Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 5:57 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by Redward

Macrawn wrote:

I don't know if it has been "solved" but changing coal to coke, then shipping it (or going a step further and making it steel) is just such a money factory that it seems so much harder to try and win any other way. Owning the colliery also gives a good leg up because everyone needs to use that building and it costs 2 food. It makes the game feel narrow and solved. I personally would like to see a second edition of the game open up some more avenues for victory.

The existence of a strategy bottleneck is not proof of the game being solved in any technical sense but is a valid complaint nonetheless. To some degree or another it exist at here and i agree that a new edition or an expansion that opens up options would be cool.

---

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 6:07 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by TheJRMY

jamuki wrote:

What does they mean with "solved"?

"Solved" is a term we take from game theory. Any game that involves randomness (Le Havre, for example) is innately unsolvable. Solving a game requires the construction of a decision tree. Then you can follow the paths to achieve the desired outcome. Each side has a best possible choice at any given moment. Once these best choices have been determined, the game is solved. At the beginning of a game, assuming both sides are going to make the best possible choices, you know who is going to win and in what way.

Examples of solved games include Connect 4, Tic Tac Toe, and Hive. Games like Go and Chess are solvable, though the number of choices possible at any point are well beyond anything we can calculate, even by our most complex computers.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 11:01 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by davypi

TheJRMY wrote:

Any game that involves randomness (Le Havre, for example) is innately unsolvable.

I have to mildly disagree with this. Nim is a game that can have a random setup. You can change the number of piles and how many tokens are in each pile. You can also change the victory condition from taking the last token to forcing your opponent to take last. I have also seen variants where players are limited to how many tokens they can take. You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable. In theory, Le Havre could be solved if the seven resource tokens were known from the first turn and the special buildings pile was public knowledge (or removed).

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 11:06 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by turbothy

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 11:14 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by davypi

Macrawn wrote:

Owning the colliery also gives a good leg up because everyone needs to use that building and it costs 2 food. It makes the game feel narrow and solved.

With respect, I think its claims like this that demonstrate why Le Havre is not a solved game. I have played games where players were able to acquire more energy by picking up the neglected wood offer and converting to charcoal than they would going the coal/coke route. Similarly, with four or five players, it is possible to win the game through buildings instead of shipping. Admittedly, it does require that the other players are continually cutting each other out of the key locations needed for big shipping. You need some experience with the game to recognize when these opportunities are open, but coal/coke/steel is not the only game in town. I have to agree though that it is unfortunate that these other avenues are too often the exception rather than the norm.

I personally would like to see a second edition of the game open up some more avenues for victory.

At one point there was discussion of releasing another expansion for Le Havre that would replace the standard buildings. The idea would be that you have two ways to play, much like the France/Ireland choice you have when playing Ora. Unfortunately this never came to light and, given how old the game is, I'm skeptical that it ever will.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 17, 2015, 11:15 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by golden_cow2

Le Havre is random. The restocking tokens and the special buildings are not open information. Players have to make game-affecting decisions before they know what game they're really playing, so to speak, and the optimal moves could be different for each.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 6:40 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by RutgersKev

In other worker placement games I've played if one person does strategy A and one person does strategy B the scores are close and it really comes down to one or two times where the loser does not take an optimal move.

In Le Havre, it seems, that if one person plays strategy A (big shipping) and one person plays Strategy B (anything else) Player A is going to win even after making one or two mistakes during the game. And in my opinion, although I love the game, it knocks it down a few spots behind some of the other newer worker placements.

Now, the argument is, everyone knows Strategy A is the best, so therefore the skill in the game is playing Strategy A better than the others.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 10:16 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by TheJRMY

davypi wrote:

TheJRMY wrote:

Any game that involves randomness (Le Havre, for example) is innately unsolvable.

I have to mildly disagree with this. Nim is a game that can have a random setup.

I assume once setup occurs information is open. As such, only the setup is random. I should have said randomness in gameplay.

---

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 10:17 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by TheJRMY

turbothy wrote:

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

And you cannot solve each separately, as the information randomly determined at setup (buildings in each stack) is not open information.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 10:43 am

≫ Next: Reply: Le Havre:: Reviews:: Re: A Landmark Game

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by grant5

TheJRMY wrote:

turbothy wrote:

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

And you cannot solve each separately, as the information randomly determined at setup (buildings in each stack) is not open information.

I hope you don't mean that the order of the buildings after setup is hidden, because that is wrong. That's probably not what you're saying, but it sounds like that's what you're saying.

Reply: Le Havre:: Reviews:: Re: A Landmark Game

August 18, 2015, 3:20 pm

≫ Next: Reply: Le Havre:: Reviews:: Re: A Landmark Game

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by RetroBeliever

To state it succinctly, this was one of the most thorough and eloquently written reviews I have ever had the privilege of reading. I felt smarter just reading your review, as if my mother had grasped me by the ankles and immersed my head into the waters of board game erudition. Moreover, your review has served to further convince me that Le Havre belongs near the top of my "Next to Buy" list.

