might be combined in the calculation or something. A high % Nana going into rage mode might explain a disparity in behavior when Popo is at 0%. (relevant protip: Fox's lasers have splash damage so they're damaging both Climbers)
     
    #131 Bones0, Apr 28, 2015
    Last edited: Apr 28, 2015
    shuall likes this.
  12. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    That's fair, I'll try it again, keeping that in mind. However, if nana's percent is even slightly different (+-1%) than popo's then we wouldn't see the changes where popo 79% has her not doing aerials, and 80% she does aerials.

    Also, nana AI is going to be slightly different than popo AI, or just slightly simpler. With sopo any level, when the opponent is out of range and on the ground, he will shoot an ice block with more probability as the level of AI goes up. When the opponent is far away from nana, she will simply stand next to popo. When you get within a certain range, she will sometimes roll out or at high popo %'s dash attack. When you get even closer, if popo is above 60%, she will down-b. She doesn't seem to have a long range AI, like a computer sopo, and instead, her long range AI is to simply go idle, or run back to popo.

    I did some testing, and while it's not conclusive, it still gives me some information.
    2 minutes per cpu level with sopo (done with 20XX to easily get sopo, but I redid some of them with vanilla melee at the relevant levels 3, and 4 and saw no differences in the things I was checking)
    30 seconds dashing back and forth in front of sopo
    30 seconds running from one side of the map to the other (essentially evade)
    30 seconds jumping above sopo in center stage
    30 seconds shielding in front of sopo (with shield breaks)
    All done on battlefield with human controlled demo-kirby

    Code:
    moves   jabs    smashes tilts   proj    b-moves aerials grabs   throws  combo
    33      24+11   3       4       2       2  (0)  0       0       0       0
    45      29+18   9       4       3       3  (0)  0       0       0       0
    41      25+17   5       3       5       7  (2)  0       0       0       0
    62      41+12   5       3       4       6  (2)  7       0       0       0
    85      38+17   9       10      8       12 (4)  16      0       0       0
    91      26+14   11      6       5       11 (6)  18      9       8       5
    106     30+16   5       5       6       9  (3)  20      14      8       3
    91      16+10   11      4       12      17 (5)  19      13      8       7
    127     30+8    8       6       29      32 (3)  21      14      8       5       *ab
    
    jabs - are of the format "[total]+[double hit]"
    b-moves - are of the format "[total] ([non-projectile])"
    combo - I'm noting here as a throw to charge fsmash. One peculiarity is that popo's combo seems to be fixed whether or not he has nana, I was wondering why popo would pummel, wait a second, and then throw, and it's because there is probably a down-b being inputed during the wait, but no nana to execute.
    *ab notes the only time during testing where popo did a neutral-b while in the air

    Things that seem discrete
    does not use non-projectile b moves until lvl 3
    does not aerial until lvl 4
    does not grab until lvl 7 lvl6

    Things that could be controlled by probability, but would need more rigorous testing
    'aggressiveness' number of moves per 2 minute period
    between lvl4 and lvl5, the frequency in which sopo uses all moves other than jab doubles
    double jabs at lvl 9
    projectiles
    smashes on shield (seem to be replaced by grabbing)

    Things I didn't take into consideration, but may be relevant
    combos? the only obvious one is grab->pummel->blizzard->charge forward smash
    running vs walking
    short hops
    rolling
    L-cancelling
    crouching (he does this, and I don't know why)
    platform vs ground AI
    defensive AI

    Things of note
    before sopo was lvl 4, his response when my shield would break was to upsmash, after lvl 4 it was an aerial (usually nair)
    during evade, popo would walk towards me sometimes doing neutral b, it was only when he got within a certain range that he would switch to other moves; at lvl 9, sopo would neutral b much closer, and more frequently than the other levels.
    I don't think lvl 8 is less aggressive, I think I was just getting hit away from popo more during that test, and so he had less chance to hit me

    Differences between nana AI and sopo AI
    nana does not use probjectiles (if so, not anywhere near as much as sopo)
    nana does not side b, or up b
    nana does not combo (charge fsmash)

    Otherwise, she shows similar discrete changes with sopo cpu levels.

    I'm going to run tests next by damaging only popo and noting any changes at 20% intervals.
    I may test damaging nana, and seeing if she changes her behaviour.
    Also, putting nana into AI in different ways (teeter, desync double jump, popo on ledge)
     
    #132 shuall, Apr 29, 2015
    Last edited: May 2, 2015
    Bones0 and DerfMidWest like this.
  13. BigVegeta

    BigVegeta
    Expand Collapse
    Smash Rookie

    Joined:
    Jul 29, 2014
    Messages:
    5
    Location:
    Boston, Massachusetts
    It's very important to note that when in teeter Nana will always face an opponent. Test this by walking from one side of the stage to the other while popo teeters. Here is where I noticed she will want to face an enemy. Also as a result nana will never teeter if the opponent is behind popo and will easily be able to teeter if popo is facing the opponent. I'm willing to be you were either playing with no CPU on the or the test dummy was I front of popo making it impossible to set up the situation Shuall and I use to test her actions
     
    #133 BigVegeta, Apr 30, 2015
    Last edited: Apr 30, 2015
  14. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    Nana goes into AI when popo is grabbing ledge as well, but you have to constantly refresh popo if you don't want to die. I'm thinking of testing nana on the teeter on platform edge, on ground edge, and ledge grab desyncs to see if she acts differently.
     
