Unofficial statistics site

Discussion in 'Guilds' started by Farbs, Jan 2, 2014.

  1. Farbs

    Farbs Blue Manchu Staff Member

    Cheers. I think HD23 has picked the right way forward here, looking for a positive standing rather than maximum. I'll set that up when I get the chance.

    Fun side note: Internally, "Standing" is actually stored as Shame, a non-negative integer. 0 Shame is interpreted as 10 Standing for the user interface 'cos it seemed nicer.

    EDIT: Done.
     
    ParodyKnaveBob, Jayce and Flaxative like this.
  2. Flaxative

    Flaxative Party Leader

    Maybe now that it is obvious to everyone that going first in Card Hunter is a strict disadvantage (and that the best first move, and thus the best move at any point in an optimally played game, is passing), this line should be changed? Since it's false? Just a thought. Whether or not you and/or the rest of the team think that this is a flaw in the game, Farbs, it's pretty clear that the 'go second advantage' is a big factor on all ranked maps, and thus that the player win balance doesn't mean much in terms of scenario balance.
     
  3. Farbs

    Farbs Blue Manchu Staff Member

    I think it's worth showing the win balance anyway, since it will show whether/when the balance swings dramatically for particular maps. For example, it demonstrated that the extra distance to the VP in the previous collection of ranked maps didn't throw the game wildly out of balance. That was cool & useful.
     
  4. Flaxative

    Flaxative Party Leader

    Except it didn't, apparently—I have to retract my arguments against SLG from the other thread, and here's why.

    1. If all maps swing balance toward Player 2 by approximately 4 or 5% on average, then Player 2 probably has an advantage in the game.
    2. Thus, a "balanced scenario"—one which does not grant either player any additional advantage—should have a Player 1 win rate of 45.5% or so.
    3. Crundyup's Bridge gave Player 1 the closest to a 50% rate of any map—49%.
    4. So Crundyup's Bridge probably wasn't even.

    Maybe this means that maps should be designed like Crundyup's Bridge in order to mitigate Player 2's innate advantage. But I think that the line about balanced scenarios settling around 50% is shown to be false by the stats to date :\
     
    ParodyKnaveBob and Kalin like this.
  5. Farbs

    Farbs Blue Manchu Staff Member

    I'll adjust that line. The new wording will appear when I update the site with guild functionality, hopefully later in the week.

    In other Card Hunter Meta news, Challonge have updated their API which should make it pretty easy for CH Meta to list CH tournaments and pipe match results through to them. So that's cool.
     
  6. Brian Lancor

    Brian Lancor Kobold

    Flax,
    I don't think this is proof that player 2 has an innate advantage. Player 1 Always has the option of passing, which then leads to a situation which is completely equivalent to having gone second. It may mean that the players going first are more likely than not to do something that is worse than passing. However, if a player is playing optimally, I don't see how going first can be a disadvantage.
     
    PaladinGP likes this.
  7. PaladinGP

    PaladinGP #1 in Spring PvP Season

    This is a very good point, though I would write something slightly different. There's still a slight disadvantage to going first, as you declare something about your hand by showing whether or not you are willing to commit to moving first, the slightly more vulnerable position (e.g. you move, they whirlwind, and you've wasted a move card).
    I tend to consider as player one "Do I think my hand is good enough that I am content to act first: that is, do I think giving the oppo the chance of passing the hand out, and them taking it, will make the situation worse for me?" It's obviously a bit of a guess as to what the opponent is running, and whether they're likely to take this opportunity.

    I see the disadvantage of going first, but I'm surprised it translates into results! I can only guess, and this seems a weak theory, that often people don't consider passing at all and they play all the cards they can before passing. This part is likely, but I see it still requiring player 1 to finish all their plays before player 2 and keep going first in turn 2 where it is likely to matter more to be second. AND then not have the advantage of moving first in turn 3, where that might be good, for example. I struggle to see why the disadvantage, that I do agree with, is translating into results! Are player 1/player 2 randomly assigned, or is there something about the matchmaking?
     
