Thingol wrote: Don't know what you're talking about Min, Chess is entirely a game of skill. There is no good or bad luck...only good or bad moves. Unlike Cram, I was upon a time a very advanced chess player, but I still find his analogies pertinent.

I think your reference to Fischer's blunder being "good luck" is a broadening of the term. I believe most people that complain about bad luck here (like me) are referring to bad dice and/or bad placement, both of which are not decision-making for you or your opponent(s). When I make a bad decision, I have only myself to blame. When I roll 14-44 and just miss eliminating a player, then said player cashes a set and goes on an elimination spree to win the game, I call it...horseshit...and bad luck.

+1 I'm not a good chess player Min, but based on the argument it seems you're confusing strategy and daring with luck.   The main reason I've never played much chess is that there is NOT any luck involved and a computer that can out compute me will kick my ass every time.   When I played Battle Chess as a kid I wished just once that an attack would fail so the computer couldn't project any further ahead than I can. 

Mad Bomber wrote: I see the top players take 10 singles in a row with a loss of maybe one or two.....I seem to lose one or two against each single...... Multiply that frustration times 17 turns in a row....4 of which should have translated to wins...........these days I only seem to win when others have an epic fail...(thingol's example)....I have no skill.......oh yeah love trying to take a single over only to roll 5 ties in a row......anyone notice all the ties in a row?

Was that satirical? Or, are you on a crazy bad luck streak and should I join all your open games?  =D

Minimal wrote:

Okay but my OP asks the question of whether my winning streak was because of luck.

You'll likely need a computer algorithm that gives the NCAA's BCS rankings a run for their money, but I suppose there's no reason it's not possible.

Everyone's skill level should give them some baseline expected win% against an "average player." Then, the likelihood of winning against the "average player" will be spread over a curve peaking at that point, tailed at either end by the effects of dice, starting placement, number of players, seat, and a bunch of other less tangible factors.

You can then take those curves for various players and overlay them on each other. Using some calculus mumbo-jumbo, you can calculate the overlapping areas under the curves and use that to determine, with some pre-determined amount of acceptable confidence the likelihood that one player should beat the other.

If you want to get more precise, you could try to model your skill curves as the likelihood of winning against players similar to the ones in an upcoming match rather than against an "average player".

As far as I know, there is a bustling market for people skilled in this sort of analysis in the fantasy sports, broadcasting, and gambling industries (among other less glamorous areas, like insurance). If someone were to find themselves as having a knack for making accurate performance curves, I think they could do pretty well for themselves out there.

The years between when I did the sort of math explained above are getting pretty long, so I apologize in advance for what a mess of it I probably made.

"When I make a bad decision, I have only myself to blame."
And when your opponent makes a bad decision?

To those that make the argument that everything in chess is deterministic I should remind you that the physics of rolling dice are also deterministic yet this act produces results that are random. You imply my definition is *too* broad. What exactly is broad about it?

Everything that has an influence on the result of your game can be categorized in one of two ways.

A) within your control and a direct result of your actions (i.e. skill)
B) outside your control and not at all a result of your actions (i.e. luck)

Now, of course, there may be some grey area in between. My two categories aren't all-encompassing but if you honestly believe that your opponent making a 1 in 200 game blunder against you is not luck than I think I have no further comments on the matter.

VERY cool Squint. I'll look at all of those stats when I'm not so tired.

 if you honestly believe that your opponent making a 1 in 200 game blunder against you is not luck than I think I have no further comments on the matter.

While you could indeed consider yourself "lucky" that your opponent made a 1 in 200 blunder, the fact remains that the event was under the control of one of the players, and to the extent that your opponent's blunder was the "unlucky" result of them happening to have to flu or being in the midst of a break-up, I would be tempted to concede that there is no such thing as determinism in any game.

So as a response to your question regarding what 'broad' is, I would offer that we tend to more 'narrowly' define luck on this site in terms of the pure mechanics of the game. The continuum being that either the actions of the players determine the outcome, or the actions of the mechanics of the game as designed and then executed by the engine determine the outcome regardless of what players do.  You may find this too narrow, but it can be very clearly delineated and then hopefully more easily quantified.

Very interesting discussion all around & good points, but I think this has come down to an argument about semantics. Basically we all agree that things within your own control are skill, not luck.  The disagreement is whether things under your opponents control are categorized as skill (not your skill, your opponents skill) or luck (because they are not under your control).

