Quantifying Skill in AW

Skrivet av VRIL, 21.02.2014 at 08:31

But tbh ELO is the best indicator for player skill at the moment. Mostly its accurate and there are only a few exceptions of players who farmed.

Thing is not everyone duels actively, and not everyone duels similar ranks.

Skrivet av The Tactician, 21.02.2014 at 09:27

similar ranks.

ranks dont matter, but for the ones who duel actively its pretty accurate.

Ok but a fair amoount of people dont duel actively... Also this thread is not about measuring skill, but defining it.

This thread is about defining skill in AW. Even though dueling is an insignificant part of AW (far less than 1% of active AW accounts have an Elo of 1100 or more), Elo purports to measure 'skill' in dueling. Given two Elos, and their weights (which I cannot find documented anywhere), in chess, each rated player would know before a match the liklihood within a range, of an expected win/loss/draw.

(as of my last check)
The #1 dueler, by Elo, hasn't lost any of their past 100 duels, earned <1 Elo per duel, and their average opponent had an Elo of <1000. This dueler's Elo was 1596.
The #8 dueler, by Elo, lost an average of .37 Elo per duel, and their average opponent had an Elo of 1123. This dueler's Elo was 1427. This dueler won more duels than he lost, by far, but the duels he lost cost him dearly in Elo.

So Elo measures relative skill in dueling AND, it could be said, the wisdom in choosing which duels to fight.
---
The question is: How much does the ability to beat one player in a duel apply to 'skill' in AW?
One would surmise by the shockingly low-level of participation, not much.[b][/b]

Skrivet av Goblin, 21.02.2014 at 16:45

Skrivet av SuperiorCacaocow, 21.02.2014 at 16:29

It wasn't done easily, he needed more than a year for it. Now that if becomes more difficult as he progresses, try to imagine when he could possibly break the 1700 mark at that rate.
I agree with the rest you wrote though. Unfortunately atWar hasn't enough players to get ELO working at this point, and I don't see this happening for a long time, if ever.

Yes he needed a long time ...but he did stop farming from time to time when he was at first place and he isnt trying to do it every day ...look at it this way. With even such low elo he is geting he still got very far with "only" 300 duels.

But seriously the worst thing about all this ...is that blucher is being rude to all of us. He knows that players are pissed about what he is doing, doesnt respond to challenges to a duel from others or makes some kind of excuse ...i asked him for a ancient world map duel (his teritory ...i played it only once, have no idea about units there or countries), he says "sorry i just made a game" ...??? ...then a low rank joins his game and he sends him duel request.

EDIT: ok amazed that his last duel was rank 9, but that dude is a scenario player so ...how tough can he be (sorry scenario players )

The best way to evaluate skill would be to create an algorithm that accounts for upgrades and players' expansion over income ratio.

"Experience" being left out of the formula for obvious reasons; for instance, legitimacy; but also accountability, we cannot incorporate experience along side "mechanical skills" such as expansion.

Now, the formula I proposed is obviously quite simplistic and doesn't account for all the aspects involved in being "skillful"; such as one's experience, positional expansion (not just maximum city-expansion) and of course, tactical mechanics.

Therefore, this formula would have to be done, and then, with a little bit of logical input, we could add our general views of a player's experience, tactical mechanics and so forth.

But in the end, why the need for this? lol. I think observing a player's gameplay is enough to dictate his skill on a general basis; no real need for specification.

I have come to a conclusion: OP is a troll.

Skills is a subjective quality.

Skills is in AW is not like a skill tree you can unlock or rank up, it is something you develop that can not be measured or quantified.

You can say the person is pretty "pro" or "nooby", but that just mean you are comparing them to other players.

It is like the color green. I see green, and you see green, but how do we know both see the same green? What looks green to you, could be my red, but we both grew up calling that specified color we both see, "green".

Just like in real life, skill is measure as simply this: "can you do it, or not?" ["can you do it flawless, or not"]

This is all that is of skills.

Statistics can be an indicator of skills, yet once again, you are putting humans down on paper, and quantifying their behavior.

So to conclude, every man home is his castle, but in this case, home is what you like to play.

Skrivet av zombieyeti, 20.02.2014 at 23:21

...

I agree with Cthulhu in that skill will be viewed differently by different players.

With regard to your original post, I think maximizing SP is only one of many ways to measure skill in AW. For those that play competitive games, skill could be defined by your ability to beat the other player in a specific context (EU+ 1v1 or 3v3). SP is more or less relevant in this case. For example, in a 3v3 context, you can sacrifice yourself to make your team win. Clan proficiency and Elo are probably better indicators of sckill for those playing competitively, but both these metrics have their problems and can be exploited. It would be interesting to develop a more robust method to measure ability in competitive play. By the way, I think this is the whole idea behind tournaments (perhaps more testing than measuring though).

That said, AW is more than just the EU+ competitive scene. It is many games in one. There are the scenarios, UN/roleplay, games with alliances and also casual games. All of these require different skills to play well. I think there is no single, effective and universal way to measure skill for all these playing styles.

Skill=the ability, coming from one's knowledge, practice, aptitude, etc., to do something well.

Sorry, gone for a bit. Very busy.
---
I want to determine what the community consensus is on what skill is ... then see if that thing/those things can be measured.
I haven't seen any recent serious attempts at gaining deep understanding of the game, perhaps people just want to (by and large) just click away and not analyze, or they have secrets and want to keep these secrets under wraps. If eitheror case is true, this may explain why such an incredible game is a mostly-undiscovered gem.

I can't even get straight answers on basic battle mechanics.

AFA AW's 'complexity' -- AW is more complex (and chaotic) than Chess or poker. I think Chess (and most forms of Poker) have tremendous complexity.
You may not agree.

