Recommend
19 
 Thumb up
 Hide
6 Posts

Tichu» Forums » Strategy

Subject: Advanced Strategy in Tichu rss

Your Tags: Add tags
Popular Tags: [View All]
Edward
United States
New York City
flag msg tools
badge
Avatar
mbmbmbmbmb
This article assumes that you have already mastered the tenets of Aaron Fuegi's excellent article on Tichu strategy.

Fundamentals

Every potential play in your hand can be measured along two probability axes:

* Probability that you will have a chance to play it while not on the lead (let's call this the play probability)

* Probability that if you play it, it will win the trick (let's call this the win probability)

Your cards can therefore be divided into four categories:

Losers: cards with low play probability and low win probability. Example: low singles (like a 2 or 4); low pairs (22, 44).

Sloughs: cards with high play probability and low win probability. Example: middle singles (like a 9).

Nulls: cards with low play probability and high win probability. Example: long straights.

Winners: cards with high play probability and high win probability. Example: Aces, Dragon, bombs.

Note that cards will shift between these categories. Kings normally begin the game as sloughs, but are promoted to winners after the Aces and Dragon (and sometimes Phoenix) are out. Low pairs, which are normally losers, may become a null in the endgame. The Dog is a loser that is sometimes a null.

Low singles and pairs tend to be losers. Low full houses and triplets are often losers as well, because the nature of the card distribution means that if you have a full house others are likely to have full houses as well. Complex sets like consecutive pairs and straights, no matter their value, tend to be nulls.

Among singles, the Dragon and the Aces are winners, with the Phoenix being more unreliable. High pairs and triplets are also winners, but you rarely want to play AAA as that is turning three potential winners into one.

Tichu Calls

What you need for the Tichu call is at least one winner (or start with the Mah Jong), and subsequently at least one winner for each loser (although you can exit with a loser if you do not expect to be bombed). Generally, it does not matter how many nulls you have.

How risky a Tichu call is depends on how many sloughs you have, and whether your nulls really are nulls. In some games you will be able to get away with sloughing off multiple singles on a trick, or playing 34567 and seeing it hold. In others, your opponents will King and Ace every single you play, and even a ten-card straight might get played on. This is where your risk appetite (and current score) comes into play.

More rarely, a Tichu call might be risky if you are unable to play your winners: the only guaranteed winner is a bomb, and if your opponents have a lot of complex sets they may exit before you can get the lead back with an Ace. Even the most rank novice can appreciate the power of the Phoenix, but the Dragon all-but-guarantees its owner at least one lead, and if you don't have any losers or sloughs, one lead is all you need.

As a defender against a Tichu call, your goal is to slough off cards on their leads while hopefully blocking their sloughs. In addition, hope that some of your nulls can be converted into winners if the Tichu-caller plays a complex set he believes to be a null. Against a true Tichu hand (a winner for every loser), you have no real chance of stopping them. Most people, however, are not that conservative with their Tichu calls (and if they are, you should be beating them handily).

If your partner calls Tichu, you should avoid playing winners unless your partner indicates an inability to win the trick. Even sloughing is discouraged after a certain point (around 9) because there is so little room to play at that level, and there is a high chance you are blocking his sloughs. Assuming you do not have any winners (you should have passed them to your partner!), your primary goal is to maintain as many types of nulls as possible, so that if your opponents get in you are hopefully able to stop them with your now-winners. Return the lead to your partner with the Dog, or a loser that they have a winner for (usually a single, but sometimes a pair depending on what your partner has led).

Grand Tichu

I will call Grand Tichu whenever I have at least 3 Aces, Ace-Phoenix, Ace-Dragon, or Dragon-Phoenix in my initial eight. If you follow this algorithm, you will not make every Grand Tichu you call; in fact, you'll miss about 30% of them. But if you make every Grand Tichu you call, then you are calling too few Grand Tichus. In every hand, someone will exit first. If you fail to capitalize on your GT-worthy hands, you will quickly fall behind opponents that are consistently making an extra 100-200 points when they are dealt the strong hands.

Passing

A common passing dilemma is whether to make the risky pass (for the straight) or the conservative pass (for the pairs and triplets). In general, straights are preferable to pairs, because 33 is usually as much a loser as 3, whereas 23456 can convert 5 losers into a null. The main consideration is how much benefit you derive from taking the risk, and how much risk you are taking (double-ended draws like holding 3456 are naturally much more likely than gutshot draws, where you need one specific card).