Reply: Le Havre:: Reviews:: Re: A Landmark Game

August 18, 2015, 5:24 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Reviews:: Re: A Landmark Game

$

0

0

by JohnRayJr

Thanks!

It's good to hear that people still find this review and consider it a worthwhile read, several years later :)

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 6:59 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Reviews:: Re: A Landmark Game

$

0

0

by TheJRMY

grant5 wrote:

TheJRMY wrote:

turbothy wrote:

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

And you cannot solve each separately, as the information randomly determined at setup (buildings in each stack) is not open information.

I hope you don't mean that the order of the buildings after setup is hidden, because that is wrong. That's probably not what you're saying, but it sounds like that's what you're saying.

Then I have been playing this game wrong.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 18, 2015, 7:41 pm

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by grant5

TheJRMY wrote:

grant5 wrote:

TheJRMY wrote:

turbothy wrote:

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

And you cannot solve each separately, as the information randomly determined at setup (buildings in each stack) is not open information.

I hope you don't mean that the order of the buildings after setup is hidden, because that is wrong. That's probably not what you're saying, but it sounds like that's what you're saying.

Then I have been playing this game wrong.

Oh yeah, this is a critical part of the game!
-Shuffle all the building cards being used
-Split into 3 stacks.
-Organize each stack by the tiny number in the corner, smallest on top.
-Splay each stack so that the bottom row of each card, which shows the building name and material cost, is visible.
-Any player may examine any card in any stack.
-Now plan your strategy for the game based on what you see in the stacks.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 12:36 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by TheJRMY

grant5 wrote:

TheJRMY wrote:

grant5 wrote:

TheJRMY wrote:

turbothy wrote:

davypi wrote:

You can randomize all of these conditions, yet algorithms exist that solve the game for any of these given states. I think the correct claim here is that any game with open information is solvable.

Correct. Le Havre isn't random, it just has a large number of opening states - which makes it harder to solve of course, since you'd have to solve each of them separately.

And you cannot solve each separately, as the information randomly determined at setup (buildings in each stack) is not open information.

I hope you don't mean that the order of the buildings after setup is hidden, because that is wrong. That's probably not what you're saying, but it sounds like that's what you're saying.

Then I have been playing this game wrong.

Oh yeah, this is a critical part of the game!
-Shuffle all the building cards being used
-Split into 3 stacks.
-Organize each stack by the tiny number in the corner, smallest on top.
-Splay each stack so that the bottom row of each card, which shows the building name and material cost, is visible.
-Any player may examine any card in any stack.
-Now plan your strategy for the game based on what you see in the stacks.

Wow. Yeah, that changes everything. So the game could be solved for each setup. That would, however, be a very complex solution. No one is going to memorize it, unlike Connect 4 or Tic Tac Toe.

---

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 2:29 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by turbothy

You still have the initially unknown distribution of the 7 offer tiles and the hidden special buildings. So I have to agree with golden_cow2 that it's not actually solvable, because even if you were to create decision trees for all the billions permutations of the initial state, the players won't know which decision tree to choose from the start because there is hidden information.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 4:56 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by Ponton

The real hidden information are the special buildings. The offer tiles are not that hidden. You know exactly how they look like, you just don't know the order they come into play. For the first turn that's 720 possible permutations (one is face up, six are not => 6!=720 permutations).

Edit: Of course, you could argue similarly for the special buildings. We know which ones are out there (like short of 120, right?) and there are 6 in the stack, in order, resulting in (120 choose 6)*6! combinations.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 6:10 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by turbothy

Right, but you still can't choose the correct decision tree because the information is hidden from the start. Even if you were to solve all permutations, you can only guess as to which is the correct choice.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 6:12 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by grant5

People. Come on. The game is not solvable, even disregarding the hidden information in the special buildings. For one, there is far to much variability that enters the game based on the actions players take throughout the game.

Second, and most important, during the first round all players are making decisions based on incomplete information due to the replenishment tiles not all being flipped up yet. So you couldn't even try to calculate those permutations, because you could never predict the player actions during that first round, since information is still hidden when the decisions take place.

So even from a theoretical standpoint (i.e. could a computer do it even though a human never could), the game is not solvable.

Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

August 19, 2015, 6:57 am

≫ Next: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

≪ Previous: Reply: Le Havre:: Strategy:: Re: "The Game's Been Solved"?

$

0

0

by steveg700

So, it's not "solved" as game theory would define the word.

However, since the consensus is that there's a dominant strategy, the game may "feel" solved. While it's argued that the game doesn't suffer for having a dominant strategy, because amongst experienced players it's a competition to execute that strategy better than others, the result is that gameplay can feel highly formulaic since all players are pruning off the same branches from their decision tree in every game. There's no exploration of the game, only a refinement of a proven stratagem.