personally i think doing only a hundred or so trials to crank out the graphs is a joke. does odds oracle (or any other commercial poker s/w) offer the ability to crank out equity graphs with a larger/customizable number of trials and/or a table/spreadsheet of equity/% flop coordinates?
Tom Coldwell11 years, 11 months agoYeah, as Tian said, you can set to any number. The only issue is that to do every possible flop against a range of every possible AA hand you might need the kind of computing power used by Pixar to make Toy Story (or a willingness to wait AGES).
xazervl11 years, 11 months agoCan anyone help me to figure out difference between "Graph HvH" and "Graph HvR" buttons? I read the manual on the propokertools site but i didn't understand it...I thought we use first one when we make graph vs specific hand (AcAd5c9h for example) and HvR vs all AA, but Tri Nguyen in his book and Scratch in his arcticles use ppt v1 with only one graph button that equals HvH ppt v2. When we use HvR we get more precise result?
Tom Coldwell11 years, 11 months agoThe HvH function deals a ton of flops against specific hands from the assigned range as opposed to giving your equity on fewer flops but against the entire range. Because in reality you'd require perfect knowledge of your opponent's hand to play as the HvH graph would recommend, it's a less accurate representation of how you'll actually play (for example, the HvH sim would see you making crazy hero folds like JTh8h7 vs. AhAJhT on 7h53h).
Hopefully that made sense. Either way, you should certainly be using the HvR button in situations where you don't know villain's exact hand.xazervl11 years, 11 months agoThanks a lot, now I've understood it completely.
Aleksandra ZenFish11 years, 11 months agoTom, this is great in depth analysis of a hand, my question is when we re just in a hurry ( or lazy )
cant we just do a quick EV calculation like
W : % we win G : how much we win L : How much we lose EV=W x G - (1-W) x L
http://propokertools.com/simulations/show?g=oh&h1=KhJhQdTd&h2=AA%2A%2A&s=generic
EV=0.41 x 200 - (1-0.41) x 100
82-59=+23
Aleksandra ZenFish11 years, 11 months agoPS ~ number looks to me approx to EV 30.5 -9 you did in end calculation on minute 18.08Tom Coldwell11 years, 11 months agoThis calculation isn't related to what I did at all and any similar result is pure coincidence from what I can tell. There's no reason that I can see why this should give us anything approaching a realistic number (indeed, your +$23 is nowhere near the +$9 I came up with).
In fact, I can create a hypothetical example to show you why something of this sort couldn't work:
Lets assume that we have a hand w/ 41% equity against AA (the same as the KQJT hand) but that we can only flop 100% equity or 0% equity (turn and river can't change anything). In this case, we'll flop the winner on 41% of flops and the loser on 59% meaning we'll stack off with all the equity 41% of the time and save our money 59% when we're dead. Using my calculation, we are (((201.5 *1) - 19.5 - 71.5)) * 0.41 + (-19.5 * 0.59) = (110.5 * 0.41) - (19.5 * 0.59) = $45.31 - $11.51 = $33.80. As you can see, despite having the same hot/cold equity as KQJTds, my imaginary hand performs FAR calling a 4-bet as reflected in the numbers I have gotten ($33.80 and $9.00 respectively). However, the type of calculation you're trying to create would view both hands as identical and fail to draw what is a huge distinction between the two.
Now obviously reality isn't this extreme, but the fact that your calculation doesn't take into account equity distribution is big flaw (I'm also can't see how it relates to the issue at hand which is another pretty major problem).
Hopefully that made sense.
Aleksandra ZenFish11 years, 11 months agoIt does make sense, it doesnt account equity distribution at all hence big differences in outcomes
I wanted a shortcut :( that math looks awefully long
Tom Coldwell11 years, 11 months agoJust takes practice. When I was doing the vid, I could get most calcs from creating graph to final result in under a minute.
