"Luck"

[playBunny]

Fact: The winner of a game is almost always the player who gets the most luck. ("Luck" being a mathematically defined and measurable value)

Fact: Strong players win more games than weak players.

Conclusion: Strong players therefore get more luck.

Possible implications:

Your choice... see poll

[dropper]

Backgammon can't be learned by silently watching. The reasons for moves (the skill) is fairly invisible without knowing the strategies behind them. It is very easy to perceive that you're being unlucky and your opponent is being lucky when actually you are being out played.

Start with learning about diversification and duplication if you're not familiar with those concepts.

http://www.bkgm.com/articles/

[ah_clem]

Sounds like somebody doesn't understand the difference between correlation and causation.

[playBunny]

Sounds like somebody doesn't understand the difference between correlation and causation.

Perhaps someone who does could explain what someone who implies that they do might be saying in terms of whatever they're referring to? ;-)

[Tom]

I think the biggest difference between a Weak player and a Strong player is that the strong player makes moves that allow them to leverage more lucky rolls at any given time.

Putting it another way, given a specific sequence of rolls for both players a Strong player will seem to have more lucky rolls, even if they traded sides and re-ran the sequence.

Now... Luck has to be defined as a roll that allows a large change in equity of the match.

And while some may think that 4 doubles in a row is lucky... sometimes it can be very unlucky!

tom

[sixty_something]

it is this Zen-like chicken or egg question which makes me return, again and again, to this game to play and to study .. if luck and skill are considered as truly independent, which technically they are, then strong players will be equally as unlucky as lucky with individual rolls .. regardless, there is no doubt when playing highly rated bots or players the thought "you lucky bastard" often crosses my mind when unseen or unexpected variations unfold to my opponents advantage .. therein, IMHO, lay the highly charged emotions so often encountered in this game .. of course, when they unfold to my advantage, it is my skill, not luck that reaps the rewards

but the perception of luck, whether in backgammon or other endeavors, is a different story .. perceived luck and perceived skill are the buffers that enable and even attract me to this game .. continual defeat when facing more skillful players in a game like chess, where luck is completely absent, can be simply ego crushing .. however, the chance of demonstrating my skill with an occasional victory against an unlucky much more skillful player or bot makes coming back for more not only entertaining but enticing

[dropper]

nicely said, sixty 

[Tom]

if luck and skill are considered as truly independent, which technically they are, then strong players will be equally as unlucky as lucky with individual rolls ..

I do not consider them independent.

What makes a Roll so Lucky?

When it changes the advantage you have greatly.

So you could say that a Skillful player moves their checkers so there are more possible rolls that are deemed Lucky than their opponent.

For example:

A strong player will have 2-3 possible rolls that will allow them to make a point on the way to building a prime.
So they then have a 33% chance of getting lucky (building that prime to block you in)

The individual rolls themselves are not lucky... is rolling 6-6 lucky? sometimes, and other times it is very unlucky (stuck on bar)

Skill is what you do with the rolls you were given and maximizing the number of rolls that benefit you in the future (ie are deemed lucky)

tom

[sixty_something]

I do not consider them independent.

therein lays the rub which brings me back to playBunny's original comment describing "Luck" as "being a mathematically defined and measurable value"

if the definition of "Luck" implies that used by GNUbg and Snowie, i.e. a measure of change in equity as a result of a single roll, then i totally agree with you, Tom .. bots make choices which create situations that optimise equity gain which yield more "Luck" and are calculated with supernatural precision .. skilled players effectively do the same, but are mere mortals computationally

therefore, it is fair to say "Luck", defined as above, is made not rolled .. however, it is the interplay between "Luck" and the good old fashioned luck of the roll that makes this game so challenging to learn, so much fun when winning, and so annoying when losing

p.s. thanks dropper

[eddieballgame]

"Luck".......does it effect "how" you move your checkers? I think not.  

[blitzxz]

Fact: The winner of a game is almost always the player who gets the most luck. ("Luck" being a mathematically defined and measurable value)

Fact: Strong players win more games than weak players.

This has been discussed before. The answer to this dilemma is that stronger players truly are more often luckier (according to bot luck analyzes) but by smaller amounts. This is because stronger player needs less luck to win and small amount of luck is more likely to happen. When weak player is lucky he usully is very lucky (and wins the match) but this happens more rarely. All in all the amount of luck evens out.

So the mathematical luck is not complitely independent of skill. Expected value is always 0 but the distribution depends from skill.

[roadkillbooks]

I think this is correct. Say the less skillful player needs exactly a specific joker to dominate the game.  That swing of luck is large compared to there preroll equity.  A skillfull player my distribute there pieces to take advantage of  number of rolls.. Say one specific joker still dominates but 2 more combinations still improve greatly..Even when the spicific 1/18 joker hits the gain in equity would be less because the other rolls were also set up.  Likewise when the other rolls hit it is still 'lucky'.
Anyway, I just like to say..If i was never lucky I wouldnt ever play backgammon.  This is just a random comment of this topic and I havent acully checked the numbers. 

roadkillbooks

[don]

Duplication and Diversification.  Use these two principles in your favor, you will have more rolls that are good for you, bad for your opponent.

The opposite is also true.

In the long run, this is what is perceived as 'luck' by the ignorant.

