Zeebo's Theorem - Nobody ever folded a full house.
Reasoning: Nobody is good enough to fold a monster. Most players aren't even good enough to fold a hand that looks like a monster but really isn't.
Application: There are two basic applications to this theory. The first is that if you put your opponent on a full house and you can beat them, don't be afraid to overbet/push the river. This is particularly true when there is three of a kind on the board. Players will call with an incredible range of full houses in that spot. It is true that some villain may fold 22 on a board with three aces. However, you have no way of knowing if they have 22 or TT so go ahead and felt them. You are losing value if you don't. And sometimes they'll call with 22 anyway.
The second thing to realize is to never try to bluff anyone off a full house. If you have 22 on a board with three Aces, don't expect to be able to push 66 off his hand.
This theorem also generally applies to any monster over monster situation, from straight flush over quads/FH/nut flush down to set over set.
Reliability: This is the most reliable theorem. Nearly 100%. Somebody will post and argue that it is actually 100%.
Reference:
http://captzeebo.supok.net/2006-02-12.html
Clarkmeister's Theorem - When you are OOP HU on the river and a 4-flush comes always bet.
Reasoning: Simply put, a 4-flush is an ideal bluffing situation.
Application: Bet a lot of 4-flushes, particularly HU, OOP on the river. You will get a ton of folds. Most everyone is folding non-flush hands (that beat you) and small flushes.
Reliability: Yes, sometimes villain has the nut flush or calls with the K-high flush. Nothing you can do there. But over the long haul this is a VERY profitable spot to bluff.
Keep in mind though, you ARE turning your hand into a bluff. If you have a hand you don't want to turn into a bluff (very villain dependent) like top set or the K-high flush then check/calling can be fine.
BelugaWhale Theorem - When you are the preflop raiser and your turn bet is raised or check/raise, it is time to re-evaluate one pair hands.
Reasoning: In raised pots, most players will just call down with one pair (be it pocket pair or top pair) type hands as well as draws. The turn is where most players who flopped a monster stop slowplaying and try to build pot. Or, they raise if they hit their draw.
Application: A raise on the turn is a signal to re-evaluate where you are at. It is not and automatic fold but you need to consider if villain has a monster or just hit his draw.
Reliability: Against fish and bad players in general, with the exception sometimes of LAGs and maniacs, this is a VERY reliably theorem. However, it is also an extremely popular and well known theorem, perhaps the best known. A lot of good players, particularly 2+2 players can try to exploit this theorem, especially by floating. So depending on the player (a decent player who is ALSO capable of making a play) you may need to discount this theorem considerably.
Yeti Theorem - A flop three bet on a dry (preferably paired) board is always a bluff.
Reasoning: On a paired (or otherwise very dry flop) a player with an overpair is unlikely to want to stack off because usually the only hand he gets action from is a monster that crushes him. For this reason, someone who DOES have a monster usually will usually slowplay here. Since neither strong hands like top pair and overpairs don't 3bet here and monsters don't 3bet here the only hands left that 3bet are bluffs.
Application: If someone 3bets you in this situation, 4bet/push.
Reliability: These days this is mostly considered a joke theorem or a sarcastic excuse for spew. This is because Internet games between regulars are so much more aggressive than pre-Internet (ie live) games. For many players, "fast play is the new slow play" so players will stack off in these spots both with overpairs AND with monsters. Aggression is often used to conceal hand strength as much or more as slowplaying is used.
Against some players (ABC TAG) this theorem does still hold merit however.
