Check out the patch notes for the July 16th update, P.E.T. Part 1: http://forums.mabinogi.nexon.net/discussion/23332/
[NEW MILLETIANS] Please note that all new forum users have to be approved before posting. This process can take up to 24 hours, and we appreciate your patience.
If this is your first visit, be sure to check out the Nexon Forums Code of Conduct. You have to register before you can post, so you can log in or create a forum name above to proceed. Thank you for your visit!

Jousting Finals

Comments

  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 6, 2019
    Helsa wrote: »
    I found that to be an interesting tidbit of information.
    It is interesting indeed. (Just not to my liking)
    Helsa wrote: »
    In the write-up I had suggested in situations where you do have to make a pure guess, knowing your oppenents history can suggest a better choice. In the heart of it luck, yes, but as you said earlier using knowledge of your opponent to spot "better" choices.
    Well, there was no knowledge at all, because I saw them for the first time ever.
    Helsa wrote: »
    BTW I am curious, what would be an example of "Statistically approved "best practices" often mean higher overall frequency of successful cases, but not higher chances in each case."
    No-no-no, they're basically called so because of being statistically approved. That's what I meant. If they weren't statistically approved, people wouldn't call them "best practices". I'm highly likely to be mistaken, but I'd call Lileas' suggestion to start from W a "best practice".
    Helsa wrote: »
    To be fair, one can actually be in a "sufficent luck" situation even after the threshold is reached. If the player in the winning position is not aware that they are or how to execute the winning strategy. In this situation the player in peril "would be lucky" that the other player didn't know to or how to press home their attack. Have i got that right?
    Frankly speaking, that basically covers my whole theory. If you're lucky enough, your opponent acts in a way beneficial for you (intentionally or carelessly).
    I'm always imagining a thing like:
    You: Hmm, should I start from W?..
    Opponent: He may be starting from W, I should probably use A.
    You: But what if he thinks I'm starting from W and plays A? I should maybe start from A or S to negate that...
    Opponent: He may think that I'll try to hit his opening W with my A... and play another one... I should probably go S or D then...
    You: But he also can use D to negate my S (theoretically countering his A, which is theoretically countering my W). Maybe I should just risk and start with D to have better options further.
    Opponent: OK, no more time to think. I'll sacrifice my D for his theoretical W.
    DxD
    Opponent: Wow
    You: Oh...

    This thinking line can go on forever, and it's pure luck in terms of which option they eventually choose (or suppose you to choose) (RAWR) :D
    P. S. Of course I realize that few people would have such thinking line.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Musicat wrote: »
    Helsa wrote: »
    I found that to be an interesting tidbit of information.
    It is interesting indeed. (Just not to my liking)

    I understand. As you said, you prefer analysis of perhaps more precise things void of luck, like say, Chess, for example. In the case of this Poker AI, that finding could perhaps at least suggest a kind of mathematical or logical justification for taking more risks than people, empirically have considered "proper". On the other hand it could've arisen from something quite simple. In competition, those with good reputations for being particularly skilled at the activity, can benefit from their competitors being intimidated by them. This results in those competitors either playing more cautiously than normal or more recklessly than normal. When I was still actively competing in Jousting, I noticed, or at least thought I noticed this behavior first hand. In the case of this study, maybe when the AI chalked up some wins, the human Poker champions became intimidated and played more cautiously, allowing the AI to take more chances? When the first Chess AI finally beat the then current world champion (Kasperov?). The Chess grandmaster said that the fact that the computer was a cold precise calculating machine monster that could not be intimidated or get tired, was "something he'd never faced before". Perhaps this happened in the AI vs Human Poker experiment.
    Musicat wrote: »
    Helsa wrote: »
    In the write-up I had suggested in situations where you do have to make a pure guess, knowing your opponents history can suggest a better choice. In the heart of it luck, yes, but as you said earlier using knowledge of your opponent to spot "better" choices.
    Well, there was no knowledge at all, because I saw them for the first time ever.

    As you say it now, I realize that one could have taken it that way. My bad; that's not what I meant. It was not meant as a personal criticism but rather a general recommendation to everyone, "moving forward". Because you are correct, situations may be random but people aren't; they have patterns. The patterns aren't precise, of course, but they are definitely biased.
    Musicat wrote: »
    Helsa wrote: »
    BTW I am curious, what would be an example of "Statistically approved "best practices" often mean higher overall frequency of successful cases, but not higher chances in each case."
    No-no-no, they're basically called so because of being statistically approved. That's what I meant. If they weren't statistically approved, people wouldn't call them "best practices". I'm highly likely to be mistaken, but I'd call Lileas' suggestion to start from W a "best practice".