--
don

[blitzxz]

And to make my point more clear we can also think an simplified example. There's perfect player and player who always loses 10 % match equity because of his errors. This means that on average the perfect player wins 60 % matches. In every one of those matches he need 40 % luck. But the error maker wins 40 % matches but in every one he needs 60 % luck. So the perfect player is more often luckier but by smaller amount and it evens up. And in every one those games the luckier player also wins. Of course this not the reality but it's how it generally affects the distribution of luck.

Subjective luck is complitely different thing. Player who is bad in backgammon can't see his errors or his good luck. And he explains his losses with bad luck and doesn't take any fault on himself. But this has lot to do how you see the world any way. Are you explaining events with "luck" or reason. The reason however is also wrong (and actually source of luck thinking) because ramdom events are only controllable statictically. I have said before that if you're taking part only in one major tournament in year your bad/good luck could easily last longer then life time. Good gambler will make sure that he plays so often that the skill will eventually prevail and on the mean time he doesn't bother about lucky or unlucky streaks. And also one interesting point is psychological. People who think they are lucky are generally lot more happier then people who think they are unlucky or even people who don't believe very much in luck. It's all about how you see the world.  

[stog]

y, if i diversify & duplicate, do they throw a double of the exact number? (man going through mid game crisis)

[roadkillbooks]

y, if i diversify & duplicate, do they throw a double of the exact number?

And what don says is right on also.  If you duplicate and reduce there lucky numbers to double 4's (and they will get them if u are in crisis) their luck will be huge.  IF they leave a hit open to 45..u will obviously hit that number twice as much..So while they were twice as lucky to make their hit..you will do it twice as often..and will thus have twice as many games in which you are lucky.

thanks  guys i think we put it together very nicely...so to say "If I wasnt lucky I wouldnt play backgammon" is a multifaceted comment that appears modest but has its hint at sarcasism since when we talk we use our own definition of the words.  I do get tired of several people that consistenly go.."lol, what a joke" etc... I mean even bad people win games with a couple of big doubles at the end of the game..what's wrong if a consistently lucky person does it too?

[KDP]

Learning Understanding Consistency Knowledge

[pck]

From the poll options:

"Strong players create situations such that they can't help but get better dice, on average"

As much as I can see the attraction in the thought that good plays open up more chances to get lucky, I still think that the notion is misguided. Good plays open up more ways for me to win - but only in the sense that they open up more ways than worse plays.

Apologies in advance for the terrible long-windedness of this.


Let's say I make a good play GP, after which I have a certain equity E. Then, after my opponent's turn, I roll some roll R.

Now even before I rolled R, the "good opportunities" (= ways for me to win the game) I'm entitled to  - due to my having made the GP - were already contained in (being measured by) E. (That is precisely the job of the concept of equity.) Had I made a worse move, E would be less, meaning I would need more luck in the rest of the game to win it. "More luck" here means the collected luck of all subsequent rolls.

Do note the difference of "avoiding having to be unnecessarily lucky in the rest of the game" (we usually call this "skill") and a "lucky roll", which significantly increases your equity, but is not due to skillful playing or decision-making. Call this distinction "D".

As has been mentioned in this thread, the luck contained in an individual roll is measured as the difference in equity of my previous position and the position that will result from the best play I can make with R.

Now let's say R realizes (makes possible the realization of) one of the opportunities the GP opened up for me (make an inner point, block, etc.), or perhaps at least part of such an opportunity. Am I to be counted lucky to have rolled R? The answer depends on the probability of the opportunity. Let's say many different rolls will lead to it. Call this "Scenario A". Then my equity will not increase much (perhaps not at all) after making the best possible move R allows. Why? Because the (high) probability of the realization of my opportunity was already accounted for in E. Hence, the difference in equity will be small, and I will not be counted lucky on behalf of R. (The roll that allowed the GP might of course still have been very lucky.)

If there are few rolls that will lead to the opportunity in question ("Scenario B"), but the opportunity is nevertheless realized, my equity will increase by a substantial amount, and in this case I will be counted lucky on behalf of having rolled R. Again, the probability of the opportunity's realization was already accounted for by E (this time by E being low, or more precisely, the part of E which corresponds to our opportunity being small).

So I'm only counted lucky if my roll opens up significant new chances for me to win the game that didn't exist before, or were estimated to be low by E. But the novelty of these chances means that they cannot be due to my having made the GP earlier. (If they were, E wouldn't have been low in the first place.)

So making the GP earlier does not and cannot lead to luckier rolls later ("lucky" here means lucky on a per-roll basis). Both Scenario A and Scenario B are possible. The GP itself (i.e. the fact that I demonstrated skill) has no influence on the frequency of those. I don't get A-Scenarios more often just because I made the GP. It is rather that with the GP I didn't waste any of the opportunities I got when I rolled the GP's roll. This is what distinction D was all about.

To summarize:
Good plays do open up more ways for me to win - but only in the sense that they open up more ways than worse plays. However, they do not open up more ways to get lucky, because winning the game after a good play needs less luck than would have been needed if I had made an inferior play - which is why the good play deserved to be called so in the first place.

Good playing doesn't lead to more luck, it reduces the amount of luck necessary to win the game, compared to what would have been necessary had an inferior play been made. From a hindsight-perspective, these can seem to be the same thing. But we must not confuse luck on a per-roll basis with our attempts to avoid unnecessary increases of the amount of luck necessary to win the game (i.e. our attempts to make good plays). The two are of course related. But as players we have influence only over the latter.


pck