As M57 says, I think in past discussions on the site, "luck" has basically meant anything that the wargear engine pulls a pseudo-random # for.  A clear subset is 'dice luck' which relates to what most people would first think of when talking about their luck and what the various existing luck stats/charts cover.

Ozyman wrote:

has come down to

Has started with

A) within your control and a direct result of your actions (i.e. skill)
B) outside your control and not at all a result of your actions (i.e. luck)

Now, of course, there may be some grey area in between. My two categories aren't all-encompassing but if you honestly believe that your opponent making a 1 in 200 game blunder against you is not luck than I think I have no further comments on the matter.

 Luck by definition is an event which occurs beyond one's control.  While your opponent's blunder is generally beyond your control (unless you took steps to cause that blunder), it could be considered lucky, but since the blunder was not beyond his control it was also not by definition unlucky.  But, while you might get "lucky" in a single game chess is still not a game of luck because every outcome is calculable and controllable. 

But, this whole thread while entertaining is moot anyway. I'm surprised no one has made a poker analogy. But, the only way to really know is just keep playing.  In the long run, luck will always level out.  It's why you always see the same faces in the WSOP.

I will always have further comments on a matter. =P

A noob in one of my games on this site just handed the win to the other guy in our game.  Whatever your thoughts on the definition of luck, I don't take any extra comfort knowing that it was human stupidity and not an algorithm that screwed me.

Minimal wrote:

A noob in one of my games on this site just handed the win to the other guy in our game.  Whatever your thoughts on the definition of luck, I don't take any extra comfort knowing that it was human stupidity and not an algorithm that screwed me.

That's never fun and it happens to us all.  I tend to think of it as noise more-so than a luck element.  I would think that in theory it should cancel itself out because it is more random than not.

Again, I bet the best players have a much higher noob-gave-it-to-the-other-guy to win ratio then all the rest of us.  Same as dice luck, or starting position luck, or other guy mistake luck. 

Luck is part of the game, and it can't be totally circumvented, but some of the players here manage to do just that about 90-95% of the time... and I think that's the skill level we should all be striving for. 

I was really interested in how far up the rankings I can climb but it's taking so long that it just isn't going to happen. Why on earth does it take like a solid hundred games to get anywhere in the ladder?

Cause some guys play 10 - 20 games a day...

Depends on which ranking you are talking about.  With a real quick check I found that of the top 100 globally ranked players (those with over 1885) over 30 of them got there with less than 100 games and 4 in less than 50 games.  There actually are more but I stopped counting when I got to 30+

Minimal wrote: I was really interested in how far up the rankings I can climb but it's taking so long that it just isn't going to happen. Why on earth does it take like a solid hundred games to get anywhere in the ladder?

I noticed you play a lot of 1v1, which is fine while you're climbing from 1000 - 1800, but, at some point the cost of losing a game will become too great for the number you'll need to win to offset a loss.  At that time you'll have to devise a plan on how to win greater than your share of large multi-player games whereby each win can deliver you 80-150 points and then try to go at least 35-50% win rate for such large multi-player games so your big wins will out pace your ever growing losses.  Of course it requires a different style of play then 1v1's, so it's not everyone's thing, but to accomplish the goal that you wrote above, that is what it'll take.

@Minimal, Your H-Rating is practically instantaneous.

As I close in on my 300th win I'm becoming less and less interested in my rankings, partly because I pretty much know what my range is by now, but more because I believe  the range of board types on WarGear is forcing most players who are concerned about their global rankings to become specialists.  And because of the shear number of offerings those concerned about their Championship points have to think about which boards suit their talents, and/or which ones are a waste of their time.

Also as berickf noted, it's not just about which games you play - but also how many players are in the game.

As the designer toolbox becomes even more powerful, designers are more easily able to make games that are very un-Risklike.  I'm here to play - not worry about my rankings, and I love a good dueling map.  Heck, I've designed a few of 'em.  At the risk of re-igniting the original discussion (we are OT here), I prefer boards that are more deterministically oriented, If I play 1v1 boards like Hex or Go-Geared and run up against Hugh or Andernut, I know I'm going to get my head handed to me 9 times out of 10, yet that doesn't stop me from playing them.

