Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mic Qsenoch]

I agree with what you're saying and I still maintain that mostly we're just using differing definitions of the same core concept. My thought is that, "strategy" is basically a complex decision tree you create for yourself at the beginning of the game.  Typically these decision trees only last for two to three turns before they have to be reformulated. "Execution" or "tactics" is what you do when you encounter a situation that your decision tree doesn't account for. Once you've reached the end of your decision tree, you have to formulate a new one. You can call that strategy, you can call that execution, it's really 6 of 1 half dozen of the other, because if you'll notice in my hierarchy I put a much greater emphasis on "execution" than I do "strategy".

So in this case, the decision tree could be: Turn 1, Forager. Turn 2, Forager. Turn 3, if $4 buy Death Cart, if $3, buy silver. Turn 4, if $4 buy death cart, if $3 and no silver, buy silver. Now, there are plenty of things that could happen here that don't fall within this decision tree. If that happens, that's the 'execution' part. "Good strategy, bad execution" would be if I encountered one of those situations and made a bad decision. (Of course that's assuming that Forager/Forager is a good opening, which I'm not sure it is)

Yes, I don't think we disagree much about this. I am just trying to do some propaganda against what I perceive to be a problem in typical Dominion strategy discussion. And I think it bleeds into and harms our ability to play well also. And your responses help me clarify my thoughts (though probably only for me!) Sorry for derailing your thread, and if a mod wants to split this nonsense off that's great.

I want to dismiss the "develop a decision tree for the future" approach as being counterproductive to getting a realistic picture of how my decisions will impact my future options. Rather I'd just like to picture "what options will I have in the future if I choose X now." I don't need to decide between those options yet.

If you paint yourself too rosy a picture, then when the real shuffles arrive you are at a loss for what to do. If you dream too small then you often miss opportunities that could give you complete control of the game. But if you look at your turn with fresh eyes then you can consider the options with a little less bias. The advantage of making plans is that it might save you some brainpower, but it always restricts you by making your thoughts lean towards the plan first.

[Witherweaver]

I agree with what you're saying and I still maintain that mostly we're just using differing definitions of the same core concept. My thought is that, "strategy" is basically a complex decision tree you create for yourself at the beginning of the game.  Typically these decision trees only last for two to three turns before they have to be reformulated. "Execution" or "tactics" is what you do when you encounter a situation that your decision tree doesn't account for. Once you've reached the end of your decision tree, you have to formulate a new one. You can call that strategy, you can call that execution, it's really 6 of 1 half dozen of the other, because if you'll notice in my hierarchy I put a much greater emphasis on "execution" than I do "strategy".

So in this case, the decision tree could be: Turn 1, Forager. Turn 2, Forager. Turn 3, if $4 buy Death Cart, if $3, buy silver. Turn 4, if $4 buy death cart, if $3 and no silver, buy silver. Now, there are plenty of things that could happen here that don't fall within this decision tree. If that happens, that's the 'execution' part. "Good strategy, bad execution" would be if I encountered one of those situations and made a bad decision. (Of course that's assuming that Forager/Forager is a good opening, which I'm not sure it is)

Yes, I don't think we disagree much about this. I am just trying to do some propaganda against what I perceive to be a problem in typical Dominion strategy discussion. And I think it bleeds into and harms our ability to play well also. And your responses help me clarify my thoughts (though probably only for me!) Sorry for derailing your thread, and if a mod wants to split this nonsense off that's great.

I want to dismiss the "develop a decision tree for the future" approach as being counterproductive to getting a realistic picture of how my decisions will impact my future options. Rather I'd just like to picture "what options will I have in the future if I choose X now." I don't need to decide between those options yet.

If you paint yourself too rosy a picture, then when the real shuffles arrive you are at a loss for what to do. If you dream too small then you often miss opportunities that could give you complete control of the game. But if you look at your turn with fresh eyes then you can consider the options with a little less bias. The advantage of making plans is that it might save you some brainpower, but it always restricts you by making your thoughts lean towards the plan first.

