Usually when I write articles, I try to speak somewhat authoratatively - I'm arguing a position and trying to get you all to see why I think I'm right. I'm going to switch it up a bit and just present the data this time and then talk through it with you guys, because to be completely sure, I'm not 100% convinced of what to make it yet.The idea is that we're playing against a reg who opens the button wider than the average opponent. Previous theory posts have shown the value of jamming over minraises when this is the case, but they are usually just compared to folding. I want to talk more about the "fold or jam or call" decision that is much more typical.I made a table of expectation for jamming 20bb deep against 70% minraiser (Specifically A,K,QQ-22,QJ-Q3,JT-J6,T9-T7,98-96,87-86,76,Qx2x,Jx5x-Jx2x,Tx6x-Tx2x,9x5x-9x2x,8x5x-8x4x,7x5x-7x3x,6x5x-6x3x,5x4x-5x2x,4x3x).I presumed a calling range of AA-22,AK-A5,KQ-KT,Ax4x-Ax3x,Kx9x,QxJx-QxTx - some wide-ish hands in there, but pretty by-the-book and reg-ish.Here's the table. This compares expectation from the start of the hand, so -1.0bb is the same as folding. I put everything worse than folding in red, because we're not playing that stuff. I put everything in green that has better than 0EV in the entire hand - that's a great recipe for getting it in to very likely be the best decision, whether it's jamming or 3bet/calling. One thing I immediately notice is that this makes a strong argument for jamming hands like J2s-J6s, Q2s-Q4s, K2s-K4s, etc, while at the same time, flatting hands like J7s, Q8s, K7s, etc. J3s has very poor expectation flatting, and I think many of you would fold it rather than call. However, it's a full 0.4bb better to jam than to fold against these parameters. That's a great recipe for a light 3bet shove. On the other hand, K7s is a hand with very strong expectation when flatting - historically, with it and similar hands, I'm actually in the positive for flatting in this situation, much better than the -0.3bb from jamming.What are your thoughts when looking at this table?
cool stuff. Imo the key is that I really should be sure, that villain has >70% minraise range in the actual hand. It could be, that he had such a range before and that he became tighter or that my sample size of observed hands is too small. So for example J3s: vs. 70% ist has -0.6 (which is much better than foldingI, but vs 50% (see the other chart) J3s only has -1.4, which means I should rather fold.So for me it is a great play to 3bet shove ligt with hands like Q2s, J2s-J3s, T2s-T4s, if I have strong reads, that villain has a minraise rage >70%.
Yeah, I agree, it really hammers in support of the value of jamming weak suited hands against people who are opening a wide percentage. Against people like h2olga, livb112, etc, jamming these type of T4s hands is an absolute must, and the expectation is going to be much better than flatting or folding.I'm not convinced this is as true with the better suited connectors - an analysis of my database suggests about 0EV in expectation from flatting that clump of Q8s/J8s/J7s/T7s type stuff, which despite the strong equity from jamming, suggests that flatting might be a better option.
"I'm not convinced this is as true with the better suited connectors - an analysis of my database suggests about 0EV in expectation from flatting that clump of Q8s/J8s/J7s/T7s type stuff, which despite the strong equity from jamming, suggests that flatting might be a better option."What do you think about the hands 76s, 65s, 54s and which expectation for flatting would you guess? (with my small sample size I have - 0.6bb for these hands).Btw, ist it right that 1BB=2big blinds for Pokertracker?
Correct. I've started using HEM for all my calcs now that I've figured it out more.For the range of 54s-84s, 65s-95s, 76s-T6s, 87s-T7s, I have about -30bb/100 over about 1k hands. So I think it's pretty safe to say that the worst hands in that range are probably in the -50bb/100 range.
I have a question. In the table if we look at T8s it shows that it has -0.1bb expectation when 3bet jamming over min raise 20bb deep if villain is opening 70% and raise / calling with 21.9% of hands (22+,A3s+,K9s+,QTs+,A5o+,KTo+). Can you explain how you get that number (-0.1bb)? I'm using Insane_Steve's excel spreadsheet to calculate EV of 3bet shoves and if I'm trying to recreate that particular stack depth (1000 eff stack, 50 chips bb) and villain's frequencies / ranges I get the number of +48.7 chips . So it would seem that it's almost +1bb of EV when I jam T8s. Why it's only -0.1bb in your table?
I figured out how you got the -0.1bb expectation, but I don't understand how could it be like this. I mean, using excel I'm getting +48.7 chips (~1bb) of profit every time I rejam, so according to that it's clearly a +EV jam. In the screenshot is pokerprotools shove equity tool that I think you used to get that -0.1bb of expectation. I get your number if I subtract 1 from number in the "Shover net bets". In this case 0.9-1=-0.1. But I don't understant why I need to do that. If I don't subtract 1 then it shows ~1bb of profit, the same as using excel.