Aleksandra ZenFish11 years, 11 months agook ill set up that as my weekend practice , need improve my math anyway :)
so far i used similar table and memorised approx what i can call 4 bet with what cant without actually doing it
Nom de Guerre11 years, 11 months agogreat insights by tom, but the work looks kind of painful esp when i think this could be done with less manual labor and better accuracy through software.. maybe this is something to ask odds oracle (?) to provide for in their next release.
Tom Coldwell11 years, 11 months agoThere is certainly software out there which can do this more accurately than I did. The main reason I chose this method is so that it could be copied by everyone watching, not just those who are willing to pay a load of money for specialist software.
can u use this, calling 5 bets or 6 bets and so on? or is this math calculation just for calling 4 bets?
Tom Coldwell11 years, 8 months agoYes. This works fine for 3-bets, 5-bets, w/e you like. The issue comes when the basic assumptions change (ie that villain has AA** and will stack off every flop). The reason I've titled it about 4-bets is because it's the most common spot 100bbs deep (with a 5-bet, you can then usually just run hot/cold equity as the money will be in already).
do we have to use line of best fit or can we just use the graph thats there? or is that to difficult? i think ive seen a video of someone not useing line of best fit.
Tom Coldwell11 years, 8 months agoYou need to know your average equity so I can't really see how you'd do this easily without using a line/lines of best fit. It certainly seems like the simplest method to me, but if you can find another you prefer, obviously feel free to go with that.
Tom Coldwell11 years, 8 months agoSorry, not that I know of (although I've never looked). I believe PPT Odds Oracle will do some of this for you (average equity) and you could program Microsoft Excel to do the rest I guess, but I am unaware of a program that does the whole lot.
Max equity, at least as I use it here, is a number which only matters in so far as I use it to calculate our average equity when we stack off postflop. My understanding is that if you use Odds Oracle you can get it to tell you your average equity on the flop when you have above a certain threshold. That being the case, our max equity would be rendered irrelevant (at least as far as what this video is teaching) 'cas you can just get the average and go from there.
However, if you are using PPT's free online stuff then the easiest method I can see for working out our average requires lines of best fit, max equity etc. I can't really see a quick way round that I'm afraid.
is it ever okay to cold call a 4 bet? i find it hard to calculate if a cold call of a 4 bet is +EV, because theres players/player behind too act. so how can we calculate when we are cold calling a 4 bet?
I had drafted a response that said in theory this could be okay, especially against a relatively small 4-bet. However, in reality 100bbs deep you really shouldn't run into a spot where this is advisable. To have 3 players raise in front of you and then pick up a hand good enough to call yourself is seriously tough (you need to do really well to justify calling without money invested). So as general advice I would say just toss your hands when facing a 4-bet unless there are some seriously weird considerations to the contrary - a couple of examples could be:
1) 4-bettor is 4-betting 100% meaning you can jam whatever is a value-range against the 3-bettor.
2) The 3-bettor and 4-bettor are both nitty enough that you can almost guarantee both have AA. In those cases you can probably just ship stuff which has an equity edge against two AA hands (which, for the record, is most well-connected hands).
But yeah, the general rule is fold all non-AA hands against a 4-bet if you haven't already put money into the pot (obv just jam AA). This won't cost you much (if any) value - especially when we remember the low-stakes rake monster - and it'll avoid what will be a disgustingly high-variance spot.
Feel free to keep doing calcs and thinking about this though, will almost certainly expand your knowledge and make you a better player. If you want a tip on how to think about these spots, I would start out simply making it a HU pot where you have nothing invested and go from there (if you can call a 30bb bet to put in your last 70 OTF without any dead money profitably, you may find you have a hand which can be cold-calling). If you have something which passes that original test, perhaps you can start looking at how you do HvR against AA** and something like 20%!AA (a 20% range without AA in it).
ty very much tom, much appreciated!:) guess i should not be cold calling a 4 bet as original raiser either or is that way more okay?
Tom Coldwell11 years, 4 months agoThis becomes slightly closer, but there certainly won't be THAT many hands you wanna cold-call with unless the 4-bet is small. Best thing you can do is run some sims and have a look at how you're doing.