[Mic Qsenoch]

If only I had known about this gif I could have saved some typing!

[Polk5440]

I also do not like the victory card buy rules or the copycat engine section.

Though I do see (I think) where you are coming from with the "copycat". A clueless player has no idea what is going on with complex boards. Trying to do what the smart guy seems to be doing will at least force beginners to think a little and play with cards they may not have otherwise picked up. Who knows, they may luck into a win, too. 

Also, the point of resorting to buying a village or doing what you would do in big money if you are clueless when playing the "copycat engine" seems like you are leading beginners in the wrong direction. I would remove that.

[Witherweaver]

Copying is good because you can't help but wonder why you do things as you do them.

[TheExpressicist]

Yes, I don't think we disagree much about this. I am just trying to do some propaganda against what I perceive to be a problem in typical Dominion strategy discussion. And I think it bleeds into and harms our ability to play well also. And your responses help me clarify my thoughts (though probably only for me!) Sorry for derailing your thread, and if a mod wants to split this nonsense off that's great.

I want to dismiss the "develop a decision tree for the future" approach as being counterproductive to getting a realistic picture of how my decisions will impact my future options. Rather I'd just like to picture "what options will I have in the future if I choose X now." I don't need to decide between those options yet.

If you paint yourself too rosy a picture, then when the real shuffles arrive you are at a loss for what to do. If you dream too small then you often miss opportunities that could give you complete control of the game. But if you look at your turn with fresh eyes then you can consider the options with a little less bias. The advantage of making plans is that it might save you some brainpower, but it always restricts you by making your thoughts lean towards the plan first.

Agreed. Whether you frontload the computations in advance ("strategy"), or do them on the fly ("execution"), ultimately it's still the same computation. Ultimately that's the reason why I say "execution" is more valuable than "strategy"; people tend to be much worse at "execution" than they are at "strategizing". This gets a bit into a specific individual and how they process information, but I suspect this is because people tend to be much better at data storage than they are performing complex calculations quickly. 

 

[eHalcyon]

Instead of just going copycat, why not use what you've observed up to that point to just experiment as you will?  If you and your opponent mirror, great, you can analyze where you diverged and what mistakes each player made.  If you don't mirror, you can still consider mistakes made and get an idea of how the two strategies match up.  If you get absolutely crushed, you can focus on your opponent's choices and think of that player's choices on your own time.  Copying on the fly isn't really helpful. 

The more you do this, the better an idea you have of Dominion as a whole.  In other words, play more games.

[TheExpressicist]

I also do not like the victory card buy rules or the copycat engine section.

Though I do see (I think) where you are coming from with the "copycat". A clueless player has no idea what is going on with complex boards. Trying to do what the smart guy seems to be doing will at least force beginners to think a little and play with cards they may not have otherwise picked up. Who knows, they may luck into a win, too. 

Also, the point of resorting to buying a village or doing what you would do in big money if you are clueless when playing the "copycat engine" seems like you are leading beginners in the wrong direction. I would remove that.

Well, remember this is not really meant as a "This is how you build a good engine". It's more like, "You have a great learning opportunity here. Here's how to make the most out of it." (That said, I do need to clarify that it's the same *ACTION* buying rules as Big Money. Obviously you rarely want treasure in an engine.)

But yeah, part of the idea here is, if you choose a card that is completely out of sync with the engine, you should try to learn from it and figure out *why* that card didn't work, but the other cards your opponent picked DID work.

[TheExpressicist]

Instead of just going copycat, why not use what you've observed up to that point to just experiment as you will?  If you and your opponent mirror, great, you can analyze where you diverged and what mistakes each player made.  If you don't mirror, you can still consider mistakes made and get an idea of how the two strategies match up.  The more you do this, the better an idea you have of Dominion as a whole.

In other words, play more games.