For example, in the following hand:

Dog 334566 89TTKKA

I would pass 3 6 A. Getting that 2 or 7 will instantly convert quite a few losers into nulls, and you might be able to threaten a quick exit if you can get in with your KK. If you don't get it, your hand is not that much worse off than before, and in any event certainly not a Tichu-threatening or Tichu-denying hand.

But with this hand:

234455 789QQKKA

I would pass 3/2/A. Your 4455 is probably a loser but one that can be covered by QQKK: if you can get a lead, QQKK goes from a null to a loser-covering winner. (If you are playing against opponents that you know often pass low cards, I would pass 7 8 A instead.) Passing 5 4 A is tempting, but turns a somewhat-decent defensive hand into utter junk if you don't get that 6.

The Dragon

It is important for both partners to immediately identify if they have the Dragon. Once it is no longer a threat (i.e., it has been played or the partnership holds it) then four Aces are all promoted to winners.

When I must give away the Dragon, I give it to the opponent that has demonstrated the least strength thus far (under the assumption that Tichu players try to concentrate their strength) or the opponent that plays before me. If you are fortunate to have both a bomb and the Dragon, look to bomb the Dragon while still being able to exit.

The Phoenix

An unreliable winner, more useful as a way to convert low losers into a null. If you cannot make use of the Phoenix and do not plan to call Tichu, pass it to your partner. Nevertheless, it is the most powerful card in the game and the card I most like to see in my hand. It essentially eliminates any passing dilemmas, because either way you pass, the Phoenix will likely be able to repair your hand.

The Mah Jong

It is a loser, but still a better card to pass to your partner than a Q (assuming you have no strength and are looking to pass a good card to your partner). You are passing a guaranteed-slough that may link into a low straight. It is probably worse than a K because K's may eventually become a winner. I will also pass it to my partner if he is sitting before someone who called GT, in order to wish out an Ace (or 5 or 6, in order to break up a low straight). Wishing out that Ace may stop the GT caller from a critical slough.

I will not pass this to an opponent, for the reasons listed above.

The Dog

I pass this to any opponent that calls GT. I also pass it to an opponent that I know from experience is aggressive (or conversely, to the partner of an opponent that I know to be passive). I will sometimes pass it if my hand is so lousy that I have no hope of ever making a winner. At all other times, I will keep it or pass it to my partner.

Card Play

The order of play from your hand should be: losers -> sloughs -> nulls -> winners. If you do not have the lead, you can usually exit so long as you have one winner for each loser. For example, if you are holding 23AA and the Dragon/Phoenix are out, you should be able to exit on the next single that is played (barring any bombs). If you do have the lead, you can afford to have one extra loser.

If you do not meet these criteria, then it is often in your interest to let an opponent go out first, so that you can gain some kind of free lead or promotion. For example, if you hold 23A and sit after an opponent who called GT, do not waste your A trying to stop them unless you have some reason to suspect that they seriously misplayed their hand (e.g., they were bombed out or unexpectedly lost a trick). Better to allow them to exit. A conservative opponent will exit with a winner (the most bomb-proof way to make the GT), and therefore you will hopefully gain the lead for free.

Likewise, if you are holding 23KK in the same situation, you are probably best off waiting for the Aces to go out so as to promote your Kings. Against an opponent that called GT, they likely hold all the Aces in their partnership, and if you let him make his GT, your K winners will be able to salvage whatever points remain and help you exit 2-3 with your partner.

Sometimes you will have no way to exit no matter how you play (e.g., you are still holding the Dog, with no winners, and your partner and the opponent after you have both exited). In that case, assuming your partner went out first, your only goal is to take as many points as possible. It is OK if you strand yourself with nothing but losers so long as you get as many points as possible into any trick that you do win.

In general, do not take the lead from your partner. This includes overplaying him (with the rare exception of overplaying his Kings to block an opponent's Ace). Of course, his 7 is unlikely to hold up, and you can slough a 9 onto it, but do not overplay his Q with your K, or his A with your Dragon! Your winners are best used to steal or deny the lead from your opponents, not each other. The exception is when you know that you can exit, because you have nothing left but nulls and winners. Even then, you should hope that your partner can exit first, and then you can take the lead soon thereafter to go out 1-2.

