Great analysis. This type of video is what I am looking for in content:
In-depth hand analysis using current software.
For me personally, there is a bit of a learning curve using oo, as I'm relatively new.
Having a guide to help with my own hand review is awesome.
- quick question: @17:30 you discount ev gained when V has set, which he has 25%. You do same when V has flush. This makes sense. Why do you discount those relative ev's again by total % V continues when we have flush ( 25/62, 37/62)?
Then @20:30, why do you discount the total ev when villain continues (3828) again by the total 62%?
I am probably missing something from the basic definition of ev, but having an explanation would be helpful.
Thanks Lefthook. I am glad you enjoyed it. I will preface this by saying it is far from an impossibiltiy that I made a mistake but let me explain the part you are questioning.
We are trying to evaluate our EV of making a flush overall.
First we assume the turn is the Tc (makes us a flush) and we evaluate our EV on that turn.
To do that we look at the 3 following scenarios:
1) He has a set 2) He has a flush 3) He has niether
1) We have some implied odds 2) We have full implied odds 3) We have 0 implied odds
We next are reducing those 3 scenarios into 2 scenarios:
A) Times we have implied odds B) Time we don't
A) = 1) + 2) B) = 3)
So to find our overall implied odds in A) we are taking a weighted average of the implied odds we have in 1) and 2). 25% he has set 37% he has flush; which means we have nonzero implied odds with our flush 62%= (25+37).
The other 38% we still win the pot and have EV but we just win the current pot because we have no implied odds
So the overall ev of us making a flush can be written as
25%/100%(EV_when he has set)+37%/100%(EV_he has flush)+38%/100%(EV_He has neither)
Then this is weight averaged when calculating our Overall EV because we only make a flush ourselves 20% frequency.
Zach, great video. No PLO expert but arguments are very well designed and clearly put on the video.
a question on the math on deciding to pot flop or not:
to achieve 43.6% equity against his range you used a uniform probability distribution for each range 1/3. Given that he is tilted/stuck and recreational, shouldn't a probability distribution that isn't flat and consider heavy tails on wider ranges work better ? It gives a bit more of subjectivity to the modelling and I am sure that this is more well-fitted when you have specific reads but in this situation I rather use this other probability distribution format.
If the 3 ranges I prescribed were A) Nitty Range B) Medium Range and C) Tilty Spewy Range
then your idea to weight towards the tilty spewy range would be ideal. However the three ranges I prescribed were A) Somewhat Wide B) Wide and C) Really Wide.
meaning, even A) my tightest range, already took consideration into the read that he was probably tilting and on the looser side.
Even the tightest range I used still contained lots of draws that a nitty range wouldn't.
Hey Zach, nice analysis! There's nothing I'd like to add or correct, just a cpl things I'd like to comment on.
1. There's a difference between 70%-15% and 70%!15%:
70%-15% is 151615 combos, or 56% of all hands 70%!15% is 148907 combos, or 55% of all hands
The difference is fairly subtle, and for this analysis it doesn't really make a difference. The hands that differ are the ones in the 14%-15% range. You may well have intended exactly what you put in there, but for the benefit of others I'm pointing this out. I see many mistakes in OO syntax, and a lot of it is due to... well, many things in the syntax being very subtle. And, for some other analysis it may well make quite a significant difference.
2. For me, the main takeaway from an analysis like this is not that stacking off on the flop is -EV, or that calling is +EV. I kinda knew that already. The main benefit is in quantifying the call EV, which you did a good job of imo. And taking that a step further, if you had the same hand with a K high fd, then the call would probably be right around even, and with a Q/J high fd it's a fairly clear fold. I think a lot of people know that intuitively also, but seeing it calculated here adds extra weight to it.
What's also interesting to think about is how these things change if you (somehow) have another club (blocker) in your hand, if you didn't have the pair OTF, or if the board is 762 instead. Intuitively, the extra club would probably not make a huge difference, whereas the pair gives you a fair bit of equity here. A 762 board would obviously change the situation quite a bit, but probably making the call more profitable. What do you think?
