Statements like Dave’s always amuse me.
If two players play (close to)equally well, then of course the random factors will decide which wins!  If there is no luck and two players play equally skillfully the game would be a draw.

I played a two player 7 Ages with a friend for several hours and he won by a handful of points.  When I said something along the lines of, “ten hours of gaming and the winner was determined by a card draw at the end,” he intially took it to mean I was objecting.  I explained I thought that my statement should be interpretted to mean that we’d both played well enough with respect to the other player that either could win depending on the result of the last battle and we both could feel good about how we’d played.

I object to luck if as a result of it, a player is put out of the game early.  For example, if the (randomly chosed) starting position is so bad that one cannot make up the difference.

Luck can also be objectionable if there are two dice rolls/card draws/etc. in the whole game that distribute hald of the victory points.  Two times is not enough to average out the luck and may not leave other avenues to mitigate the results.

“Luck” in the form of another player throwing the game, intentionally or otherwise bothers me some.  Think Puerto Rico and the newbie always picking Craftsman.  You certainly want to be in the chair to the left of him.

Luck in the final draw a la Killer Bunnies doesn’t bother me at all.  (There are a myriad of other reasons that I won’t play that game, but the end draw isn’t a contributor to my “boycott.")

Luck of the draw in a game like Vikings doesn’t bother me either, I feel that there are ample decisions to be made that can mitigate an unfavorable setup.

It’s spelled Svea Rike.

Good points and questions, Kris.  For me, the determination is based on whether I feel I’m in control of my fate.  Let’s take some very random games first:  Verflixxt! (That’s Life!), Diamant (Incan Gold), and Button Men.

Verflixxt! is a roll-and-move game.  Consistently unfortunate rolls can kill you; however, you choose which of your three pawns or the visited guards will use that roll each turn.  Because I’m deciding what to do every turn - even if it is deciding how to use a bad roll - I generally find this an acceptable application of luck, even if it is per-player.

Diamant boils the players’ choices down even further:  Each turn, you decide whether to progress deeper, or run back home to keep your findings (and perhaps pick up leftovers along the way).  The luck of the draw kills everyone equally (except those that made the choice to leave), so again, acceptable.  I’d rather play something with deeper implications to my decisions, but this is a fine diversion.

Button Men is entirely dice, and some turns are choice-free (there is one attack I can make, and I must make it) - something that usually disgusts me in a game, but nearly every turn involves odd decisions of which di(c)e with which value(s) will I use to capture which of my opponent’s di(c)e.  The original rules are elegant, but what really makes the game interesting for me is the CCG-style interaction of various die abilities in all the expansions.  In case you couldn’t tell, I like this one.  Even though every turn is rolling dice, it’s typically die rolls by choice, which imparts to the player some sense of control.

Now let’s look at LCR:  You roll a die, then do what it says.  Couldn’t be much simpler, but there are zero decisions for the player to make - not even superficial ones.  You don’t even get a bright colorful board to look at like you do in Candyland (another zero-decision game).  Bzzzt.

Now how about Die Macher?  Brilliantly designed game.  There are aspects that are purely card driven, and a relatively new player may not understand the more subtle ways of manipulating the other players to bail them out of consistent card problems (i.e. never having whichever issues happen to be favorable at the right time).  Normally, this wouldn’t bother me, because you simply play again; however, when we’re talking about a five-hour game, that doesn’t seem like such an acceptable answer to me.  So, though Die Macher is overflowing with decisions, I find the amount of luck in it damaging, due to its length.  I still acknowledge its brilliance, and I can enthusiastically recommend it to anyone who considers five hours a trifle.

