Archive for Tagpoker Forum A friendly place for winning poker players to share their strategy to making money playing poker.          Tagpoker Forum Forum Index -> Heads Up Poker Strategy rekop To Shove or not to shove - a quick method? One of the experienced members of the forum commented in a recent thread that "There is no such thing as a coin flip". This statement got me thinking (indirectly) if there was a way to calculate whether a shove or calling a shove is profitable pre flop while sitting at the table without the years of experience and hours number crunching required in Pokerstove. Ok, sounds to good to be true, and it probably is, but here it goes anyway: (Firstly it is worth reading red dog's post which is stickied in this forum which explains what EV is and how it is calculated.) Now back to the coin flip senario - what he was saying (Nacht I think it was) was that if for example you call an All In and your cards are 50:50 to win or lose then it is +EV. Why?  Because the 50 per cent of times you lose, you lose only your call of the All In but the 50% of times ou win, you win the villain's All-In PLUS the blinds or whatever what is the pot already. I added that, at least with calling an All In there is also a % of the time the villain is bluffing with air. Conclusion: If somehow you knew you were a 50:50 to win, you should go for it. Now the way to solve the shove or not shove question, awaying from the table, is to estimate a range of hands you think the villain has and pit each of them against your exact 2 cards and get an average - you then know what your equity, or chances to win, is. This then has to be compared against what you can potentially win or lose (the pot odds).  Pokerstove does this for you. Red dog gave us the formula to understand how it is calculated Quote: Overall Equity = (odds of no one calling * stack increase when no one calls) + (odds of someone calling * odds of you winning vs opponents calling range * stack increase when called and you win) - (odds of someone calling * odds of you losing vs opponents calling range * stack decrease when called and you lose) For those of you who are frozen with dreaded flasbacks to school, what this formula is simply saying is your EV is: All the money in the pot already because he folds to your shove PLUS all the loads more money in the pot when he calls but you win (or suck out!!) MINUS your money you shoved with, everytime he calls and wins. So here's my idea: 1) Create or find a list of all 2000+ starting hands ranked in order. It would begin with AA and end in 23o. (not 27o interestingly as this is HUs - and the 7 high will win more than a 3 high on unpaired boards.) Off the top of my head I think this could be done by pitting every single hand combination against a range of every other combination out there and the putting them in order. 2) Print it out and stick it onto your wall beside your computer. (It may work out that there are more than one hand at a particular rank) 3)Find out what range of hands the villain needs to have the coin flip situation. So you have K7s which is ranked at say position 100. Because he could have any two cards at the beginning (including AA) you then need to go down to position 200 and see what is in that position. (In other words you put K7o as the centre, or average of his range.) Say number 200 is j7o. 4) Based on the player type if you think J7o is the bottom of his real range to call then shove. If his range could be lower then it is even better to shove. 5) For times when, for example J7o is definitely not in his range (maybe his range ends at position 180, then only experience can help. A shove here, while losing in equity of showdown, can win if the pot is already juicy (higher blinds to all in ratio) 6) When calling a shove remember that the bottom of his range using the rankings if even more profitable due to the bluff factor which Dan Harrigton estimated at a ball park figure of 10% if you have no read on your opponent. Well that's it in a nutshell, thanks for your patience if you managed to read this far and apologies in advance for wasting your time if it is completely flawd! nachtwacht flaws in your idea: 1) There are 1326 starting hands but only 169 in reality because for example, AhAc has the same value as AdAs. You would indeed have to put every one of those 169 into pokerstove (great, we can't do without, you will have to learn how to use it and pit that against different ranges for tight players and loose players and anything in the middle. That is a lot of work.... 2) You would indeed need a wall 3) We are not looking for "coin flip situations". Here is your biggest flaw... there is a very big difference when blinds are at 10-20 or when they are at 100-200. At 100-200 there is a lot more money in the pot so you would need way less than your "coinflip". This means we would need to make lists for not only different