Skrivet av Goblin, 21.02.2014 at 18:50

Yeah ...wisdom in chosing duels ... ...wisdom ...like anyone couldnt done the same thing #1 dueler did.

btw. zombi ...what is your story bro? Whats with this obsession about what skill is? It's a online game and not really that complicated game...

Skill = expirience + brains : luck

Skrivet av zombieyeti, 06.03.2014 at 21:01

Skrivet av Goblin, 21.02.2014 at 18:50

wow the troll is stronk in you

p.s. why you quote upside down?

Really not trolling ... really *really* trying to understand battle mechanics etc.

Skrivet av Khal.eesi, 06.03.2014 at 21:10

Skrivet av zombieyeti, 06.03.2014 at 21:01

Skrivet av Goblin, 21.02.2014 at 18:50

wow the troll is stronk in you

p.s. why you quote upside down?

... and I shall quote from the bottom (sorry)

Skrivet av Goblin, 07.03.2014 at 04:10

AW is more complex then Chess? o.O ...318,979,564,000 possible ways to play the first four moves of chess and there are
169,518,829,100,544,000,000,000,000,00... ways to play the first ten moves of chess.

Yea i can do milion openings with turkey xD ...wtf mate.

well technically, since youve 72 reins and a choice of over a dozen countries to disperse those 72 units in different quantities over theres actually significantly more opening moves for turkey than a game of chess XD

Skrivet av zombieyeti, 06.03.2014 at 21:01

I want to determine what the community consensus is on what skill is ... then see if that thing/those things can be measured.

Definitely "those". If you do find consensus in the forums - I would be surprised - it will still represent only a fraction of the community. Many players communicate mostly in game chat. Besides, it's undeniable that AW is many different games. You would have to develop a way of measuring skill for each game type. For competitive (1v1) play, the best way in my opinion is tournament wins.

Skrivet av zombieyeti, 06.03.2014 at 21:01

I haven't seen any recent serious attempts at gaining deep understanding of the game, perhaps people just want to (by and large) just click away and not analyze, or they have secrets and want to keep these secrets under wraps. If eitheror case is true, this may explain why such an incredible game is a mostly-undiscovered gem.

I can't even get straight answers on basic battle mechanics.

By "clicking away", I think you mean playing the game... seems you're implying there's something wrong with just playing the game without analyzing it I personally have an interest in game mechanics but I don't have any problems with others not sharing this interest.

IMO, one of the reasons game mechanics aren't well understood is that, while there have been various attempts at exploration, there is some opacity at the level of the game creators and the material made available to us. When we do have material to go with, it's hard to tell if it is current our outdated... Without help from the admins, understanding will progress slowly as it will have to be based on empirical evidence. I don't think it should be necessary to resort to experimentation to understand game mechanics and find it completely understandable that players wouldn't have the patience to conduct these experiments.

Skrivet av Grimm, 07.03.2014 at 12:21

Skrivet av zombieyeti, 06.03.2014 at 21:01

I want to determine what the community consensus is on what skill is ... then see if that thing/those things can be measured.

Definitely "those". If you do find consensus in the forums - I would be surprised - it will still represent only a fraction of the community. Many players communicate mostly in game chat. Besides, it's undeniable that AW is many different games. You would have to develop a way of measuring skill for each game type. For competitive (1v1) play, the best way in my opinion is tournament wins.

Skrivet av zombieyeti, 06.03.2014 at 21:01

I haven't seen any recent serious attempts at gaining deep understanding of the game, perhaps people just want to (by and large) just click away and not analyze, or they have secrets and want to keep these secrets under wraps. If eitheror case is true, this may explain why such an incredible game is a mostly-undiscovered gem.

I can't even get straight answers on basic battle mechanics.

By "clicking away", I think you mean playing the game... seems you're implying there's something wrong with just playing the game without analyzing it I personally have an interest in game mechanics but I don't have any problems with others not sharing this interest.

IMO, one of the reasons game mechanics aren't well understood is that, while there have been various attempts at exploration, there is some opacity at the level of the game creators and the material made available to us. When we do have material to go with, it's hard to tell if it is current our outdated... Without help from the admins, understanding will progress slowly as it will have to be based on empirical evidence. I don't think it should be necessary to resort to experimentation to understand game mechanics and find it completely understandable that players wouldn't have the patience to conduct these experiments.

Your considered response is much appreciated. I thank you.
- I absolutely see, and agree, how different 'game modes' would call for a different set of tools and skills. Should this topic be broached again, the starting point would be to specify the game type.


- Re: Opacity and exploration (and maybe why the pursuits are futile): I am disappointed but (as of now) not surprised about the lack of understanding. The experiments I have participated in seem to indicate that either some of the 'rules' have changed, or were not well understood to begin with. I can understand that the average player should not be expected to run the experiments (and not that you said this) but I think that most players are 'interested' in the game mechanics to a greater degree than the information is available. Your comment speaks to my larger pursuit, of course: Understanding Game Mechanics.

Shameful analogy: In the Game of 'Life', most 'players' don't run the experiments, but most 'players' have a stake in the outcomes.
- The rules are well understood but the outcomes are chaotic (f.e. law, economics, poker).
- Professionals run the experiments and feed us knowledge (physics, chemistry, computer science).
- The 'game mode' has been explored to exhaustion and everyone gets that the objective and mechanics are highly personal (spirituality, happiness, religion).

When Newton, then Einstein, then Schrödinger gained progressively better understanding of the universe, the underlying rules of the universe didn't CHANGE, just our understanding of them (I hope). If the AW 'Universal Constants' change without documentation the difficulty of gaining meaning-through-expermentation is ... non-trivial.

That being said the more popular games have legions of players who conduct the experiments and discuss the results where the information isn't documented, and these games have a much-less-'wargame' bent.