DonoDrapero
34 points
Part 3 with hands where the AI donks
It seems some of those leads are objectively bad, no?
E.g. BTN opens, SB flats, BB (AI) calls w/ 67o. Flop is AJ6r, AI leads 2x pot.
July 18, 2019 | 10:08 a.m.
The 6max bot Pluribus beat a group of professional players in a recent match. The hand histories are hard to navigate through, but some people were kind enough to paste interesting ones in the 2+2 thread ( https://forumserver.twoplustwo.com/29/news-views-gossip/first-ai-beat-multiplayer-poker-game-vs-top-pros-like-chris-furgeson-darren-elias-1747396/ ).
Since solvers have been around for a while (and therefore we can already check unexploitable strategies for HU) the most interesting aspect of a match like this to me is how an AI approaches multiway spots. And it's surprising, because based on the hands it seems to be wildly aggressive. E.g. opens CO w/ K7dd, BTN calls, BB calls. Flop is 854r, bot cbets 2.3x pot. Is there an assumption BB should be leading most of his strong range? My thinking was overbets are not used multiway, because a single players pot share rarely allows him to.
Another example - 4way flop, bot shoves over cbet w/ 65dd on T42dd, SPR 10. I can sort of understand this one, but it's still interesting to think this is the highest EV line according to the bot.
July 12, 2019 | 9:25 a.m.
How do you approach studying multiway pots? Do you believe Monkersolver multiway outputs are solid? What other study tools would you recommend if someone was willing to try anything?
June 2, 2019 | 3:22 p.m.
How concerned are you about the possibility of someone using real time assistance when you see plays like this (e.g. unusual line/sizing that is very hard to apply correctly, but it was done with one of the few right combos)? I'm not talking about LLinus in particular, but personally when I see plays like this I become quite uncomfortable.
April 24, 2019 | 1:06 p.m.
Great video (though I prefer 6max). You mentioned ActionFreak being a HU boss, can we get your HU PLO power rankings?
Dec. 19, 2018 | 6:01 p.m.
Hm, ok. Let me rephrase then - how much study time would you recommend towards exploting the population and/or specific regs in tough high stakes games? Is it in your opinion a waste of time since people re-adjust accordingly if you try to exploit spots significantly? I don't play 25/50+ NLHE, but I'm curious how the best approach the game these times.
Nov. 7, 2018 | 2:21 p.m.
Nice vid.
How much do you think is is worth to deviate from our designed 'GTO' strategy on high stakes nowadays? How much do you try to exploit when playing really tough competition - close to never / a little / quite a bit?
Nov. 6, 2018 | 7:20 p.m.
23:00 top right. What hands are you calling down with here?
April 22, 2015 | 10:49 a.m.
Interesting. How much of mixing your actions with the same hands is based on actual work off the tables (leading to a high degree of confidence that mixing your action with 76o on that A42r is beneficial) and how much on intuition (hunch that it's good, but not based on work in CREV or whatever)? You seem to be advocating more than one option quite often and you always struck me as a player who is very precise in his strategy so I found that to be a bit surprising.
April 19, 2015 | 1:02 p.m.
28:40 A3o. You say your hand falls within the threshold of calling a river bet. Don't we have enough diamonds in our range to fold non-diamond hands?
33:50 If you have seen him bet bottom pairs and you block the Ace isn't this precisely an argument for a check? Seems like his range is almost always air here.
44:15 What makes 98o a good 4bet? Playability post, because neither an 8 or a 9 is in his 4bet calling range usually? Do you ever 4bet bluff weak Aces (A2-A5o), because of the Ace blocker?
Feb. 25, 2015 | 3:30 p.m.
I really appreciate your responses, I think it's insanely valuable that we get to ask questions to a player of your caliber. Thanks Ben.
Feb. 7, 2015 | 9:35 p.m.
Just posting to say I really appreciate your videos and I look forward to you crushing 25/50+ in the future.
Feb. 7, 2015 | 9:30 p.m.
My NL game is rusty, couple of questions for you:
9:00 why is Ac7 a bet here, what does it accomplish? (similar question for A2o at 27:40). Is it that when you cbet and get an A turn you still want to have a decent amount of Aces in your range?