Because new players only have a certain amount of mental real estate. If you're spending most of your mental energy trying to figure out your own plan, you're wasting a golden opportunity to learn from a player who already has a plan (which is probably a lot better than yours would be in the first place).

Plus people are experts at rationalizing their poor decisions. If you work really hard to build the Little Engine That Couldn't, it's a lot more likely that you will try to irrationally justify why you tried doing it Your Way. Copying the opponent makes it very impersonal and takes that potential bias out of the equation.

[Mic Qsenoch]

Because new players only have a certain amount of mental real estate. If you're spending most of your mental energy trying to figure out your own plan, you're wasting a golden opportunity to learn from a player who already has a plan (which is probably a lot better than yours would be in the first place).

If the goal is just to expose yourself to higher level play, then watch videos or read logs between high level players. Then you get two people playing well, not just the opponent you're copying. And using videos or logs means your mind isn't distracted by a bunch of clicking on cards. So you free up even more mental real estate for other stuff.

If you want to play the game, you should be actively engaged in what your deck is doing since you have way more information about those decisions. You can't see your opponent's hand and may be missing a lot of the subtlety related to how they are building and playing their engine.

You can absolutely learn a lot by watching your opponent while playing, but I'm not sure this "shut off most thinking about your play and just watch them" is a very quick way to get better.

[eHalcyon]

Because new players only have a certain amount of mental real estate. If you're spending most of your mental energy trying to figure out your own plan, you're wasting a golden opportunity to learn from a player who already has a plan (which is probably a lot better than yours would be in the first place).

Plus people are experts at rationalizing their poor decisions. If you work really hard to build the Little Engine That Couldn't, it's a lot more likely that you will try to irrationally justify why you tried doing it Your Way. Copying the opponent makes it very impersonal and takes that potential bias out of the equation.

If they've already played hundreds of games of big money, they aren't new players anymore.

As for rationalizing, just put that in the guide.  "You may think you had a good plan and just ran into bad luck, but you're probably wrong.  Even experts make mistakes, and you're no expert at this point.  Accept that you aren't perfect and learn from those mistakes."

[Polk5440]

You can absolutely learn a lot by watching your opponent while playing, but I'm not sure this "shut off most thinking about your play and just watch them" is a very quick way to get better.

That's the danger with watching videos, though, too. Your thinking can just shut off because you are not being forced to play yourself. You are being entertained.

I do agree STUDYING videos and logs of the best players are the best way to learn quickly, but that's not the same thing as simply watching.

(That said, I do need to clarify that it's the same *ACTION* buying rules as Big Money. Obviously you rarely want treasure in an engine.)

That is how I read that section and what prompted the comment: Same treasure/victory card buying rules.

[TheExpressicist]

The "copycat" strategy is really just an educational gimmick. It's a good way to get someone invested enough that they take it seriously but not so invested that they start rationalizing their mistakes.

If getting people to think rationally was as easy as just telling them to be rational, there wouldn't be any billion dollar weight loss, self help or stop-smoking industries...

[WanderingWinder]

The big money strategy is wrong on most of the points, but whatever, it's not so hugely consequential - you go the biggest gist of it right. (Specifically, I think you way overvalue "big draw", which is good).

Honestly I don't think a few hundred games of big money is worth it. Somewhere between two and maybe 50 max. In terms of overall what's best, it's pretty niche though.


The biggest problem, though, is the engine section. And the biggest problem with that is that you give no indication whatsoever of what an engine is. You want to say that your deck is usually going to do more powerful and more consistent things if you can get to the point where you are playing all your (important) cards every turn, and once you can get your deck to coalesce to that point, you can really start unlocking the powerful potential of most of your cards, as well as having something which is really consistent. You want to explain that this means you want to have a large amount of +cards relative to overall cards in your deck. And then villages come in because you need to be able to have the actions available to play your action cards.


In terms of your A beats B beats C stack, I think that that is mostly wrong, but it's at least not very helpful based on the very subjective and relative nature of "good" and "bad". The biggest point I think you could make here is that good execution beats bad execution, which is pretty just the biggest truth.


