Let's talk Modern Art
ACK ACK
(Aldie! When you're back up, I'm sending some cash your way.)
So let's talk Modern Art strategy.
I've played about 10 times and I love the game. Last week I was
defeated terribly twice. This week I came in a very-close-to-first
third place and was defeated terribly once. (BTW, I own the game, so
the people I play with are slightly less experienced than I am and I
suck.)
I first bought and played the game several months ago, and very briefly
scanned some of the strategy articles; you'll see some familiar
terminology.
When I first played the game, I naively thought that bidding up to the
estimated value minus $1000 (EV-1) was a good bet because I'd gain
$1000 on the trade. But this is stupid. Why? An example: Let's say I
buy a painting from player X for $28,000 and resell the painting at the
end of the round for $30,000. Okay, not much profit for me, but some.
Better than nothing? Hell no! I made only $2000, but player X made
$28,000 because he or she was dealt that painting for free, and thus
profited entirely with no risk. So, I've discarded this theory of just
making any profit I can on anything. (Yeah, like I said it was a naive
theory.)
So then, if I can buy a painting for the estimated value divided by 2
(EV/2), how's that work? Example: I buy a painting from player X for
$15,000 and resell the painting at the end of the round for $30,000.
Player X made $15,000 and I made $15,000. Okay...getting somewhere.
This was the strategy (although modified a little, in that I actually
tried to sell many more paintings than I bought, overall) I used when I
came in a close 3rd place. This seems to make a lot of sense, but
hasn't put me into the winner's circle quite yet.
(Note that you never WANT to buy exactly at EV-1 or EV/2; you want to
buy for as cheaply as possible. Those are just the values for which
buying may be a good idea.)
After game one--the close finish--of this evening, I realized that if
it is your auction, the EV/2 doesn't make sense. In that case, because
you'll be paying your money to the bank rather than another player,
EV-1 is more reasonable; any profit is better than no profit. Except of
course that the EV is just that...an estimated value, so there is some
speculation involved. And furthermore, after the first or second round,
if you can keep track of who is in the lead, EV-1 may make sense as
long as you buy from someone who is quite a bit further back in the
running than you are. It doesn't really matter if you help out the
straggler, as long as you are helping yourself. There can be only one
winner, after all.
So I tried to implement my new, improved, refined-and-double-timed
strategy in the second game of the night and was destroyed.
Here are my questions:
Are my strategies correct? Or even reasonable?
To the Modern Art players who are more experienced than I am: How much
of the game depends on the luck of what cards are dealt to you?
What strategies am I missing? How do you manage your hand?
Has anyone tried a 6-player variant with this? (All my games have been
5-player, but it seems that it would work with 6, although be even more
chaotic. And it is possible that a player wouldn't get the opportunity
to auction even once in a season.)
Thanks. This post comes out of a love for a game that keeps me thinking
about it long after I've said, "Great game. I'll get you next time. Do
you want to play again?"
ACK ACK
PS: I hope I'm not just dredging up old grievances here. I imagine all
this has been answered before, and if so, I can sit and wait for BGG to
come back online.
Akke
> After game one--the close finish--of this evening, I realized that if
> it is your auction, the EV/2 doesn't make sense. In that case, because
> you'll be paying your money to the bank rather than another player,
> EV-1 is more reasonable; any profit is better than no profit.
In your own auction, EV-1 is much too high, because you'll get a much
higher profit selling at EV/2 to an opponent.
> To the Modern Art players who are more experienced than I am: How much
> of the game depends on the luck of what cards are dealt to you?
I think you need at least a couple of double auction cards to have a
chance to win.
> What strategies am I missing?
Are you using tactics like playing a low cost fixed cost auction of a
artist that you have plenty of cards of, to raise the interest in this
artist?
--
regards,
Akke
ACK ACK
>
> In your own auction, EV-1 is much too high, because you'll get a much
> higher profit selling at EV/2 to an opponent.
>
Hmm...An excellent point, thanks Akke.
But if I can prevent my opponents from gaining anything and myself gain
a little something by bidding up to somewhere between EV-1 and EV/2? An
average of the two: (good gracious) Avg(EV-1)(EV/2)? Hmm...
Much to think about.
Dang, I like this game. It's 3:40 am my time and I'm still thinking
about it.
RichardV
> So then, if I can buy a painting for the estimated value divided by 2
> (EV/2), how's that work? Example: I buy a painting from player X for
> $15,000 and resell the painting at the end of the round for $30,000.