Thanks for poinitng out that difference. I did not realize it but once reading the syntax it makes perfect sense given ! is not and - is minus. As you pointed out it doesnt matter for this exercise.
Also good points about the extrapolations we can make about other strength draws we might have.
What's also interesting to think about is how these things change if you (somehow) have another club (blocker) in your hand, if you didn't have the pair OTF, or if the board is 762 instead.
I think the extra club would actually make a signifcant difference given it would cut down our implied odds for him being over flushed. And in reagards to 762cc flop, I think that certainly makes a huge difference because straights are the hands that kill our hand most and eliminating the top of his range is certainly a big boost to our equity.
Enjoyed the analysis, one nit pick: as you widened his range, it looked like you added mostly more dubious drawing hands (wrap on fd board, etc.). In my experience fish are far more likely to play made hands aggressively and draws passively, so I'd expect him to show up with more sketchy two pairs than weak draws. Doesn't necessarily apply to this villain, of course, but I've found it to be a pretty solid read for weaker players.
It seems like our EV mostly hinges on how often he will stack off with worse flushes on a flushing turn. Although he hits flushes ~37% of the time, I kind of doubt that he will get all the money in; he might be able to find a bet/fold, in which case our EV goes down rather significantly.
In overall, I agree that our flop call is most likely +EV, but it looks pretty close.
Another question is, if he checks on Tc turn, how large will you bet?
I agree about him not stacking off all flushes but I didn't want to try and model that in there given it was already getting to have a lot of assumptions. And there is some chance he has some bluffs that barrel off on flush turns although unlikely.
I'd prob bet a "normal" 2/3 pot or so if checked to. It is enough to get stacks by river and enough that a set will call incorrectly. I dont think betting pot would be great given we wont have many bluffs. I wouldn't bet small becasue he is peeling a larger bet with all flushes and sets and going smaller wont start to add in straights and wraps and 2pr likely.
watched the first 3 mins and after your description of villian this seems like a trivial raise to induce/get it in spot to me... you are def ahead of a mid 40s stuck gambling rec player's flop potting range and should just push your equity. don't like a call as turns are iffy to play. gonna watch it all ..
1. i think you have to take some bluffs into his range especially 33 44 if you think he plays 22 55 66 (i do so too)
2. you should lower the distribution of 34 22 55 66 hands if that is possible with the tool as he is sometimes (i think in general more often than not) c/r those
and those 2 points should already make your life easy i'd imagine
I believe for the reasons mentioned in the video that villains of this type call off light but don't necessarily pot into us as light.
Also regarding reducing 34 22 55 66 combos from his leading and into his cr range, I think that's true but I also think the draws such as wraps+FD will be in that cr range as well so it would be close enough to a wash. I don't think we can assume he leads draws more than made hands.
What makes this hand interesting is that its LIVE PLO and it really is a weird spot.
In a live game with stack sizes its hard to play your hand. To the average player they think you just get it in and be happy but that's not the case. Its hard to say that he has many straights in his range. Its live so his limp calling range can be any 4 cards or somewhat connected 3456 2356 type hands. I don't see many sets on this board though but with your description its possible.
Now the interesting part. You have both the Ac and Kc but didn't really discuss that for stack off ranges. I think that's really important that you have the Kc here and we can discount more flushes that normal. We also have a 6 and that blocks a lot of hands. That now weights his range to made hands that are betting for protection. I cant see him betting a J or Q hi flush here w confidence. His hand is now polarized to a bluff or straight.
The other issue is that we are facing a PSB on the turn and have to risk a lot more of our money when sizes are deep. I think its a good fold but is close. Your results show a small +EV plan but I really think that number changes when we start to take out FD from his range. Even if he does have a FD he's not going to stack off multiple streets vs you. At least he shouldn't....
Loading 16 Comments...
Hey Zach,
Great analysis. This type of video is what I am looking for in content:
In-depth hand analysis using current software.
