[SPOILER] Dark Savant graduates from blatant n00b to plain n00b

[BRickAstley]

Oh yes, and fight the good fight DS, you can do it!

[Dark Savant]

I haven't yet logged in (I might actually do so during work hours, which I've been avoiding), but I have a meta question.

If Donovan Zoi wins the 36% roll and takes my capital, he'll have a serious advantage over everyone else.  If he loses it, I have a good chance of fighting back.

He may refrain from attacking, of course, but my question concerns if he doesn't.

One thing I can do is to stack all my workers in my capital for one turn (they will be vulnerable without military cover, anyway, which I don't have this turn but will the next -- if I'm still alive ), which also happens to encourage him to attack.  If he does so and wins, he also gets the three workers, which is just a ridiculous haul for this early in the game.

I'm actually concerned about unbalancing the game in his favor.  Should I worry about that?

My warrior only gets odds on him because I founded a religion, actually.  That got me in trouble, but he'd have odds on my lone warrior without the extra +20% bonus defense I currently have from the holy city.  I had the warrior step out the city a few turns ago (I was thinking of trying to reclaim my second city with 2 warriors, then realized that was dumb), which lost me the full fortify bonus, so I don't actually have that.     I do still have enough of it for odds, though.

I'll detail it, since it's so important:

Donovan's Zoi holkan is strength 4, and is Woodsman II, so it has no bonuses.

My warrior is strength 2, no promotions, and has 40% (culture) + 15% (fortify) + 25% (warrior defense bonus) + 25% (hill), for a total of 105%, which is just enough to tilt the odds in my favor.

[Mardoc]

Coming from not being in academia past what was needed to get through college, is it usually appropriate to just go to the conclusion of those big articles like the one Bacchus listed?

So, the biggest switch from undergrad to graduate+ science is the switch from confident certainty to confident uncertainty.  Back to Dark Savant's original point, really: a scientific paper is evidence, but it's not facts.  To get a good feel for what's believed, you really need to read all the papers in the field.  Literally all: when I was in grad school, my advisor would give us tongue-lashings if he knew about any paper related to our work that we didn't also know.  And even then, you may be learning more about what the scientists believe than about what is actually true.

Sometimes an author will do this for you, in which case it's called a literature review, or review paper, or something along those lines.  What Bacchus linked is a good example.  You still have to decide if you trust that particular author, though.  Ideally this would involve reading a sampling of the papers they reference and comparing the papers to the summary of the papers - it would also involve looking for any responses, or important publications that were left out.  It's not common to write a review paper if you've got an axe to grind, but it happens.

Even a good review article only captures the overall sense of the field, what the scientists currently believe and are basing their current research plans on.  I would be generally willing to trust any statement along the lines of "Definitive causal conclusions cannot yet be drawn from this emerging body of literature," while statements like "may be linked to autism" really mean "I would love to see a good quality experiment about this!"  The authors think it's likely enough to be true that they don't think it's a complete waste of time to study.  They may not have gotten to the level of certainty where they would personally act on the information, though, and are certainly not to the point where they would be comfortable promising anything.  Could be any level of certainty from "yes, of course that's the cause, there's just a few more people I need to convince" to "well, we haven't disproven it yet and there are loud people who believe it, I guess it's worth running an experiment to shut them up".

Fundamentally, if you want 'things that the scientific community agrees to be true', you need to be looking at textbooks.  Anything more recent than that is, at best, 'things the scientific community are working on and has some ideas about' and is quite likely to be 'things the scientific community has no clue about but agrees would be nice to know.'

consequently haven't been wedded to the thiomersal/mercury hypothesis at all.

Glad to hear it.  That's one of the hypotheses that's got the most evidence of being false: they removed it from vaccines and that didn't cause any noticeable decline in autism diagnoses.


Back to the game:

I'm actually concerned about unbalancing the game in his favor.  Should I worry about that?

If you are taking the best option for yourself, and it happens to tilt in his favor, that's fine.  Probably actually better than trying to balance the rest of everyone - it's not exactly kingmaking to try something useful and get unlucky, but if you try to balance his benefit vs. everyone's benefit, it's hard to be fair.

If you stick to deciding based on your own situation, then I don't think anyone will complain.

[RefSteel]

If you stick to deciding based on your own situation, then I don't think anyone will complain.

Realistically, other players are certain to complain when and if they find out practically anything you do - and lurkers might criticize whatever decision you make in the no-players thread too. Sort of like being a mapmaker, you can't please everybody!

But more realistically, Mardoc is right. My understanding is that you're thinking of stacking the workers in Glaurung in the hope that Donovan will then be more likely to attack the city - so it's a question of why you want him to do that:

- If you're hoping he wins the ~1/3 fight so you don't have to play anymore, there's probably a better solution - including solutions that guarantee you no longer have to play.