If one of the opponents have already exited, then the analysis shifts because the paramount goal is to avoid being swept 1-2. If you play before your partner (i.e., the opponent after you is the one that exited), you are the one that is implicitly designated to exit. Your partner should be passing on your plays, content in the knowledge that if the opponent overplays you and you cannot respond, your partner can still try to take the trick. In addition, you can now exit with a winner, and allow your partner the free lead. On the other hand, if you play after your partner, you must try to exit with a loser (leaving you slightly vulnerable to a bomb) to avoid handing your opponent a free lead.
15 
 Thumb up
1.26
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
If you call Tichu, planning to play your 9 for free, I think you might be disappointed a lot. Your RHO just has to keep going up, in front of you. Often, he doesn't really have to think about going out.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Aaron Fuegi
United States
Framingham
MA
flag msg tools
Avatar
mbmbmbmb
DaviddesJ wrote:
If you call Tichu, planning to play your 9 for free, I think you might be disappointed a lot. Your RHO just has to keep going up, in front of you. Often, he doesn't really have to think about going out.


He somewhat addresses this point:

Quote:
How risky a Tichu call is depends on how many sloughs you have, and whether your nulls really are nulls. In some games you will be able to get away with sloughing off multiple singles on a trick, or playing 34567 and seeing it hold. In others, your opponents will King and Ace every single you play, and even a ten-card straight might get played on.


Whether the base rank should be 9 or J or whatever is debatable but the point is valid - a 9 has a reasonable chance of getting sloughed while a 3 is much less likely to be. And there is a 50% (or so; actually somewhat less when the opponents have reasonably balanced hands which does happen a reasonable amount of the time) chance that your RHO is the one who is a real threat to you and so he can't afford to just block.

Saying when the Dragon is gone, Aces are straight winners is probably the thing in the post I'd most point out as wrong. I also pass the Dog to opponents less than seems to be suggested.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Curt Carpenter
United States
Kirkland
Washington
flag msg tools
badge
Avatar
mbmbmbmbmb
Nice article. It seems as reasonable as anything else I can recall seeing. If it were me I would probably make some mention of how thinking about points factors into a strategy. And an overall strategy for each hand. The mistake I see many new players make is simply trying to go out, oblivious to everything else. But consideration needs to be made to what realistic goal you and your partner can achieve on the hand, and work toward that goal. The game is won/lost at the margins.

theory wrote:
the only guaranteed winner is a bomb

Believing this leads to disappointment.

theory wrote:
In every hand, someone will exit first.

But it's not an invariant based on cards dealt. Making calls will affect the play of others.

theory wrote:
The Dog
I pass this to any opponent that calls GT

I put a bit more thought into it. The worst possible outcome is getting 1-2'd. Passing the dog to GT caller increases the chances of that disaster. If the goal is to set the GT caller (which may sound obvious, but sometimes I look at my hands and am happy if we merely avoid getting 1-2'd), then either you or your partner will have to do it. If you think you have a good chance, the dog at your partner will help you more. Especially since the GT caller may assume his partner has the dog. If you gave your partner a power card, then keeping the dog for yourself might be best. And if you have a terrible hand, and don't think you can even dog to your P if you kept the dog, then maybe passing to GT's partner is best, just to try to avoid the 1-2 (also dependent on current score). Almost all Tichu best practices are contextual. But the first step is almost always to consider the realistic goal for the hand based on what you know and go from there.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Gerald Katz
United States
New York
flag msg tools
Avatar
mbmbmbmbmb
I never pass the Dog to the opponents, even to the one who called Grand. He'll play it, and his opponent will just lead a low singleton since it's likely his partner has the Dragon. Grand player might hold onto the Dog has his last card to try for Grand 1-2. He also might hold it for the middle of the round once he finally got rid of all his sure losers and doesn't care what partner leads, hopefully also setting up for Grand 1-2. Even if there's no Grand or normal Tichu call, the Dog just transfers the lead, The opponent are still in control.

If my hand is garbage or mediocre, I keep the Dog. I'm not going out first anyway. Maybe partner can. If I win a trick, I give my partner the lead to help him. If I have a great hand give partner the Dog to help me go out first because it's not in my hand and he can give me back the lead if necessary. Sometimes I have a pat hand. I'll go out or threaten to if I get a lead, but I can't get a lead on my own. Partner playing the Dog gives me the lead I need.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
I sometimes pass the Dog to an opponent if there aren't two small cards I want to give them, and my hand is "too good" to want to keep the Dog in it, but "too weak" to give the Dog to my partner (and I have something better to give him).
2 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.