> Player X made $15,000 and I made $15,000.
First, I would say that winning the game is more about selling strategy
then buying strategy. You need to release your paintings in such a way
that you drive up the value of the cards that you are holding -
especially the double auctions.
A better valuation of buying price point is
buy if P < (n EV)/(2n-1), where P is price, n is number of players, and
EV is expected value.
To show this by example: In a 3 player game with EV=20, we buy if P <
60/5, so 12 is price neutral and 11 is a gain. Andrea, Brandon and
Carl all have $100. Andrea sells at 11 (repeat experiment to cover
each opponent). If EV is correct (paintaings pay 20) then Andrea is
$122, Brandon $109, Carl $109 i.e. +13 for Andrea. If Andrea buys at
11, then Andrea is $118, Brandon and Carl are $100 i.e. +18 for Andrea.
At selling price of 12, Andrea gets +16 regardless of whether she or
the opponents buy.
> After game one--the close finish--of this evening, I realized that if
> it is your auction, the EV/2 doesn't make sense. In that case, because
> you'll be paying your money to the bank rather than another player,
> EV-1 is more reasonable; any profit is better than no profit.
No - see the above argument. Another way to think about it is:
case 1 - you buy a painting from player A, they buy a similar value
painting from you for the same price
case 2 - you buy your own painting
In both cases you now own a painting of similar value, but in case 1
you haven't spent any money on it, whereas in case 2 you are out of
pocket!
In general, if you hold more paintings by the same artist you should
definitely NOT buy your own painting. Selling to someone else drives
up the value of your remaining holdings (the buying player will want to
maximise on their existing investment, other players will like to get
into that artist).
> To the Modern Art players who are more experienced than I am: How much
> of the game depends on the luck of what cards are dealt to you?
The double auctions are critical as they effectively give you an extra
turn. My variant (up on the Geek) has 2 different ways to deal the
doubles out fairly. Basically, you keep the doubles in a separate
deck, and deal out an equal number to each player. Control freaks can
deal them all out first round, others can deal out 2 in round 1 to
allow a bit of planning, and then a few more over the remaining rounds.
> Has anyone tried a 6-player variant with this? (All my games have been
> 5-player, but it seems that it would work with 6, although be even more
> chaotic. And it is possible that a player wouldn't get the opportunity
> to auction even once in a season.)
Two 3 player games might be a better option! ;-)
Cheers
Richard Vickery
ACK ACK
I hope I never have to play Modern Art against either of you, because
I'd be doomed!
Richard, how did you derive your P < (nEV)/(2n-1) formula? Experience
of play? Very interesting. Do you think this is a "solved game"?
Still chipper,
ACK ACK
dcorban
concerning EV-1. You will see that ultimately EV-1 IS the best bid, as
long as the auctioneer isn't a threat to win. Theoretically, if the
auctioneer is the current money leader, then the other players should
bid EV/2+1.
thatmarkguy
back-and-forth bidding (in on open auction, it's usually better to pay
the extra dollar, give the auctioneer an extra dollar, and deny the
rival bidder the rest of the profit, than to stand pat and earn nothing
- this is true all the way up to EV-1) Not sure the same logic applies
to open/fixed/oncearound. I'd bet an auctioneer who tries to sell a
fixed at EV-1 finds himself making a profit of 1. And making only a
buck on a selling turn isn't a winning strategy.
The rule of thumb I heard in the 'early days' was a vague target of 2/3
EV. It's a fair balance between giving the auctioneer too much money,
but making it not worth his while to buy from himself and shut other
players out of profit (an auctioneer would seldom buy at much more than
EV/2 from himself).
Then I got to thinking that it should vary with the number of players.
If all players are 'equally' considered threats for the lead (so you
can't assume it's safe to overpay, nor can you assume it's a good idea
for everyone to collude to keep bids low) I think a good vague target
is (n-1)/n. So for each go-around (one turn for everyone), each player
will, on average, have held one auction for
((n-1)/n)*AverageAuctionResaleValue profit and won one auction for
(1/n)*AverageAuctionResaleValue profit.
Habes
so, that I almost prefer the streamlined Trendy, which gets rid of the
auciton portion.
Here's the problem. It's almost always better to outbid the other
players up to EV - 1000.
Say player 1 is selling a painting with an EV of 30,000.
Player 2 bids 25,000.
If he wins, player 1 gets 25,000 and player 2 gets 5,000. You get
nothing.