- If you think the ~2/3 chance of killing that W2 Holkan is better for your civ than letting the Holk stick around, in spite of the ~1/3 chance that you'd be eliminated, then putting Workers into your capital as lures has value, and it's just a question of whether you think that value is greater or less for the civ for each one than whatever else that worker could be doing instead.

- Like Mardoc said, don't worry about unbalancing the game, as long as you're making what you think is your own best move.

If you make what you think is your best move, and it doesn't work out, and Donovan Zoi ends up winning the game (or becoming a slightly more difficult snack for some other player to digest, or anything in between) there will be poor sports who try to blame you when they end up among the 24 non-winners of the game, but honestly that's a problem with them, not with you.

Also: I'm glad you're enjoying the off-topic discussion in the thread and the Orion Succession Game too!

[Dark Savant]

Coming from not being in academia past what was needed to get through college, is it usually appropriate to just go to the conclusion of those big articles like the one Bacchus listed?

So, the biggest switch from undergrad to graduate+ science is the switch from confident certainty to confident uncertainty.  Back to Dark Savant's original point, really: a scientific paper is evidence, but it's not facts.  To get a good feel for what's believed, you really need to read all the papers in the field.  Literally all: when I was in grad school, my advisor would give us tongue-lashings if he knew about any paper related to our work that we didn't also know.  And even then, you may be learning more about what the scientists believe than about what is actually true.

Here's what typically happens in my experience.

Let's say you find 20 papers evaluating "is X true".  You'll typically see something like:

• 13 papers conclude that X is true, using methodology and reasoning that appear perfectly sound.
  
• 2 papers conclude that X is false, but they are dated as they use old technology.  The conclusion is valid, and the methodology was fine back in the day, but the conclusion can't be taken seriously now.
  
• 1 paper concludes that X is true, but it's using completely different methodology from a field you do not specialize in.  It's nice to see outside validation, but boy do I not have the expertise to judge this.
  
• 1 paper concludes that X is false, but their statistical modeling is terrible.  How did this get published to begin with?  Oh, right, many scientists do not so much as understand what a p-value is.
  
• 1 paper concludes that X is true, but that conclusion does not actually follow from the data they present.  Did they just publish "X is true" to be cool like everyone else?
  
• 1 paper concludes that X is false, but their technical expertise is lacking, so you can't draw any conclusions from the paper anyway.  How did this get published to begin with?
  
• 1 paper concludes that X is false, and in this case, both their methodology and their reasoning actually seem valid.  No idea what happened here.  
  

So in that case, I roll with "X is true", and keep that last paper in mind as a reasonable argument to the contrary.  Maybe X is false, but for most practical purposes it is true unless new evidence to the contrary emerges.

This is with relatively simple questions, too.  There are many, many interesting biological/medical questions that would take tons of money or time to demonstrate even to the above level of satisfaction.

This is also why you should rarely draw conclusions just from one paper.  It gets worse in the popular press -- articles on CNN and the like often do not demonstrate that the author has a firm grasp of reading comprehension, never mind actual ability to critically evaluate a paper.  (The same happens for everything else, in my experience.)

It's kind of fun, actually.  Girlfriend points me to article on CNN.  I actually read the paper; it doesn't actually say what CNN says it does.  At least CNN doesn't lie outright; it gets worse if anyone tries to sell you things (with big business typically actually more honest than small -- that sounds strange to many people, but as with many other things, competence outweighs ethics).

Actual game:

If you are taking the best option for yourself, and it happens to tilt in his favor, that's fine.  Probably actually better than trying to balance the rest of everyone - it's not exactly kingmaking to try something useful and get unlucky, but if you try to balance his benefit vs. everyone's benefit, it's hard to be fair.

If you stick to deciding based on your own situation, then I don't think anyone will complain.

Okay, makes sense.  But boy, it feels really strange to have much of the fate of the entire game potentially riding on a single roll of the dice just because you made a single major oversight.  (If I were Donovan Zoi, I'd definitely roll the dice -- the odds are low but the payoff is high -- but I know I have an archer coming the next turn and he may not guess that, although I think he ought to.)

Maybe I should have whipped warriors, but that really doesn't get me that much if he comes at me with additional force, which I'm expecting if I survive the next couple of turns.

I was thinking of playing now, but I actually need to go home for my combat calculator (yeah I wrote my own) just so I can figure out things like what happens when a holkan tries attacking a 3-XP archer on flatland.

That means I can't play until this evening.  I've sent DZ a PM alerting him to that.

But more realistically, Mardoc is right. My understanding is that you're thinking of stacking the workers in Glaurung in the hope that Donovan will then be more likely to attack the city

Actually, it's more than there's not many places they can reasonably hide from the oncoming holkan -- it's Woodsman II so it has two moves in many cases.  I cannot provide military cover for them until my archer finishes.  He could always still attack my capital if I move the archer out to cover my worker horde, but then he could have done so immediately with both better odds and better reward.