11:50 JdT seems like one of the best hands to bluff xR here, is it not?
20:10 BB is really aggro, but he's not 3betting more than 16%. I thought less than 16% is definitely not on the aggro side
Feb. 7, 2015 | 1:38 p.m.
Thanks, Ben!
Aug. 31, 2014 | 3:18 p.m.
Ben,
Thanks for the response.
You're a good teacher, throwing back a question at me that makes me address the problem holistically. :) How much "giving up" on the turn is too much for the 3bettor? I would guess when checking the turn makes it immediately profitable for the villain to bet with most/all of his range and/or very often allows him to get a free card with a range that's worse than mine. This can be due to me not 2nd barreling enough when I have sufficient fold equity+equity and/or having too weak of a range that got there on the turn.
I realize that when I have a fold equity of 50% this becomes a very profitable bet. I was wondering you agree with the assumption I made about his turn range and his turn folding range.
Good point with CREV, it's a shame there is no game tree software available for PLO which is my main game. Any chance on making a similar HU PLO video in the future? The day will be praised.
Aug. 31, 2014 | 12:27 a.m.
I have always had trouble with giving up too many turns as the aggressor so I'm only going to focus on the A9o hand at the end of the video. What % of his range are we folding out here? My assumption is apart from 44-66,3x and AT we're not folding out better. I would also think most of his open ended straight draws fold as well. Is that fair?
Rough turn range I come up with: 88-44,AJs-ATs,A8s-A7s,A3s,KTs+,K8s-K7s,K3s,QTs+,Q8s,Q3s,JTs,J8s,T7s+,97s+,76s,65s,43s,AJo-ATo,A8o,KJo+,QJo,JTo,T9o,98o
Rough folding range to our 2nd barrel: 66-44,22,ATs,A3s,KJs-KTs,K3s,J9s+,T9s,65s,43s,5d4d,5s4s,ATo,KJo-KTo,JTo,T9o
So roughly 50% of his hands fold the turn.
I was surprised to hear you think his call with AJo was a clear mistake. But after thinking about it it can only be somewhat good if you 3bet all of your combos of 65,T9,J9,64 and cbet them on the flop 100% of the time and check a decent % of your broadways on flop, correct? What kind of strategy would you suggest for villain on turn with his range?
Greatly appreciate any response.
Aug. 30, 2014 | 5:45 p.m.
On Stars in the last 900k hands (since January the 1st 2012) a certain datamining site has only 5k hands of you HU, which is close to 0.5%. I guess you must have played those 300k HU hands on a different site?
Aug. 30, 2014 | 2:54 p.m.
He probably didn't like the video. :) What % of your volume is HU?
Aug. 29, 2014 | 4:32 p.m.
HU. BTN opens, we 3bet (lets say we 3bet the top 20%). Flop comes down XXY. Say both players are equally likely to end up with trips here and that's 20% of the time.
I'm trying to come up with something that could remotely try to resemble a GTO line for both sides. Since our range will contain a lot more high overpairs, assuming a non-wet board our opponent will for the most part have a {trips+,air} range and we have {trips+,overpair,air}. I guess our sizing should be pot, because we would like to include as many bluffs as possible while maintaining optimal bluffing frequency (so 1/3 of bluffs preferably those with some backdoors, 2/3 overpairs+trips). As we are up against a {trips+,air} range we don't need a check-raise range. Check-call some trips and stronger overpairs, check-fold the bluffs (which I assume should be less than 50% of our checking range).
I realize this is simplifying things to the extreme as it is not taking into account stack sizes, flop raises and future streets. I also imagine you can create a fairly balanced checking strategy. Or one where you bet a smaller sizing, but include a check raising range. However I'm wondering if you see any glaring flaws in what I come up with. Obviously I won't create a GTO strategy, but I'm more interested in improving my ability to create strategies that are fairly difficult to exploit regardless of how our opponent is playing. And wondering how you approach these sort of boards (hit rarely, but when you do hit hard) in general.
Also an easier scenario - this time the flop comes down 722r. The 3bettor has WAY less trips+ in his range than the caller. Does that mean he should be checking his entire range here?
It's Deepsolver. Highly recommend, IMO they have the best nodelocks out of any tool
July 25, 2022 | 9:31 a.m.