Minimal wrote:

A noob in one of my games on this site just handed the win to the other guy in our game.  Whatever your thoughts on the definition of luck, I don't take any extra comfort knowing that it was human stupidity and not an algorithm that screwed me.

I´m with you on this one. =)

berickf wrote:

Minimal wrote: I was really interested in how far up the rankings I can climb but it's taking so long that it just isn't going to happen. Why on earth does it take like a solid hundred games to get anywhere in the ladder?

I noticed you play a lot of 1v1, which is fine while you're climbing from 1000 - 1800, but, at some point the cost of losing a game will become too great for the number you'll need to win to offset a loss.  At that time you'll have to devise a plan on how to win greater than your share of large multi-player games whereby each win can deliver you 80-150 points and then try to go at least 35-50% win rate for such large multi-player games so your big wins will out pace your ever growing losses.  Of course it requires a different style of play then 1v1's, so it's not everyone's thing, but to accomplish the goal that you wrote above, that is what it'll take.

One of my favorite maps is 1v1, but at a certain point I realized that I needed two wins against decently ranked player to break even for one loss to a noob.  

I also decided around then that I wanted to play more variety and build my CP instead of just playing the maps I was best at. My CP jumped way up and my GR dropped down and hasn't changed too much since. 

Because of time I also dropped from 70+ games at a time to about 20.  Of course now when I'm jonesing to play something , but I have no turns waiting I hang out in the forums looking to talk about them.

The rating system used here seems very flawed. Systems like Glicko and ELO change the point adjustments considerably depending on each player's rating. Under ELO an expert playing a novice can expect +1-2 for a win and something like -20 (or much more) for a loss. On this site even with the largest rating disparity going into the game the adjustment is only about half. For this to be accurate a 1000 should have about a 25% chance to beat a 2000 which I find extremely hard to believe. I think the smartest strategy to climb in rating on this site is just to play only noobs for this reason.

@Recon They must have played really large games. In fact it might just be "lottery" winners.

1000 does not mean that a player is necessarily a "noob", only in-as-much as they are new to the site.  They can be expert risk players, however.  It's just the starting point in the ranking system.  So, to put such a score in context one would need to know how many games are built into settling on that ranking. 

So, when you say you should choose to play only noobs, it makes me think you are actually saying you should choose to play "poor" players, those with 1000 or less points after significant game-play to settle on their ranking.  As that would be the only way to know for sure that they could possibly be an "easy" win? 

Personally I would never seek to play repeated games against low ranking players unless it was on such a specialized scenario like 1v1 Blind Ants whereby I knew I would win 99.9% of games against such players... But how boring would that get?  Playing the noob standard of Wargear Warfare, however, is not such a safe bet.  You get your score up to 2500, say, and then are picking up only 6-8! points a game for every win against 800-1000 point players.  Then, in one game, the other player gets Australia defended by a neutral blockade in Asia right off the mark and you have no chance versus all his ineptitude at playing in general, and he picks you off for 63 points...  Only 9 more wins to make up that setback! Then, 4 games along on your road to recovery your dice completely fail you and.. oh, only 6+9=15 more wins to get back to your height.  Only, by now, the noobs are all running scared from you so you make your own game and some player with 1250 and very few games played jumps in, darn he was good... -40 more.... Where are all the noobs!  Need 20 wins against them to get back on track and you're wondering how you're ever going to get to be a +2800 player?

So I'll reiterate.  The trick for GR is to, as M57 said, to specialize at something (which means you stop playing Wargear Warfare) and maintain a crazy high winning percentage at it or to learn to play a patient multi-player (10-16 players) game and win more then your fair share of them while out-dueling even a few high ranking players in each game. 

Then as itsnot says, if you then want to grow your CP, it's all out the window again.  So, if you want to rank high in "all" the standings, probably better to focus on your CP first and then once you've locked up your top ten spot there then to work on your GR.

Also, no one gets high in the rankings playing lotteries.  Any one lottery win will only get someone 200 points, or so, good case scenario at a lower ranking, and far less as they climb the rankings.  Then, as it's called a lottery, they have a much better chance of losing then winning which doesn't pay off as your ranking climbs because for every one you entered you'd be risking even more then half of what you hope to win...  Not a recipe for success.  Better to learn to win a 12-16 player game 40%-50% of the time then to win a 16 player lottery 6.25% of the time.

Hope that helps?