  15. BigVegeta

    BigVegeta
    Expand Collapse
    Smash Rookie

    Joined:
    Jul 29, 2014
    Messages:
    5
    Location:
    Boston, Massachusetts
    Here's a brief video showing how someone was able to prove my theory. Shout outs to my roommate Zila for knowing how debug. Unfortunately the video is showing a flawed test because Nana is taking splash damage but I can assure you her percent does effect her level ups. I'm still unsure if her percent will effect things like action frequency but i'm home now and cant wait to get working on it.

    https://www.youtube.com/watch?v=Q9us8alj_z8&feature=youtu.be

    EDIT: Here's a revised list of my observations, i noticed Shuall has some levels and percents a little off:

    LV 0: 0-19%
    LV 1 : 20-39%
    LV 2: 40-59%
    LV 3: 60-79 % (CPU's below level 3 will RARELY use a special move. They will only use ranged special moves within a fairly specific range. Once level 3 nana can start using blizzard)
    LV 4: 80-99% (CPU's below level 4 will NEVER aerial. This is when Nana will start using aerials)
    LV 5: 100-119%
    LV 6: 120-139% (CPU's will grab shielding opponents. Nana abides by this rule)
    LV 7: 140-159%
    LV 8: 160-179%
    LV 9: 180%
     
    #135 BigVegeta, Apr 30, 2015
    Last edited: Apr 30, 2015
  16. drunk SLOTH

    drunk SLOTH
    Expand Collapse
    Smash Rookie

    Joined:
    Nov 14, 2014
    Messages:
    5
    her ai doesnt change from this reset,

    her AI has always been affected by Popo's(or the main climbers %)

    plup told sometime before apex, that it was old school knowledge that had gotten washed up.

    after hearing this I wanted to know more so
    to test this, i freeze glitched a human ICs, lvl 1 cpu, and lvl 9 cpu.

    no difference in her behavior across these 3 UNTIL i got popo's % higher.
    the main 2 things i noted was after 60-70% nana would blizzard
    after about 120% she would go for attacks more fluidly. Ex, jab, then a pause, then dash attack, then a pause, then fsmash

    i did not go as in depth with the tests as it seems BigVegeta/shuall has but I definitely agree with this as being a legitimate thing,
     
    #136 drunk SLOTH, May 18, 2015
    Last edited: May 18, 2015
  17. shuall

    shuall
    Expand Collapse
    Smash Apprentice

    Joined:
    Jun 26, 2013
    Messages:
    154
    Location:
    Philly
    The most I'd heard or read before was that she smashes more when popo is above 100%. I don't even know if that's right now.
     
  18. DerfMidWest

    DerfMidWest
    Expand Collapse
    Fresh Eskimo

    Joined:
    Mar 31, 2011
    Messages:
    4,063
    Location:
    Cleveland, OH
    uuuuhm I'm not sure how much we actually knew. It was common knowledge that she didn't really use smash attacks when popo is bellow like 40% or w/e and that she'll get more aggressive when popo is at high percent, but I think that's about it.
     
  19. Achilles1515

    Achilles1515
    Expand Collapse
    Moderator

    • Moderator
    Joined:
    Jun 18, 2007
    Messages:
    3,211
    Location:
    Cincinnati / Columbus OH
    So I know this has been beat to a pulp, but the formula for Nana's CPU lvl at any given moment is:

    Nana CPU Lvl = Floor (Popo_percentage / 20.00)

    This is the actual code that gets executed to perform this calculation.

    nana.png

    @Fly_Amanita , @shuall , @Bones0 , @DerfMidWest
     
    #139 Achilles1515, Jun 10, 2015
    Last edited: Jun 10, 2015
  20. Fly_Amanita

    Fly_Amanita
    Expand Collapse
    Master of Caribou

    Joined:
    Aug 24, 2007
    Messages:
    4,224
    Location:
    Claremont, CA

    Thanks for that. It's nice to actually see this.
     
  21. Bones0

    Bones0
    Expand Collapse
    Smash Legend

    Joined:
    Aug 31, 2005
    Messages:
    11,053
    Location:
    Philadelphia, PA
    So Nana's % is irrelevant? Good to know. These should be the percentages at which we can expect an AI level change:

    < 20% - Level 0
    20% - Level 1
    40% - Level 2
    60% - Level 3
    80% - Level 4
    100% - Level 5
    120% - Level 6
    140% - Level 7
    160% - Level 8
    180% - Level 9
     
    #141 Bones0, Jun 11, 2015
    Last edited: Jun 15, 2015
    DerfMidWest likes this.
  22. Fortress | Sveet

    Fortress | Sveet
    Expand Collapse
    ▀▄▀▄▀▄▀▄▀▄▀▄▀▄▀▄▀

    • Premium
    Joined:
    Dec 21, 2005
    Messages:
    16,258
    Location:
    Northern IL

    Is there a possibility for a level 0 computer before 20%? They check for greater than 9 but don't check for less than 1...

    Thats assuming the comments are accurate. I don't know enough ASM off hand to verify.
     
  23. Achilles1515

    Achilles1515
    Expand Collapse
    Moderator

    • Moderator
    Joined:
    Jun 18, 2007
    Messages:
    3,211
    Location:
    Cincinnati / Columbus OH
    Yes, it seems as though there is a level 0, likely titled "dumber than a box of rocks." There is actually a level 10+ as well. I haven't messed with it in over a year, but if I remember correctly, and levels from 10-255 all act the same. And it's not "better" than a level 9, it's just...weird. Try it in Dolphin.

    P2 CPU Lvl is 255 (1.02)
    004800853 000000FF

    The "FF" can be changed to "00" for lvl 0 if you want to try that as well. I don't know if I've ever tried that.
     

Share This Page

Users Viewing Thread (Users: 0, Guests: 0)

We know you don't like ads
Why not buy Premium?