Copying your opponent in the same game really doesn't work. You end up not getting the same shuffle luck as them, and what you want to do at any given point really really depends on that.


Building an engine is much more than buying random villages at some point. If you don't want them to buy treasure, you really need to say that; it's not at all obvious you don't want treasure, and very often you actually do, even if it's only one or two silvers. Actually, you don't talk about keeping thin at all, which is very much not obvious, especially given that you haven't talked about what engines are trying to do (but really, not obvious anyway).

[eHalcyon]

The "copycat" strategy is really just an educational gimmick. It's a good way to get someone invested enough that they take it seriously but not so invested that they start rationalizing their mistakes.

If getting people to think rationally was as easy as just telling them to be rational, there wouldn't be any billion dollar weight loss, self help or stop-smoking industries...

If telling them bluntly doesn't work, I don't think your gimmick does either.  If nothing else, you can blame it on first player advantage or shuffle luck.  McQ has pointed out how you lack a lot of info behind their decisions, making the exercise much less educational. And since it falls apart after turn 4, you're really doing your own thing anyway!  Might as well make your own choices from turn 1.  If you are so irrational that you can't acknowledge the possibility of your own imperfection, you are not going to learn much no matter what educational gimmicks you try.

[Gherald]

Gherald's Guide to Getting Better at Dominion

1. Play opponents who are better than you

2. Pay attention to what your opponent is doing, and why it is working

[markusin]

Gherald's Guide to Getting Better at Dominion

1. Play opponents who are better than you

2. Pay attention to what your opponent is doing, and why it is working

3. Post game logs in the "Help" subforum to find out why what you did didn't work.

[Davio]

Gherald's Guide to Getting Better at Dominion

1. Play opponents who are better than you

2. Pay attention to what your opponent is doing, and why it is working

3. Post game logs in the "Help" subforum to find out why what you did didn't work.

4. Try wacky stuff that might or might not work. You won't learn a lot from always doing the same thing.

There are a lot of specialized cards in Dominion and you might miss out on the fun when they're actually useful if you only play by the books.

They key to becoming a good player is understanding why and when certain combinations work and why and when they don't.

[TheExpressicist]

@Gherald: The problem is, when you're new, everyone is better than you. And unless you have a cohesive strategy, almost everything is going to work against you. And, furthermore, players who are actually good are very unlikely to be playing you.  So you're going to be learning from a bunch of mediocre players.

@WanderingWinder and @eHalycon: For context, see my discussion with Mic Qsenoch. After playing a bunch of games under a new account following these rules, I need to either scrap or seriously retool the engine section. Even though I told myself I was committed to following the rules, when it came time to do the "copycat" strategy, I ended up just building my own engine rather than copying my opponent. If I'm not able to follow my own "copycat" rules, I doubt a new player would either.

@WanderingWinder: Remember that the point of the Big Money section is not to say "This is the best way to play Big Money." It's to say, "If your opponent can't beat you when you're using this method, their strategy is probably no good and you should try to learn from that. If they DO beat you, they probably have a pretty decent strategy and you should try to learn from THAT."

[WanderingWinder]

@WanderingWinder: Remember that the point of the Big Money section is not to say "This is the best way to play Big Money." It's to say, "If your opponent can't beat you when you're using this method, their strategy is probably no good and you should try to learn from that. If they DO beat you, they probably have a pretty decent strategy and you should try to learn from THAT."

This can't be something I'm remembering, since it isn't something I knew to start with. But the bigger point is that it's bad advice. There are plenty of bad strategies that beat big money, particularly on a given board. And quite frankly, I think people can figure out what is good well enough just by looking at what wins, especially coupled with what is by far the best advice in the article of always trying to play against stronger opponents.