We don't call with pairs and aces, right? Odds Oracle counts we will have 35% Eq on 38% flops. So, should we call with hand like this or there is a mistake in this calculation?
Tom Coldwell11 years, 2 months agoThat method doesn't solve this problem, it merely begins to. Once it's told you how frequently you'll flop enough equity to go w/ it on the flop, you need to solve how much equity you'll have and then do calculations using those two numbers.
To see how this number can't be enough, imagine two scenarios where you flop >35% w/ 40% of hands. In scenario 1, you always flop 100% equity, whereas in scenario 2 you always flop 35.1% equity. Quite clearly, one of these is a great spot, the other is really, really awful.
I do not quite understand how to use the soft. Just rule of thumbs: we don't call 4b (if villian 4b only AA) 100bb deep w pairs and aces, except hands like JJTTds, super connected doublesuited pairs. Is that so?
Tom Coldwell11 years, 2 months agoYes, as a rule of thumb, aces, pairs, and a lack of connectivity are bad. The exact thresholds differ depending on the stack-sizes which I why I can't be super specific there. You would be best served trying to run a few calculations of the type outlined in this video. Failing that, folding most pairs and A-high hands to the 4-bet and calling most else should work out well enough (oh, and probably don't fold a double-suited hand that often).
in the average equity calculation. just calculate (88%+35%) / 2 = 61,5% ? am I missing something or why you do the hard way?
Tom Coldwell11 years, 2 months agoYou are, of course, correct. (Max + Min)/2 = (Max - Min)/2 + Min = 1/2Max + 1/2Min. The reason I showed the harder way was so people could see the process working (as in why it's correct), but in retrospect it would probably have been a good idea to put this method in afterwards given it's greater simplicity. Good catch!
Does anyone use Equilab omaha? I have this program downloaded, it is free has a graph function and seems to be fine. But for some reason no one seems to mention it. Is there a reason? Should I not use it?
Isn't this method giving us too much ev when calling a 4-bet? With this method, we assume that we know our equity otf and will never stack off with less than our required equity. For instance we will nearly always stack off on the flop with a fd, but some times villian will have dominating one with his AA, and we will get in with way less than out required ev.
Tom Coldwell10 years, 10 months agoGood question, but no, it doesn't do this. It assumes we always know our equity against villain's range - AA** - which we do (on the assumption that villain only 4-bets AA** which, for most low stakes villains, is the case). Hence, when we have a flush draw, it averages our equity against those AA hands which are simply a naked overpair across those which have nut flush draws, top sets, flush blockers, even lower flush draws (say we hold KQT8 w/ a flush draw and villain has AA25 w/ the suited 25).
As a result, it means that at times, we will stack off dreadfully, but never dreadfully against a range, only the specific hand (consider calling w/ JJTTds, the flop comes J44, we fist-pump stack off, villain flips up AA44 and we have one out. This method assumes we're gonna stack of dreadfully there 'cas obviously we are).
Ah, I thought PPT simulated our hand against all the AA combos one after another and then picked out all those where we flop less than required equity. But simulating flops against a range as you say ofc makes much more sense. Thank you!
Is this assuming that we are calling the villains all in flop bet regardless of flop? If this is the assumption should we really be calling on all flops? If not what is the decision making process for calling the all in on the flop?
i was wondering what hand and board combination will yield you just over 35% ev?
rephrased: at what kind of flop should I be looking at bare minimum to GII?
gutshot + two bdfd? or just a lone pair without any possible other draws?
That seems to be a bit below 35% but depends if you have more backdoor straights. Also if you 3-bet bigger you might not need the 35%, just play around with propokertools :)
Loved the video. I think the formula can be expanded to include the partial fold frequency on 4-bettors part when when we smash the board and he misses completely like "789" no fd/bdfd, and we'll have some stackoff frequency when we're below our target EV (like dominated flush draws, dominated pairs and non AA sets). So for example:
($Calling * it's %) + ($Calling below EV * it's %) + ($Winning w no showdown * it's %) + ($Folding * it's %) = $EV
So something like this:
Great video, using the dashes (as you like to do for your bulletpoints) in front of the formulas can be a bit confusing as people might mistake them for minuses
Loading 45 Comments...
personally i think doing only a hundred or so trials to crank out the graphs is a joke. does odds oracle (or any other commercial poker s/w) offer the ability to crank out equity graphs with a larger/customizable number of trials and/or a table/spreadsheet of equity/% flop coordinates?