    Qui11 and Phaselock like this.
  8. Flaxative

    Flaxative Party Leader

    It is randomly assigned. And while there may be "first action plays" equal to or better than passing, such as putting unholy wellspring on your firestorm mage, they are probably a minority both of potential "first action plays" AND of actually taken "first action plays." And if in all those cases, player 1 simply passed, you would either see a shift to player 2 losing most of the time OR player 2 simply passing back... I think that playing the first card of the game is probably a huge disadvantage at most levels of play, both based on the stats and on the obvious power of reacting to what your opponent does (rather than being proactive).
     
    Qui11 likes this.
  9. Brian Lancor

    Brian Lancor Kobold

    Grr,
    Of course you're right GP. Passing first isn't the same as going second at all. Also, I can imagine going first is particularly bad against your brand of WW/WWE deck. Against most other decks a first turn move that puts someone in a position that the opp can't realistically get line of sight on is pretty innocuous. But, using your moves first against a ww is bad news. And, passing first then gives you the option of redrawing if you didn't get your WW/WWe or NS,etc.
     
  10. Farbs

    Farbs Blue Manchu Staff Member

    I think this might deserve its own thread. I'll just add for now though that it seems AI players are always reported as player 1 by the API, and they tend to lose.
     
  11. PaladinGP

    PaladinGP #1 in Spring PvP Season

    That's a huge point, and probably explains most of the imbalance.
     
  12. Snorre Liljegren

    Snorre Liljegren Mushroom Warrior

    i wish we could see at ones profile how many times one started as 1st and 2nd :)
     
    ParodyKnaveBob likes this.
  13. Qui11

    Qui11 Kobold

    I was checking out the AIs' Player Profiles earlier this evening. Gary had a win rate of 31%, Melvin 34% and both Mom and Cardotron2000 ( =[Unknown] ) are 42%. Two things struck me about Gary's profile, ie:

    Gary

    Standing:-468
    Best Win Streak:1
    Current Win Streak:0
    Wins:2,542
    Losses:5,639
    Total Games:8,181
    Win Rate:31%
    Time Since Last Battle:9 minutes
    Forum Profile:Gary*!

    First, we know that Gary is dim, but he isn't malicious. Unless he is programmed to tank, he doesn't deserve this standing. Hopefully no well intentioned real life players are suffering a similar stigma, even if not so magnified by games played. (the other AIs had respectable standings, 10 9 7 iirc)

    Second, "Best Win Streak:1" is impossible. Maybe not mathematically, but you get what I mean.

    Finally, that you even provide this information in the first place is wonderful. Thank you.
     
  14. Farbs

    Farbs Blue Manchu Staff Member

  15. Jarmo

    Jarmo Snow Griffin

    Quoted from Farbs' About Standing page (emphasis mine): "Players earn Standing by winning games, or by losing games where they earn at least two points and their opponent earns at least four. Any other loss is interpreted as an early resignation or non-attempt, and costs you Standing.

    I understand that everyone gets blitzed from time to time, but that shouldn't happen often enough to reduce your Standing more than regular games increase it. Unless you are Gary."

    So, Farbs is aware of the phenomenon. Gary often gets zero points. I also suspect especially brand new people also sometimes quit against him before earning four points even when they're in the lead due to just not wanting to continue playing multiplayer at that time or real life intruding etc. Both of these tank Gary's rating.

    Streaks are only calculated if you have a positive standing. Thus, Gary's aren't.
     
    ParodyKnaveBob and Qui11 like this.
  16. Thanks for adding guilds, it's really made me care about winning again.

    On a side note, for individual/guild stat tracking (either or), is there a way you could add
    1) Best Rating Achieved
    2) Last 25/50/100 game win/rate?
     
    ParodyKnaveBob likes this.
  17. Flaxative

    Flaxative Party Leader

    It would also be cool if you could view a contextual history showing only fights between Guild A and Guild B!
     
    ParodyKnaveBob likes this.
  18. I'd like to see rewards for guild rank - possibly guild rating/rank emblems before player names would be cool too. I know that's not a stat, but its stat based
     
  19. Flaxative

    Flaxative Party Leader

    Definitely. I asked Farbs about guild tags in another thread and he said that he might do it after leagues.
     
  20. Farbs

    Farbs Blue Manchu Staff Member

    In the short term I'm looking at loosening up character name restrictions, so people can put guild tags on their characters, and adding guild member titles via a new ceremony.
     

Share This Page