For me personally, there is a bit of a learning curve using oo, as I'm relatively new.
Having a guide to help with my own hand review is awesome.
- quick question: @17:30 you discount ev gained when V has set, which he has 25%. You do same when V has flush. This makes sense. Why do you discount those relative ev's again by total % V continues when we have flush ( 25/62, 37/62)?
Then @20:30, why do you discount the total ev when villain continues (3828) again by the total 62%?
I am probably missing something from the basic definition of ev, but having an explanation would be helpful.
Thank you for your time.
Thanks Lefthook. I am glad you enjoyed it. I will preface this by saying it is far from an impossibiltiy that I made a mistake but let me explain the part you are questioning.
We are trying to evaluate our EV of making a flush overall.
First we assume the turn is the Tc (makes us a flush) and we evaluate our EV on that turn.
To do that we look at the 3 following scenarios:
1) He has a set 2) He has a flush 3) He has niether
1) We have some implied odds 2) We have full implied odds 3) We have 0 implied odds
We next are reducing those 3 scenarios into 2 scenarios:
A) Times we have implied odds B) Time we don't
A) = 1) + 2) B) = 3)
So to find our overall implied odds in A) we are taking a weighted average of the implied odds we have in 1) and 2). 25% he has set 37% he has flush; which means we have nonzero implied odds with our flush 62%= (25+37).
The other 38% we still win the pot and have EV but we just win the current pot because we have no implied odds
So the overall ev of us making a flush can be written as
25%/100%(EV_when he has set)+37%/100%(EV_he has flush)+38%/100%(EV_He has neither)
Then this is weight averaged when calculating our Overall EV because we only make a flush ourselves 20% frequency.
Is that more clear? If not feel free to ask more.
Zach, great video. No PLO expert but arguments are very well designed and clearly put on the video.
a question on the math on deciding to pot flop or not:
to achieve 43.6% equity against his range you used a uniform probability distribution for each range 1/3. Given that he is tilted/stuck and recreational, shouldn't a probability distribution that isn't flat and consider heavy tails on wider ranges work better ? It gives a bit more of subjectivity to the modelling and I am sure that this is more well-fitted when you have specific reads but in this situation I rather use this other probability distribution format.
If the 3 ranges I prescribed were A) Nitty Range B) Medium Range and C) Tilty Spewy Range
then your idea to weight towards the tilty spewy range would be ideal. However the three ranges I prescribed were A) Somewhat Wide B) Wide and C) Really Wide.
meaning, even A) my tightest range, already took consideration into the read that he was probably tilting and on the looser side.
Even the tightest range I used still contained lots of draws that a nitty range wouldn't.
Hey Zach, nice analysis! There's nothing I'd like to add or correct, just a cpl things I'd like to comment on.
1. There's a difference between 70%-15% and 70%!15%:
70%-15% is 151615 combos, or 56% of all hands
70%!15% is 148907 combos, or 55% of all hands
The difference is fairly subtle, and for this analysis it doesn't really make a difference. The hands that differ are the ones in the 14%-15% range. You may well have intended exactly what you put in there, but for the benefit of others I'm pointing this out. I see many mistakes in OO syntax, and a lot of it is due to... well, many things in the syntax being very subtle. And, for some other analysis it may well make quite a significant difference.
2. For me, the main takeaway from an analysis like this is not that stacking off on the flop is -EV, or that calling is +EV. I kinda knew that already. The main benefit is in quantifying the call EV, which you did a good job of imo. And taking that a step further, if you had the same hand with a K high fd, then the call would probably be right around even, and with a Q/J high fd it's a fairly clear fold. I think a lot of people know that intuitively also, but seeing it calculated here adds extra weight to it.
What's also interesting to think about is how these things change if you (somehow) have another club (blocker) in your hand, if you didn't have the pair OTF, or if the board is 762 instead. Intuitively, the extra club would probably not make a huge difference, whereas the pair gives you a fair bit of equity here. A 762 board would obviously change the situation quite a bit, but probably making the call more profitable. What do you think?