    Okay, so the example of the Monty Hall problem is such a case because there is a 2/3 chance of success in switching, with the caveat that it is a chance, not a guarantee, but a best practise nonetheless.

    I must confess, I've been studying the jousting minigame since late 2012, and you just taught me the second thing-that-I-should-have-already-known-but-didn't of this year. The first was the realization that the jousting NPC's at the jousting stadium only joust in one particular lane (the game calls them arenas) and no other: Adair only in one, Gornevez only in two, Melor only in three, Lota only in four. I'm embarrassed to say that I never noticed that until this year. The second was, as you say, Lileas recommends starting with W; I hadn't realized she gave tactical advise. Mathematically speaking, with no knowledge of your opponent, that is the best start. I have a video on that, explaining the two reasons why, naturally.
    Musicat wrote: »
    Helsa wrote: »
    To be fair, one can actually be in a "sufficient luck" situation even after the threshold is reached. If the player in the winning position is not aware that they are or how to execute the winning strategy. In this situation the player in peril "would be lucky" that the other player didn't know to or how to press home their attack. Have i got that right?
    Frankly speaking, that basically covers my whole theory. If you're lucky enough, your opponent acts in a way beneficial for you (intentionally or carelessly).
    I'm always imagining a thing like:
    You: Hmm, should I start from W?..
    Opponent: He may be starting from W, I should probably use A.
    You: But what if he thinks I'm starting from W and plays A? I should maybe start from A or S to negate that...
    Opponent: He may think that I'll try to hit his opening W with my A... and play another one... I should probably go S or D then...
    You: But he also can use D to negate my S (theoretically countering his A, which is theoretically countering my W). Maybe I should just risk and start with D to have better options further.
    Opponent: OK, no more time to think. I'll sacrifice my D for his theoretical W.
    DxD
    Opponent: Wow
    You: Oh...

    This thinking line can go on forever, and it's pure luck in terms of which option they eventually choose (or suppose you to choose) (RAWR) :D
    P. S. Of course I realize that few people would have such thinking line.

    When I first started playing, I had that very conversation with my self in my mind. I ended up on W, without, at the time, knowing W's special properties. Of course being a Giant, I might've been biased to start with W as well.

    W is especially noteworthy for the degree of counterfactual thinking it requires. Pick any opening where at least one jouster picks W. Now, start thinking about pass two. Nearly as involved, right? This is why I say moving to W can be done as a sacrifice. You move to W, potentially sacrificing some health for a chance at a reversal of fortune.

    Positions with W are not like a double S position; ah those halcyon days of simple double-S positions. Full disclosure, there is one situation where one won't necessarily move to A from SS, in a knee-jerk fashion: if a giant faces a human, and it's early in the match, say an SS opening, the giant may want to, instead, consider crossing to W and probably take that A-hit, in the next pass. A clean A-hit is the same strength as a combo bonus. If the match is going to be SS AA WW, then the giant is still down by the same amount but they have a lot less personal health to work with. The giant could opt for SS AA SW (provided the human doesn't cross to D in that third pass). In this situation the giant is still down in damage but by only half the amount. The downside is that being at W in a WS-split is less perilous than being at S: human 70% damage at W, giant 80% damage at S. It's not game over, for the giant, but they're in trouble.

    There is another very well known "best practise" in jousting and I'd be surprised if you were not aware of it and use it all the time. That is the one where you choose, for your next pass, the attack that your opponent is defeating in the present pass. This tactic arises from the rule that attacks cannot be repeated two passes in a row. In 9 of the 16 possible positional situations that one can react to, the best practise is to ensure that you do not outright lose the next pass REGARDLESS of what your opponent does. Since ones choice is set in the rule governing the best practise and since the goal is sought regardless of your opponents actions, the goal of the action is achieved in the absence of luck. Of course there is luck in the degree that you succeed, but ensuring no outright failure can be achieved in the absence of luck in these 9 of 16 situations.
    Musicat
  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 6, 2019
    @Helsa , I love your thoughts about intimidation and stuff, and concerning those AI matters mentioned. It's true about mental vulnerability of humans in comparison with AI, and you're right about the "big shark factor" (a master's presence's influence). That's what may badly affect participants in any game and what I avoid feeling everytime (yesss, avoid feeling... what?).
    Yesss, that was Kasparov.