There may be a better option, but there aren't many options to keep workers away from a 2-mover, and I don't know where he actually moved yet.

[Mardoc]

This is with relatively simple questions, too.  

Like, for example, "Can mice survive if I disable this gene, and if not, what symptoms do they exhibit?", right? Not even "What does this gene do in mice", let alone "what does this gene do in people and is it the key to curing cancer?" My experience is in chemical engineering, and our questions were along the lines of 'what IR peaks can we see on a platinum surface while it is catalyzing a moderately well-known reaction,' with the idea that would eventually build toward 'how does platinum actually speed up the reaction?'

[Dark Savant]

This is with relatively simple questions, too.  

Like, for example, "Can mice survive if I disable this gene, and if not, what symptoms do they exhibit?", right?  Not even "What does this gene do in mice", let alone "what does this gene do in people and is it the key to curing cancer?"  My experience is in chemical engineering, and our questions were along the lines of 'what IR peaks can we see on a platinum surface while it is catalyzing a moderately well-known reaction,' with the idea that would eventually build toward 'how does platinum actually speed up the reaction?'

Speaking of reactions, I always thought it was so amusing that the reaction mechanism of burning gasoline (or anything organic) is so poorly understood. Some of that is because we are either not doing it for the end produ cts (burning gas in a car), or we outright don't want things to burn at all (like a wooden house burning down).

Something specific like the Wittig reaction, yeah, people know how that works.

Actual game:

Okay, done with work (that Wednesday deadline is not happening unless I fix a dumb problem fairly early tomorrow) and dinner; time to look at the game.

Right, holkans have negate first strike. So that 1 first strike archers have don't come into play. That means any archer defending on flatland outside a city will have to take Combat I, and Donovan Zoi will still have ~73% odds on one, which he might easily take if it means grabbing a worker stack, and that'd end all resistance. So I should avoid that.

There's also a pretty serious problem, in that if Donovan Zoi knows where my copper is, his holkan can easily beat me to it with his Woodsman II promotion. I was considering moving my first archer to camp the copper, but he can block that. There is so much forest and jungle around, I cannot prevent this even with moving my workers in his way (the idea being he can't capture two in one turn). And it actually hurts me that another forest grew near my capital this turn -- this is almost always a good thing, but not here!

I'm looking at zerg-rushing a holkan camped on a hill with multiple archers.

Okay, so I shouldn't actually have left my warrior in Tiamat, because the holkan otherwise would not have the 5 XP to pull this off. Is this the sort of thing people with early-game rush experience can easily foresee in advance? Because that sure wasn't clear until after the fact!

If he guessed where one of my workers is and grabbed it, that's actually not so bad (oy, I'm in a state where having an early worker capture is actually preferable), because I can definitely get better defense in time then, and I can also get an archer to the copper in time.

[Dark Savant]

Turn 44 (2240 BC)

I think it's increasingly unlikely, but there's a chance this turn will be my last, so I'll post pictures this turn!

Situation before I move:



The holkan moved onto the cow, so I can't use it this turn.  Donovan Zoi actually has time to pillage it before he camps my copper, but I'll move in a worker to repair it if he does this and moves on -- he won't be able to see it there and if does backtrack, I'll beat him to the copper.

Also note the forest that grew in on the hill I spent one turn building a random mine on.

That's far enough away that I can move a worker to help another chop another forest by next turn.  The third has nowhere reasonable to go other than hide in my capital.  Speaking of my capital, here's what it looks like now:



Yeah, I'm just shy of the overflow cap there.

I can produce 2 3-XP archers in the next two turns, and I can have a third chop go into a third archer if I want 3 archers in 3 turns.  I'll have to see if 3 Combat I archers are enough to zerg rush a holkan on a hill with 10% fortify defense.

That's of course assuming Donovan Zoi doesn't just try to end my game on the spot, which he might do.

C&D

I have graphs on Donovan Zoi now, and they don't currently show him with any more holkans than I can see.

I have enough EP to spare to start getting graphs on 2metraninja, actually.

[Dark Savant]

I can't believe I'm calculating the odds of zerg rushing a holkan on a hill with archers, but here I am.

The first archer has ~4% odds of winning outright, and ~15% of losing without inflicting any damage.  If that 15% result happens, the other two archers likely cannot kill it.

If I inflict so much as a single hit of damage, odds are very good the other two archers will pull it off.

I'd get better odds with more archers, of course, but my judgment is that I should not wait.

[Dark Savant]

Nope, I lost the roll to defend my capital. 

Ah well.  What could have I done better about that that I didn't already post?