Thanks for poinitng out that difference. I did not realize it but once reading the syntax it makes perfect sense given ! is not and - is minus. As you pointed out it doesnt matter for this exercise.
Also good points about the extrapolations we can make about other strength draws we might have.
I think the extra club would actually make a signifcant difference given it would cut down our implied odds for him being over flushed. And in reagards to 762cc flop, I think that certainly makes a huge difference because straights are the hands that kill our hand most and eliminating the top of his range is certainly a big boost to our equity.
Enjoyed the analysis, one nit pick: as you widened his range, it looked like you added mostly more dubious drawing hands (wrap on fd board, etc.). In my experience fish are far more likely to play made hands aggressively and draws passively, so I'd expect him to show up with more sketchy two pairs than weak draws. Doesn't necessarily apply to this villain, of course, but I've found it to be a pretty solid read for weaker players.
I agree. That would help our ev I'd imagine.
Good vid Zach, I enjoyed it. :)
It seems like our EV mostly hinges on how often he will stack off with worse flushes on a flushing turn. Although he hits flushes ~37% of the time, I kind of doubt that he will get all the money in; he might be able to find a bet/fold, in which case our EV goes down rather significantly.
In overall, I agree that our flop call is most likely +EV, but it looks pretty close.
Another question is, if he checks on Tc turn, how large will you bet?
- midori
I agree about him not stacking off all flushes but I didn't want to try and model that in there given it was already getting to have a lot of assumptions. And there is some chance he has some bluffs that barrel off on flush turns although unlikely.
I'd prob bet a "normal" 2/3 pot or so if checked to. It is enough to get stacks by river and enough that a set will call incorrectly. I dont think betting pot would be great given we wont have many bluffs. I wouldn't bet small becasue he is peeling a larger bet with all flushes and sets and going smaller wont start to add in straights and wraps and 2pr likely.
watched the first 3 mins and after your description of villian this seems like a trivial raise to induce/get it in spot to me... you are def ahead of a mid 40s stuck gambling rec player's flop potting range and should just push your equity. don't like a call as turns are iffy to play. gonna watch it all ..
gotta go right now so just a short comment -
2 things :
1. i think you have to take some bluffs into his range especially 33 44 if you think he plays 22 55 66 (i do so too)
2. you should lower the distribution of 34 22 55 66 hands if that is possible with the tool as he is sometimes (i think in general more often than not) c/r those
and those 2 points should already make your life easy i'd imagine
Privko, thanks for the input.
I believe for the reasons mentioned in the video that villains of this type call off light but don't necessarily pot into us as light.
Also regarding reducing 34 22 55 66 combos from his leading and into his cr range, I think that's true but I also think the draws such as wraps+FD will be in that cr range as well so it would be close enough to a wash. I don't think we can assume he leads draws more than made hands.
Please tell me you did not fold.
For a small fee I can.
A year later but I just watched it...
What makes this hand interesting is that its LIVE PLO and it really is a weird spot.
In a live game with stack sizes its hard to play your hand. To the average player they think you just get it in and be happy but that's not the case. Its hard to say that he has many straights in his range. Its live so his limp calling range can be any 4 cards or somewhat connected 3456 2356 type hands. I don't see many sets on this board though but with your description its possible.
Now the interesting part. You have both the Ac and Kc but didn't really discuss that for stack off ranges. I think that's really important that you have the Kc here and we can discount more flushes that normal. We also have a 6 and that blocks a lot of hands. That now weights his range to made hands that are betting for protection. I cant see him betting a J or Q hi flush here w confidence. His hand is now polarized to a bluff or straight.
The other issue is that we are facing a PSB on the turn and have to risk a lot more of our money when sizes are deep. I think its a good fold but is close. Your results show a small +EV plan but I really think that number changes when we start to take out FD from his range. Even if he does have a FD he's not going to stack off multiple streets vs you. At least he shouldn't....
Be the first to add a comment
You must upgrade your account to leave a comment.