    I disagree with the second one, because, againagain, it depends. On one hand, I believe (or am sure, or used to know; doesn't matter actually) that people can be random. On the other hand, we might as well call such random behavior a pattern too! Like, when a person is trying to be random intentionally, maybe even forcing themselves to be so, they also follow such a pattern. Moreover, their actions tend to be more predictable than those of a naturally random person. There's, certainly, few opportunity to be truly random when having only 2-4 options, like in Jousting.

    In fact, I've never fought against any other champion than Adair. Knowing what you said about the arenas, the reason must be pretty obvious: when I come to joust, there are usually no other players (especially on former Mari server), and if there's one, we fight against each other. But I do remember seeing someone fight the elf lady champion and wondering then how they've got her as their match.

    And I didn't quite understand the last paragraph. Do you suggest switching to the option that was just used out by an opponent (in case it beat our choice)? I find it a good idea, but it's also not too safe. I have won some 2 times using that strategy, and also have lost 2 times, so there's nothing in my personal experience talking in favor of or against this.

    By the way, there was a situation recently with two elves (A and B ) using all the same moves (WW-SS-AA-SS-AA), but of course there can only be one winner, so B totally lost at the last move, getting full damage and not dealing any to A. It might be just the case where some knowledge could be applied. However, I may be recollecting a situation when there was a draw, but I'm not sure.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Musicat wrote: »
    @Helsa , I love your thoughts about intimidation and stuff, and concerning those AI matters mentioned. It's true about mental vulnerability of humans in comparison with AI, and you're right about the "big shark factor" (a master's presence's influence). That's what may badly affect participants in any game and what I avoid feeling everytime (yesss, avoid feeling... what?).
    Yesss, that was Kasparov.

    I think the prime example of this in military history is Darius III at Gaugamela. Alexander had no business doing as well as he did if not for Darius being completely intimidated by him.
    Musicat wrote: »
    I disagree with the second one, because, againagain, it depends. On one hand, I believe (or am sure, or used to know; doesn't matter actually) that people can be random. On the other hand, we might as well call such random behavior a pattern too! Like, when a person is trying to be random intentionally, maybe even forcing themselves to be so, they also follow such a pattern. Moreover, their actions tend to be more predictable than those of a naturally random person. There's, certainly, few opportunity to be truly random when having only 2-4 options, like in Jousting.

    I'm not sure what you mean by the second one; was it the situations maybe random but people aren't thing? Check out the New York Times article whose link I posted earlier in the thread. Whereas, people can choose to move randomly in order to hide patterns, I have catalogued thousands of matches on Alexina, and the summaries I've made do show clear biases. To be sure, say a jouster from WW has moved to D 55% of the time, A 35% of the time, and S 10% of the time, these numbers would suggest to try A. Of course THIS time MIGHT be one of those 10% of the times.
    Musicat wrote: »
    In fact, I've never fought against any other champion than Adair. Knowing what you said about the arenas, the reason must be pretty obvious: when I come to joust, there are usually no other players (especially on former Mari server), and if there's one, we fight against each other. But I do remember seeing someone fight the elf lady champion and wondering then how they've got her as their match.

    Lota, is seen the least since she is in Arena 4. If she showed up it was either because there were 7 people in the round and number 7 got her, or there were an odd number more where arena 4 became available just in time for that last oddly numbered person. BTW you can go against them in practise but it'll cost you 100 gold and it's still a random selection but maybe that was it.
    Musicat wrote: »
    And I didn't quite understand the last paragraph. Do you suggest switching to the option that was just used out by an opponent (in case it beat our choice)? I find it a good idea, but it's also not too safe. I have won some 2 times using that strategy, and also have lost 2 times, so there's nothing in my personal experience talking in favor of or against this.