Hopefully that made sense. Either way, you should certainly be using the HvR button in situations where you don't know villain's exact hand.
cant we just do a quick EV calculation like
W : % we win G : how much we win L : How much we lose EV=W x G - (1-W) x L
http://propokertools.com/simulations/show?g=oh&h1=KhJhQdTd&h2=AA%2A%2A&s=generic
EV=0.41 x 200 - (1-0.41) x 100
82-59=+23
In fact, I can create a hypothetical example to show you why something of this sort couldn't work:
Lets assume that we have a hand w/ 41% equity against AA (the same as the KQJT hand) but that we can only flop 100% equity or 0% equity (turn and river can't change anything). In this case, we'll flop the winner on 41% of flops and the loser on 59% meaning we'll stack off with all the equity 41% of the time and save our money 59% when we're dead. Using my calculation, we are (((201.5 *1) - 19.5 - 71.5)) * 0.41 + (-19.5 * 0.59) = (110.5 * 0.41) - (19.5 * 0.59) = $45.31 - $11.51 = $33.80. As you can see, despite having the same hot/cold equity as KQJTds, my imaginary hand performs FAR calling a 4-bet as reflected in the numbers I have gotten ($33.80 and $9.00 respectively). However, the type of calculation you're trying to create would view both hands as identical and fail to draw what is a huge distinction between the two.
Now obviously reality isn't this extreme, but the fact that your calculation doesn't take into account equity distribution is big flaw (I'm also can't see how it relates to the issue at hand which is another pretty major problem).
Hopefully that made sense.
I wanted a shortcut :( that math looks awefully long
so far i used similar table and memorised approx what i can call 4 bet with what cant without actually doing it
can u use this, calling 5 bets or 6 bets and so on? or is this math calculation just for calling 4 bets?
do we have to use line of best fit or can we just use the graph thats there? or is that to difficult? i think ive seen a video of someone not useing line of best fit.
tom, do u know of any good software that can do these calculations? i am willing to pay for it. what is the best for this?
a software that can do all this for me?
I have used an excel spreadsheet to input formula etc. Works like a charm :)
Just wondering how we find max equity in Odds Oracle. I don't want to have to draw lines every time ;)
Thanks for the question.
Max equity, at least as I use it here, is a number which only matters in so far as I use it to calculate our average equity when we stack off postflop. My understanding is that if you use Odds Oracle you can get it to tell you your average equity on the flop when you have above a certain threshold. That being the case, our max equity would be rendered irrelevant (at least as far as what this video is teaching) 'cas you can just get the average and go from there.
However, if you are using PPT's free online stuff then the easiest method I can see for working out our average requires lines of best fit, max equity etc. I can't really see a quick way round that I'm afraid.
Hope that answers your question.
is it ever okay to cold call a 4 bet? i find it hard to calculate if a cold call of a 4 bet is +EV, because theres players/player behind too act. so how can we calculate when we are cold calling a 4 bet?
I had drafted a response that said in theory this could be okay, especially against a relatively small 4-bet. However, in reality 100bbs deep you really shouldn't run into a spot where this is advisable. To have 3 players raise in front of you and then pick up a hand good enough to call yourself is seriously tough (you need to do really well to justify calling without money invested). So as general advice I would say just toss your hands when facing a 4-bet unless there are some seriously weird considerations to the contrary - a couple of examples could be:
1) 4-bettor is 4-betting 100% meaning you can jam whatever is a value-range against the 3-bettor.
2) The 3-bettor and 4-bettor are both nitty enough that you can almost guarantee both have AA. In those cases you can probably just ship stuff which has an equity edge against two AA hands (which, for the record, is most well-connected hands).