[TheExpressicist]

@WanderingWinder: Remember that the point of the Big Money section is not to say "This is the best way to play Big Money." It's to say, "If your opponent can't beat you when you're using this method, their strategy is probably no good and you should try to learn from that. If they DO beat you, they probably have a pretty decent strategy and you should try to learn from THAT."

This can't be something I'm remembering, since it isn't something I knew to start with. But the bigger point is that it's bad advice. There are plenty of bad strategies that beat big money, particularly on a given board. And quite frankly, I think people can figure out what is good well enough just by looking at what wins, especially coupled with what is by far the best advice in the article of always trying to play against stronger opponents.

I agree that Big Money is rarely the dominant strategy on any given board. But I think it's fair to say, if your opponent LOSES to your Big Money (especially a big money implementation as simplistic as the one I'm suggesting), they probably did something wrong.

My issue that I keep coming back to is, when you first start playing, you really don't get any opportunities to play against opponents who are significantly better than yourself. Most top tier players won't play anyone with a Goko ranking of <5k. And most good players won't play people with a ranking <3.5 to 4k. Which means until you break that barrier, you're stuck playing against a lot of mediocre players. The idea of this guide is to help get you past the "Only crappy players will give me the time of day" hump, so that you can start learning against players who are actually worth learning against.

[WanderingWinder]

@WanderingWinder: Remember that the point of the Big Money section is not to say "This is the best way to play Big Money." It's to say, "If your opponent can't beat you when you're using this method, their strategy is probably no good and you should try to learn from that. If they DO beat you, they probably have a pretty decent strategy and you should try to learn from THAT."

This can't be something I'm remembering, since it isn't something I knew to start with. But the bigger point is that it's bad advice. There are plenty of bad strategies that beat big money, particularly on a given board. And quite frankly, I think people can figure out what is good well enough just by looking at what wins, especially coupled with what is by far the best advice in the article of always trying to play against stronger opponents.

I agree that Big Money is rarely the dominant strategy on any given board. But I think it's fair to say, if your opponent LOSES to your Big Money (especially a big money implementation as simplistic as the one I'm suggesting), they probably did something wrong.

This is true, but I don't see how it's relevant.

My issue that I keep coming back to is, when you first start playing, you really don't get any opportunities to play against opponents who are significantly better than yourself. Most top tier players won't play anyone with a Goko ranking of <5k. And most good players won't play people with a ranking <3.5 to 4k. Which means until you break that barrier, you're stuck playing against a lot of mediocre players. The idea of this guide is to help get you past the "Only crappy players will give me the time of day" hump, so that you can start learning against players who are actually worth learning against.

The thing is, they should just learn against the players they can play against. Even if they can't get games against the strongest players, they can indeed get games against players stronger than they are. And they can get exposed to more and learn. Your argument is that they will end up learning bad strategies this way, because what works against weaker players might not be good. But I would counter that, while this may be true to some extent, what they are learning the way you suggest is certainly not good, and what they are learning otherwise, learning for themselves rather than following this script, is way more likely to be better.

[TheExpressicist]

But I would counter that, while this may be true to some extent, what they are learning the way you suggest is certainly not good,

I agree that the methods I am suggesting are NOT optimal Dominion strategy. What remains to be seen is if this is a better method to LEARN optimal Dominion strategy. Which brings me to the next point...

what they are learning otherwise, learning for themselves rather than following this script, is way more likely to be better.

This is where I pretty strongly disagree. I think there's an overwhelming amount of evidence that the prevailing common wisdom of "how to learn Dominion strategy" isn't working. There are too many mediocre players. The median skill on Iso is Level 12. That should not be the case if the prevailing teaching methods were effective. I consider myself a "very slightly above average" Dominion player. But according to the leaderboards, I am a somewhere around a "Top 10% player". And back when I was playing more regularly, I was a "Top 5% player".