    I've heard it called the Clockwise Rule. Let me explain in more detail. Say your opponent is at S and you are at, say, W. Right now you are attacking D and they are attacking A. Since neither of you can repeat your move, they cannot use S again next turn, therefore they cannot directly attack A again next turn. So, if next turn, you choose A, the worst they can do to you is to choose A them self and tie you, they cannot move anywhere to outright, onesidedly (if that's a word) win the next pass. See what I mean?
    Musicat wrote: »
    By the way, there was a situation recently with two elves (A and B ) using all the same moves (WW-SS-AA-SS-AA), but of course there can only be one winner, so B totally lost at the last move, getting full damage and not dealing any to A. It might be just the case where some knowledge could be applied. However, I may be recollecting a situation when there was a draw, but I'm not sure.

    That's a tie-breaker. Yes, the highlander rule applies "There can be only one". When both jousters have simultaneously inflicted, otherwise lethal blows to each other the RNG picks a winner and only the loser has damage applied. In my match write-ups I say that I interpret the matches. Since some passes are ambiguous, I have to guess unless someone tells me what the last pass was.. Final passes are notoriously ambiguous because of tie-breakers. Anyway, in these situations, the game declares a winner with a coin toss. Sustained, inflicted, and potential beyond 100% damage, and jouster history have no baring whatsoever on the outcome of this; let's call it a coin toss. As an aside, sometimes "Best practises" involve seeking one of these.

  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 7, 2019
    Helsa wrote: »
    Whereas, people can choose to move randomly in order to hide patterns,
    People can act randomly without choosing to do so. That means they're naturally random. But, as I've said, this might just be called a pattern too. And, also, as I've said, there's few opportunity to really be random in Jousting.
    Helsa wrote: »
    Lota, is seen the least since she is in Arena 4. If she showed up it was either because there were 7 people in the round and number 7 got her, or there were an odd number more where arena 4 became available just in time for that last oddly numbered person. BTW you can go against them in practise but it'll cost you 100 gold and it's still a random selection but maybe that was it.
    Probably was a random thing, because there were 3 players only.
    Helsa wrote: »
    I've heard it called the Clockwise Rule. Let me explain in more detail. Say your opponent is at S and you are at, say, W. Right now you are attacking D and they are attacking A. Since neither of you can repeat your move, they cannot use S again next turn, therefore they cannot directly attack A again next turn. So, if next turn, you choose A, the worst they can do to you is to choose A them self and tie you, they cannot move anywhere to outright, onesidedly (if that's a word) win the next pass. See what I mean?
    That's just common sense. Common sense is not a strategy.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    edited July 7, 2019
    Musicat wrote: »
    Helsa wrote: »
    Whereas, people can choose to move randomly in order to hide patterns,
    People can act randomly without choosing to do so. That means they're naturally random. But, as I've said, this might just be called a pattern too. And, also, as I've said, there's few opportunity to really be random in Jousting.

    Sure, let's say "Choose to Joust Wrongly" (if that's a word). You can, though, choose to joust as randomly as the games RNG, by letting the computer play for you; I do that. The NPC's and by extension absent PC's do move randomly, not even random with a bias; I've got a video on that too, because, of course, I do.
    Musicat wrote: »
    Helsa wrote: »
    Lota, is seen the least since she is in Arena 4. If she showed up it was either because there were 7 people in the round and number 7 got her, or there were an odd number more where arena 4 became available just in time for that last oddly numbered person. BTW you can go against them in practise but it'll cost you 100 gold and it's still a random selection but maybe that was it.
    Probably was a random thing, because there were 3 players only.

    It'd be nice if there was a video of it to help get to the bottom of it. Maybe it was a group of folks paying to practise with other than Adair?
    Musicat wrote: »
    Helsa wrote: »
    I've heard it called the Clockwise Rule. Let me explain in more detail. Say your opponent is at S and you are at, say, W. Right now you are attacking D and they are attacking A. Since neither of you can repeat your move, they cannot use S again next turn, therefore they cannot directly attack A again next turn. So, if next turn, you choose A, the worst they can do to you is to choose A them self and tie you, they cannot move anywhere to outright, onesidedly (if that's a word) win the next pass. See what I mean?
    That's just common sense. Common sense is not a strategy.

    The tactic is based on a game rule: you can't repeat an attack twice in a row. It is a direct consequence of it. Again, though it is void of luck only in 9 of 16 situations. Three of the remaining seven, you can be attacked cleanly but at least, for those three situations, you can forcibly minimize the damage you might take. For the remaining four, luck and psychology hold sway.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    2019 July 7

    Nao had finals again with 13-1 participants. You can find all the action here.