opponents, but also for different blinds levels.... and also.... for different effective stack sizes.... We will need more than just a single wall I am afraid 4/5) ranges are never static. You might be facing a tight opponent, but at the moment you start shoving, after 5 shoves, his range will change. He might not call with J7o the first hand, but he might the 5th time you shove. It is very good that you are thinking but I am afraid you have come up with something very complicated stuff. Do some reading for the following subjects: Nash Sage And try this guy's website: http://www.riskoriented.com/ You will need to look at older posts... he has done some simular stuff that you are thinking about. rekop reply Thanks for the reply Nacht - I imagine a lot of members were thinking wtf! with regards to this post! I have now downloaded pokerstove and have spent quite a few hours number crunching, it''s a great tool! Quote: You would indeed have to put every one of those 169 into pokerstove (great, we can't do without, you will have to learn how to use it Smile and pit that against different ranges for tight players and loose players and anything in the middle. 1) Yes this is the standard and correct way to do it - but away from the table. But my idea is to have just one list to use as a rough approximation. It would be a single ranking of all hands (like a football league of sorts) but where each hand is pitted against every other hand and an equity value is gained. They are then put in order from best to worst. (I have seen lists on the net since posting originlly but most of them are for 9 or 10 handed games for are useless for HU) Quote: We are not looking for "coin flip situations". Here is your biggest flaw... there is a very big difference when blinds are at 10-20 or when they are at 100-200. At 100-200 there is a lot more money in the pot so you would need way less than your "coinflip". 2) I am not saying we are looking for coinflip situations at all - what I am saying is that if you know you are in a coinflip (or better) you always have +EV. What a "coinflip" means is that you are a 50:50 to win at showdown. So what range would an opponent need to average  - or be worse than 50% - your 2 cards. Trivial  example - small blinds early in game - Villain All In pre-flop. You hold KK. You look at your list and see  KK is ranked number 2. In order to fit the coinflip situation  the villian would need to hold the range AA, KK, QQ. If you believe QQ to be the absolute bottom of his range it is still worth calling. Of course if you feel he would have weakewr holdings like JJ or AQ then the call is highly +ev. Or if you felt the opponents shoving range was any two cards (1 to 169) then the coinflip situation would be ranked around position 85 (maybe a 10 8o or whatever...) Any holding better than that and you are winning more +EV of course. Having said all that i have since discovered a flaw in my logic (since getting Pokerastove!!) and that is that a ranking is not linear. In other words, it seems that a villain´s hand range from 1 to say ranking 100 DOES NOT equal, on average, ranking 50 (your holding). Rather it ia a curve....so you would need a chart rather than a list. The other main obstacle is for all the cases when the villain's average holding is better than yours - which is most often the case hence the tricky decisions. This method as it stands does  not ccount for that , but I believe some simple enough maths concerning equity and pot odds could be preformed at the table given a little practice and the visual graph ranking. Well I'm sstill working on it.... The only way to prove that this works is by a) asking some MIT maths professor or b) plugging in all 169 hands, drawing a graph, testing it and then verifying the results in Pokerstove. That's a helluva lot of work...but i will try to do it eventually. Meanwhile i will check out those nash and sage - must admit I've never heard of it. nachtwacht It appears to atleast be a little clearer to you Still, the problem i have is your 50/50 An example. You have 210 chips left. Blinds are 100 / 200 You are holding 32o in the SB. That hands is the worst possible hand in HU and you would be way worse than 50/50 against his calling range. (lets say you would be roughly 30%) Do you "push" ? Ofcourse the answer is yes... my point is, your 50/50 chart would be useless here since that hand will never ever even be close to a 50/50 but still, the money in the pot makes this a push. As far as I understand from the things you want to calculate, you don't appear to be accounting (enouhg) for money allready in the pot. Knowing wich hands will be 50/50 just isn't enough.... in some situations 45/55 is allready +ev and in the above example, 30/70 is also +ev dzikijohnny Doesn't the Sklansky Chubakof (?sp) give these numbers or something like them.        Tagpoker Forum Forum Index -> Heads Up Poker Strategy Page 1 of 1