It just doesn't make sense to me that after two months of playing, and using fairly simplistic heuristics, that I could somehow be THAT good. I highly doubt that I am some kind of prodigy or savant. My initial conclusion was simply that Dominion was a broken, solved game. But after playing several matches against elite, top-tier players, it became clear that this was not the case. So the only other reasonable conclusion I could think of was that the way people are learning Dominion skills right now is flawed and highly ineffective.

My method might not be the best one, but I think its a decent start at improving upon something that clearly isn't working.

[Awaclus]

This is where I pretty strongly disagree. I think there's an overwhelming amount of evidence that the prevailing common wisdom of "how to learn Dominion strategy" isn't working. There are too many mediocre players. The median skill on Iso is Level 12. That should not be the case if the prevailing teaching methods were effective. I consider myself a "very slightly above average" Dominion player. But according to the leaderboards, I am a somewhere around a "Top 10% player". And back when I was playing more regularly, I was a "Top 5% player".

It just doesn't make sense to me that after two months of playing, and using fairly simplistic heuristics, that I could somehow be THAT good. I highly doubt that I am some kind of prodigy or savant. My initial conclusion was simply that Dominion was a broken, solved game. But after playing several matches against elite, top-tier players, it became clear that this was not the case. So the only other reasonable conclusion I could think of was that the way people are learning Dominion skills right now is flawed and highly ineffective.

My method might not be the best one, but I think its a decent start at improving upon something that clearly isn't working.

Well, if you're an average player, your estimated skill is 25 and then there's the uncertainty which brings the level down so it's just natural that the median would be 12, regardless of how effective the teaching methods are. Even if everyone was almost as good as Stef, the median would probably still be 12.

Also, you consider yourself an average player because you're on f.ds and you compare yourself to other f.ds people. A vast majority of the players suck at the game, not because our prevailing common wisdom of "how to learn Dominion strategy" is wrong, but because those players have never heard of said wisdom, and because they are completely fine with sucking at Dominion so they don't even waste any effort trying to get better.

[WanderingWinder]

But I would counter that, while this may be true to some extent, what they are learning the way you suggest is certainly not good,

I agree that the methods I am suggesting are NOT optimal Dominion strategy. What remains to be seen is if this is a better method to LEARN optimal Dominion strategy. Which brings me to the next point...

what they are learning otherwise, learning for themselves rather than following this script, is way more likely to be better.

This is where I pretty strongly disagree. I think there's an overwhelming amount of evidence that the prevailing common wisdom of "how to learn Dominion strategy" isn't working. There are too many mediocre players. The median skill on Iso is Level 12. That should not be the case if the prevailing teaching methods were effective. I consider myself a "very slightly above average" Dominion player. But according to the leaderboards, I am a somewhere around a "Top 10% player". And back when I was playing more regularly, I was a "Top 5% player".

It just doesn't make sense to me that after two months of playing, and using fairly simplistic heuristics, that I could somehow be THAT good. I highly doubt that I am some kind of prodigy or savant. My initial conclusion was simply that Dominion was a broken, solved game. But after playing several matches against elite, top-tier players, it became clear that this was not the case. So the only other reasonable conclusion I could think of was that the way people are learning Dominion skills right now is flawed and highly ineffective.

My method might not be the best one, but I think its a decent start at improving upon something that clearly isn't working.

Okay, I figured it out. Median 'skill' (actually, level) being level 12 tells you very little - the ratings are entirely relative. If you add 50 points to everyone's rating, suddenly the median skill is 62, but nobody is actually better. Median being level 12 rather than 0 is mostly down to the fact that as you play more, uncertainty naturally drops.

FWIW, the isotropish board actually shows you as just outside the top 15%.

I think you have to judge how good someone is by a kind of eyeball test.


Anyway, your big point is, you don't feel like you're great, and you feel like you're way too close to the top of the standings. Thing is, there just aren't that many people to compete against. You can be pretty mediocre and get to the top, say, 200, just because there aren't all that many people who are playing and, more important, who are really making a concerted effort to win a lot.

PPE: Seems Awaclus has covered the same idea basically.