    Alexina only had 7 so one short this week.
    Kensamaofmari
  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 8, 2019
    Helsa wrote: »
    Sure, let's say "Choose to Joust Wrongly" (if that's a word).
    Well, if that's for jousting without really thinking, then yes. And making decisions in accordance with your gut instinct can also be called random of sorts.
    Helsa wrote: »
    It'd be nice if there was a video of it to help get to the bottom of it. Maybe it was a group of folks paying to practise with other than Adair?
    No idea. Since we've assumed that the AI opponents are random, it doesn't really matter if they were intentionally summoning her (pretending to, that is) or got her by chance. I can't see any other options to challenge Nonadairs than taking a practice round.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    You did pretty good in your match I thought; just bad luck! <winkie woo>
  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 9, 2019
    Yesss, by the way, I started well enough, and then wanted to switch to D but went W instead (after I thought that me switching to D would be too predictable for my opponent who'd certainly counter it), but it turned out that it wasn't so obvious for the opponent, and they chose one of your "best practices" (going to S after being defeated by it). This way, the thing being a "best practice" was affirmed once again while actually having been under the greatest risk of being taken down.
    Come to think of it, in terms of Jousting, countering best practices is the bestest practice, because, by doing this, you counter the majority of jousters.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    edited July 9, 2019
    Musicat wrote: »
    Yesss, by the way, I started well enough, and then wanted to switch to D but went W instead (after I thought that me switching to D would be too predictable for my opponent who'd certainly counter it), but it turned out that it wasn't so obvious for the opponent, and they chose one of your "best practices" (going to S after being defeated by it). This way, the thing being a "best practice" was affirmed once again while actually having been under the greatest risk of being taken down.
    Come to think of it, in terms of Jousting, countering best practices is the bestest practice, because, by doing this, you counter the majority of jousters.

    It is when folks are wont to use them. Some veterans are very cautious, so against them your suggestion has a good chance to succeed; I've done that very thing in the past when I still actively competed. Just think of jousting as being like quantum physics, you're trying to manipulate probability wave functions. :)
    Musicat
  • MusicatMusicat
    Mabinogi Rep: 1,755
    Posts: 180
    Member
    edited July 10, 2019
    @Helsa , Jousters in superposition? (>w<')
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Musicat wrote: »
    @Helsa , Jousters in superposition? (>w<')

    When both choose the same attacks for the match and go around together in a clockwise fashion, (such as DD SS AA WW SS) or when they choose opposite attacks and go counterclockwise around, (such as SW DA WS AD SW) I suppose they'd be entangled.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Nao had its third final in so many weeks with 12-1 jousters poking the heck out of each other. You can see all the action here. Alexina only had 7, so no finals there.
  • KensamaofmariKensamaofmari
    Mabinogi Rep: 32,720
    Posts: 7,471
    Member
    Good to see participation continue. I'll be getting a new computer soon, so I will return to the competition as well.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Good to see participation continue. I'll be getting a new computer soon, so I will return to the competition as well.

    Cool hope to see you back at the weekly Naowegian cluster-poke.

    See uh when you joust um like you hold a lance right and um you try knock the other guy off by poking them with the lance right?
  • KensamaofmariKensamaofmari
    Mabinogi Rep: 32,720
    Posts: 7,471
    Member
    Helsa wrote: »
    Good to see participation continue. I'll be getting a new computer soon, so I will return to the competition as well.

    Cool hope to see you back at the weekly Naowegian cluster-poke.

    See uh when you joust um like you hold a lance right and um you try knock the other guy off by poking them with the lance right?

    Off the horse of course.
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    Helsa wrote: »
    Good to see participation continue. I'll be getting a new computer soon, so I will return to the competition as well.

    Cool hope to see you back at the weekly Naowegian cluster-poke.

    See uh when you joust um like you hold a lance right and um you try knock the other guy off by poking them with the lance right?

    Off the horse of course.

    Of course, unless it's Mr. Ed!
  • HelsaHelsa
    Mabinogi Rep: 14,970
    Posts: 3,245
    Member
    No finals this week: Nao had 6 and Alx had 7. Pooper doopers. :(
  • KensamaofmariKensamaofmari
    Mabinogi Rep: 32,720
    Posts: 7,471
    Member
    Helsa wrote: »
    No finals this week: Nao had 6 and Alx had 7. Pooper doopers. :(

    Oof.