But yeah, the general rule is fold all non-AA hands against a 4-bet if you haven't already put money into the pot (obv just jam AA). This won't cost you much (if any) value - especially when we remember the low-stakes rake monster - and it'll avoid what will be a disgustingly high-variance spot.
Feel free to keep doing calcs and thinking about this though, will almost certainly expand your knowledge and make you a better player. If you want a tip on how to think about these spots, I would start out simply making it a HU pot where you have nothing invested and go from there (if you can call a 30bb bet to put in your last 70 OTF without any dead money profitably, you may find you have a hand which can be cold-calling). If you have something which passes that original test, perhaps you can start looking at how you do HvR against AA** and something like 20%!AA (a 20% range without AA in it).
ty very much tom, much appreciated!:) guess i should not be cold calling a 4 bet as original raiser either or is that way more okay?
can I call with JJQT$ss vs AA profitable?
http://uploads.ru/aS0BA.png
i count like this http://www.youtube.com/watch?v=QgDK_kF_UYM
We don't call with pairs and aces, right? Odds Oracle counts we will have 35% Eq on 38% flops. So, should we call with hand like this or there is a mistake in this calculation?
To see how this number can't be enough, imagine two scenarios where you flop >35% w/ 40% of hands. In scenario 1, you always flop 100% equity, whereas in scenario 2 you always flop 35.1% equity. Quite clearly, one of these is a great spot, the other is really, really awful.
Does that help?
I do not quite understand how to use the soft. Just rule of thumbs: we don't call 4b (if villian 4b only AA) 100bb deep w pairs and aces, except hands like JJTTds, super connected doublesuited pairs. Is that so?
in the average equity calculation. just calculate (88%+35%) / 2 = 61,5% ? am I missing something or why you do the hard way?
Does anyone use Equilab omaha? I have this program downloaded, it is free has a graph function and seems to be fine. But for some reason no one seems to mention it. Is there a reason? Should I not use it?
Isn't this method giving us too much ev when calling a 4-bet? With this method, we assume that we know our equity otf and will never stack off with less than our required equity. For instance we will nearly always stack off on the flop with a fd, but some times villian will have dominating one with his AA, and we will get in with way less than out required ev.
As a result, it means that at times, we will stack off dreadfully, but never dreadfully against a range, only the specific hand (consider calling w/ JJTTds, the flop comes J44, we fist-pump stack off, villain flips up AA44 and we have one out. This method assumes we're gonna stack of dreadfully there 'cas obviously we are).
Ah, I thought PPT simulated our hand against all the AA combos one after another and then picked out all those where we flop less than required equity. But simulating flops against a range as you say ofc makes much more sense. Thank you!
It can do this which is why it's very important to use the Hand vs. Range graphing function as opposed to the Hand vs. Hand graphing function.
Is this assuming that we are calling the villains all in flop bet regardless of flop? If this is the assumption should we really be calling on all flops? If not what is the decision making process for calling the all in on the flop?
i was wondering what hand and board combination will yield you just over 35% ev?
rephrased: at what kind of flop should I be looking at bare minimum to GII?
gutshot + two bdfd? or just a lone pair without any possible other draws?
That seems to be a bit below 35% but depends if you have more backdoor straights. Also if you 3-bet bigger you might not need the 35%, just play around with propokertools :)
Loved the video. I think the formula can be expanded to include the partial fold frequency on 4-bettors part when when we smash the board and he misses completely like "789" no fd/bdfd, and we'll have some stackoff frequency when we're below our target EV (like dominated flush draws, dominated pairs and non AA sets). So for example:
($Calling * it's %) + ($Calling below EV * it's %) + ($Winning w no showdown * it's %) + ($Folding * it's %) = $EV
So something like this:
Great video, using the dashes (as you like to do for your bulletpoints) in front of the formulas can be a bit confusing as people might mistake them for minuses
Be the first to add a comment
You must upgrade your account to leave a comment.