If you overbid, player 1 get 1,000 more, while player 2 gets 5,000
less, and you gain 4,000.
The game is about manipulating the paitings values, so that the
paintings you buy are worth more than thier EV's and the other players
painting are worth less then thier EV's.
Note that knowing who is winning is also very important. If it's the
guy to your right, then put up a fixed price auction, so he doesn't get
to buy it. If the guy to your left is winning, put up a once around,
so he doesn't get to buy it.
Habes
on him by overbidding the other players. If he was going to make
20,000 anyway, you might as well give him 21,000 if it means more than
a 1,000 gain for you.
dcorban
and not fall into an EV-1 trap set by the current money leader.
Even if every single auction went for EV-1:
1) EV isn't always obvious. Someone is going to lose some serious
money. Players should be playing a wider variety of artists so that
there is at least one artist that does not place.
2) The auctioneer who forces their artists to be worth more will have a
huge advantage. This is where the brunt of the luck and skill factors
come into play.
3) Only open auctions really bring out the EV-1 play. Fixed auctions in
particular are dangerous for the auctioneer. Anything over EV/2 and,
theoretically (from a pure monetary gain perspective), no one should
bid unless the auctioneer is clearly losing.
Migmeister
I love Modern Art. I find it good that you have all those strategies,
and mathematical calculations... But for my taste, I prefer to
concentrate on the way people play, rather than the actual value of a
painting. Must be my poker habit, but when I play, I try to guess, or
find out what people have in their hands. I think that to give you more
chance of winning, maybe your way to calculate value and willingness to
buy are good tools, but for my part, I enjoy the human reaction and the
art of guessing. Although I won around 50% of the Modern Art games I
played, I have a better feeling of finding out stuff, than the actual
winning.
So, I'll try to guess how much someone is willing to pay, who will be
interested in the painting, and then I'll evaluate how much I'm willing
to pay and what auction I might offer that might affect the painting
value. The way people play, normally tells you a lot about what they
want to buy, and what they have to offer.
That's what makes me enjoys the game. Don't be fooled, I love to win,
and I always try to win (my girlfriend calls me the "assassin" because
I'll try to crush as many opponents as possible, whenever we're playing
any game, even party games!! But, I'll be happier gueesing your style
of play, and finding out I'm right, than actually winning the game.
Because, you might do 3 deals that gives you 4,000, but what if I do
one that gives me 60 000? Last time we played, I bought 3 paintings
only in the four rounds (one of them was a double offer) and I won!
Anyway, I loved reading this discussion, but don't ever forget that the
more linear your strategy is, the easier it is for your opponents to
read you. Once your read, in poker, your dead, but in bidding game, it
might become more difficult to win. And if everyone at the table uses
the same strategies, it's not going to be a bidding game anymore, once
someone will put an auction, you'll all know the result, based on the
calculation.
And finally, Ack Ack, you're the perfect example. You might not be
winning many Modern Art game, but you still love the game!! :)
For my taste, and I'm talking about me, not putting anyone down, if to
win at Modern Art, I would have to calculate who's the leader, who's
not, who's making X amount of profit and use a mathematical formula to
calculate my decision of buying.... I would not enjoy it that much. In
some way, I'm probably doing the same thing. Evaluating the way players
play, guessing their willingness to buy, guessing what they might put
up for sale. Some player prefer to buy, others to sell... That's what
makes bidding game, for me, some of the best games to play.
So, happy gaming, I hope you'll find what you're looking for in that
post. I was just happy I could share my love of the bidding games!
RichardV
sales, but also partly to screw the EV-1 guys ;-)
If you are bidding at EV-1, you are locked in to trying to ensure you
get EV or better. This may force you to end the round, which is a
disaster - it is a turn of lost selling, and selling is the surer way
to make money than buying.
My calculation of the buy/sell point is the one at which the auctioneer
makes the same gain relative to the other players if they buy
themselves or if someone else buys.
If you buy your own, you make (EV-P) per sale.
If you sell to others you make a gain relative to them of (P - (EV-P)).
To allow for the number of players we consider you to be able to sell
once to each other player (n-1 sales).
So at the buy/sell point
(n-1)(EV-P) = (n-1)P - (EV-P)
(n-1)EV - nP + P = nP - EV
nEV = (2n-1)P
so P= nEV/(2n-1).
If you sell above this price you do better in immediate gain than if
you buy it yourself. For reasons I outlined in a previous post, you are
often better advised not to buy your own works even above the buy/sell
point.