I think we're subtracting 1 because we're comparing expectation not to folding but from the start of the hand, like you wrote in the article. Starting from the very beginning. Same conditions apply, same ranges, same T8s hand. 20bb deep effective stacks before blinds are posted. Hero posts 1bb OOP. Villain opens 2bb OTB. If hero folds he will lose -1bb. If he jams he will gain +0.9bb. So overall expectation in the hand is -0.1bb. All the profit from rejamming is eliminated by the fact that we posted 1bb. But we're still better off jamming because if we fold our expectation will be -1bb? If we were to jam a hand that nets us 0 chips our expectation in the hand would be -1bb or the same as folding. Or in other words we would not be gaining anything by rejamming. But we're also would not be losing more. It's hard for me to get the fact that rejam is +EV in terms of chips and in terms of big blinds (+0.9bb) but our expectation in the hand is negative, -0.1bb. I understand it but I don't quite get it, if that makes sense :)
OK, I think you've figured out most of it.If we're comparing jamming to folding, then all we have to do is be better than -1.0 in the hand. It's OK for a play to have negative expectation, because folding always has negative expectation from the start of the hand.Insane Steve's calculator is wrong in a couple of ways, most notably that it doesn't completely take into account card removal. If you have A5o in the big blind, and your opponent raises, your opponent will be raise/calling less often than if you had 78s, because your opponent has more big aces in his range when you don't have one. In Insane Steve's spreadsheet, you always have the exact same chance of getting called when you shove. That's not true. It matters what hand you have.
In propokertools, if you get 0.9 back in addition to your shove of 19, that means that you end up with 19.9bb. That's 0.1 less than where you started.propokertools will show big blind getting back 0 when shoving is equivalent to folding, which is -1.0bb from the start of the hand.
"In Insane Steve's spreadsheet, you always have the exact same chance of getting called when you shove. That's not true. It matters what hand you have."Is it a significant difference? How much is it? Is it hard to calculate? There's 4 aces in the deck, so if we take 1 out thats 25% less aces. So if we look at it this way it should be a significant difference. I presume propokertools take card removal into consideration? If so, there's no need for Insane_Steve's spreadsheet anymore. I can do all the calculations with propokertools and everything is in the same place, I don't need to look at pokerstove, copy/paste all the stuff to spreadsheet.
"Yeah, I agree, it really hammers in support of the value of jamming weak suited hands against people who are opening a wide percentage. Against people like h2olga, livb112, etc, jamming these type of T4s hands is an absolute must, and the expectation is going to be much better than flatting or folding." How about at the lower/midstakes? Do you still think its profitable to jam hands like J2s-J6s, Q2s-Q4s, K2s-K4s against agressive openers, or are they to bad/loose to fold at those stakes? Have any of you tried out the S-C test in the Endgame_Calculator_v2_21 to practice what hands to push against different villians?Is it a good way to learn you think, or would you recomend another way?
I do not recommend the endgame calculator because of it's inaccuracies. It can be OK to get a general idea but it's bad to come to hard conclusions from it because of the calculation errors.At lower/midstakes, it's absolutely still good to jam those hands over very aggressive openers a lot of the time, though not quite as much.Download propokertools and try it out yourself.
If the opening range is:22+, A2+, K2+, Q3+, Q2s, J6+, J5s-J2s, T7+, T6s-T2s, 96+, 95s-92s, 86+, 85s-84s, 76, 75s-73s, 65, 64s-63s, 54, 53s-52s, 43s, 32s and the calling range is:22+, A5+, A4s-A3s, KT+, K9s, QTs+ pokerazor suggests:22+, A2+, K2+, Q4+, Q3s-Q2s, J5+, J4s-J2s, T6+, T5s-T2s, 96+, 95s-92s, 85+, 84s-82s, 75+, 74s-72s, 64+, 63s-62s, 54, 53s-52s, 42s+, 32s as a shoving range with an expectation above 0.00001 EV (not considering previous decisions) 20 bigblinds deep before the blinds are paid Did i make a mistake or is one of the programs not working correctly? edit:okay i adjusted the EV threshold to 0.1 now it only deters at 64o and 32s (both a shove)
I'm not 100% sure what you're saying, but it seems like you're making error here:"as a shoving range with an expectation above 0.00001 EV (not considering previous decisions) "This seems like you're saying the entire strategy has that EV. That's not what we're looking for.For example, every hand in NASH is better than folding to shove. It's not like it's the entire range's equity that we're calculating.So the question isn't, if I throw 62o in here, does it make my entire range -EV? The only question is whether 62o itself is a -EV shove.
No its okay i was trying to check if i can do with pokerazor what you did with propokertools and it worked, when i did it with 0.1 EV.(it showed every hands expectation but highlight those with 0.1 and better)I didn´t get why we wouldn´t go for 0.001 EV instead though.At least in Pokerazor it showed the EV of shoving . If we fold we end up with 19 bb. So if we end up with 19.001 by shoving we´re better off, than folding.
How do we play our range at <17BB vs a player who opens >65% and cbets >65%?Can we push our weak Kx/Qx hands? How do we play K7+, Q8, push/call or fold?How about the J7+, T7+, 97+, do we push/call or fold with them? Im really lost about the thought process here..How can I make tables like above? Want to try to make some for different stack sizes vs different oppening ranges?
So, there's a huge difference between 65% and 100%. ">65%" is way too broad of a description. Same for "<17bb". As for making tables, download propokertools and play around with ranges! :)
It doesnt work.Needed a key and when I try to get it I just get to a login page, where I cant do anything.
There should be a free trial associated with the shove calculator if I remember correctly?http://www.propokertools.com/odds_oracle/trial_license_key
sweat thanks!
I had a look at PPT.But I cant find how to make tables anyware. How do I do that?
I did it by hand.(<3 my students)
I did it in a haste and probably made some mistakes but you should really try to get your hands on a copy of pokerazor.you could rule the world with its computing power and your math/poker knowledge.maybe you realize what i was trying to say now.I couldnt reproduce your ranges to do this properly.
I can't really comment on PokerRazor not knowing the program - I am pretty confident propokertool's caclulations are correct, though. Subtract 2 and divide by 2 and you get your table converted in bb/hand from the start of the hand. There are small differences in the tables, could be due to a few factors.