So, to the topical title:  The Pillars of the Earth.  Yes, the bag of brutality has bruised and butchered me in the game, but the bag represents a gamble.  The worst it treated me was in round 6/6 of a close game.  I’d taken a slight gamble, because I could flex any of several actions to boost myself to a probable win.  ...that was assuming that at least one of my builders would come out of the bag in the first eight.  As it turned out, my builders were three of the last four.  Wipeout.  However, I chose to take a (statistically reasonable) risk, rather than some guaranteed path with a lesser potential reward.  That other path probably wouldn’t have won me the game, but I certainly wouldn’t have been in a shamefully distant last place.  Yes, I felt abused by the bag, but I knew I’d chosen to depend on the randomizer, rather than act conservatively.  And it’s 90 minutes, instead of 300 minutes.  If that had happened to me all six rounds (which is theoretically possible), I would have been pretty disgusted.  As with the Die Macher “flaw”, though, an utter, constant trampling is pretty darn rare.  Were Die Macher 1/3 its playtime, I’m sure it would entice me as easily as Pillars.

Finally, I’d like to mention the family of games that include Arkham Horror and Betrayal:  There is a very heavy dose of luck to provide the surprising, suspenseful experience the game is intended to deliver.  Since the key to enjoying these games is letting go of analytical thinking, and immersing yourself in the game’s story, you actually kind of depend on the randomness to make the game interesting.  These games would be nothing but a storytelling exercise without something stochastic in nature.  I generally disprefer these games, because my fate is more chosen by the game engine than my own choices - by which I mean that my decisions have little foundation and it’s a crapshoot whether they will pan out, but I absolutely wouldn’t play these games if there were no luck to drive them.  I guess these are analogous to the Choose Your Own Adventure class of books:  Yes, you make choices, but you might make a conservative choice only find yourself reading exactly how that means you died.  Where Arkham and Lord of the Rings succeed is by giving the players tools to “cheat” and try to alter their circumstances and/or die rolls.  This encourages the collaboration these suspense games try to foster.

Right, well, enough blather for now....

That’s pretty lame to say a game has too much luck becuase the second tie-breaker is a random draw.

Galaxy Trucker has a lot of luck but I think people find that OK because you knew what you were signing up for when you started the game.  Most of the fun is watching half your ship fall off due to a unlucky laser shot from the pirates.  You can mitigate the luck somewhat, but nobody ever builds a perfect ship, so you can’t eliminate the luck.

Hey, I loved Uncle Wiggly!  Then again I think I was only four and I didn’t even know what luck was so it didn’t bother me.

These days, now that I’m older and crankier, I find I can tolerate various amounts of luck in a game only so long as there is some rational way to plan at least somewhat for the various possible outcomes.  A totally unforeseen evil fist of doom striking the player down so that there was never anyway to have won the game is the problem.  I also concur with other comments about how much time is invested, and on how much the luck is shared or distributed (many small doses rather than one large fatal one).

Some people say acceptable luck is tied to the length of the game. I always like to say that the amount of luck should be linked to the impact of other strategic and tactical choices you have to make. If these are hard and should put you up for win if you do better at them than other games, than the luck factor shouldn’t undo that. So if there is a high luck factor, the strategic and tactical choices would have to be of smaller importance too, as making a really good decision and seeing it undone by bad luck is not good.

Well said, Surya.  The same is actually true for other “simple” factors than luck:  In the Babylon 5 CCG (for example), some players find it extremely frustrating to spend significant time plotting, scheming, and maneuvering only to have their plans undone by something simple.  This is frustrating in any game, whether that simple thing is a roll of a die, a card draw, or the play of a simple, cheap spell-canceling card in Magic: The Gathering that undoes a play with a lot of conditions.  (Fortunately, Magic’s development team tends to avoid overly complex requirements.)

I also agree with Surya’s comment.  One of biggest pet peeves is a game that combines deep planning with a high luck element.  The former entices you to carefully work out your strategy, but the latter often makes all that planning pointless.  It’s the reason I rate Wallenstein so low (and I actually have more problems with the random way that the actions come out than with the cube tower).

I have not played Wallenstein, but in Shogun I have found the opposite is true: the luck is undone by good strategy. If you stick to building and defending what you have, chances are you have a very good chance at winning. Al least in the games I played.