Jan. 15, 2021 | 5:03 p.m.
If you know a lot of MTT regs, you could ask them for their HHs. I think you can even buy some MTT HHs. Just know that you are breaching site TOS if you use that DB for your HUD while playing (you are probably still violating even if you do not use it for HUD).
If you had enough HHs from the same site(s) you could even try to filter out a big part of the regs.
As to your exmaple with the 38/4, sure you can do it that way. But even if he gets to the river with range x, that you approximate better, than what PIO would output if you let it run on the flop, his river response will still deviate from the PIO strategy heavily. So if you do it that way, be sure to nodelock his river strat as well.
Personally I would think all this work is not worth the hassle, but maybe I am wrong here.
Jan. 8, 2021 | 1:46 p.m.
I do not know whether he is still active here, but I suppose Nick Howard could share some infos on the methodolgy of how to go about what you suggest. As far as I know he and his team have used a huge DB of hands to analyze pop tendencies and compare them to PIO outputs. Based on these results they developped strategies for specific lines, where they saw people being to value- or bluffheavy or under-/overfolding.
As for your approach (putting in assumed rec ranges manually for very specific spots) there is always a bias coming from YOUR assumptions about their range.
I do not think that it is possible to run scripts like this (everything is possible obv, but I suppose it would be a waste of time).
I get the feeling that you are somehow stuck with the idea, that by finding a good way to put rec ranges and sizes into PIO, you could find the holy grail of how to exploit them perfectly. The problem is that rec /= rec and you will end up just searching for something that is not there. Just by looking at standard sims you should be able to make educated guesses, how to deviate from solver outputs when a player is doing something very differently from what PIO suggests constantly. And if not sure, you can always nodelock.
Jan. 8, 2021 | 12:45 a.m.
I would either make it a tad smaller or if you are not comfortable playing postflop, just jam pre. On the flop it does not make any sense to jam, you fold out everything that you have beat and get called by everything better (yeah he might sometimes call QQ and sometimes fold A4s or something, but that will happen very rarely compared to him folding lower pairs and calling you with Ax and sets). Flop you can Cbet very small or check and then proceed from there. Obviously the spot sucks, but I do not agree with you that most players will hammer away with smaller PPs so when facing a lot of aggression you can let it go.
Jan. 6, 2021 | 12:41 p.m.
I dont think it makes sense running sims that already stray far away from GTO by using wacky ranges or settings. There is a huge spread of weaker players with a bunch of different tendencies and you will not be able to cater to each and everyone of these by trying to approximate your sims to somewhat match their strategies. Even if you set the sim up what you perceive to be settings that could represent their actual play, the results that PIO will spit out will not be nowhere close to what they are actually playing.
After running sims that are closer to GTO play you can still use nodelock if you have a hand vs a rec and have the feeling he is playing this and that way, how can I exploit that.
Also keep in mind that if you were to just replicate PIO strategies w/o any exploits at all, you would completely crush, then add just a little exploit based on intuition or because you have a good grasp of an opponent's tendencies and you destroy even more...
Jan. 6, 2021 | 12:33 p.m.
Did you use FGS for that ICMizer sim? Looks like you did not and it will widen the shoving range a lot if you use it.
Also there is a flaw in your logic with the 97s. He will get called by one of the other players anyway, so why let them have the chance to win his chips when you can clickback over his jam and have them fold a ton in a spot where they are highly disincentivized to get involved.
Really good series and very well played MTT, you definitely earned that score :) Really nice calldown with KQ BvB vs CL and also great fold with the set (I think it was 55) in one of the earlier parts.
Jan. 6, 2021 | 2:30 a.m.
Hi Kyriakos thanks 4 the vid,
great that u found a topic to talk about that did not receive much airtime on RIO so far!
6.50 You elaborate that BB does not want to have a x/r range on most boardsbecause of how wide his range and how strong the IP players's ranges are (and also OR range being uncapped).
At what stack depth do we start implementig a x/r range? I imagine that the shallower we get, the less important it becomes to protect ourselves from multiple barrels or flop 3bets by the opener and the more focus we want to put on gii as much money as possible when we are an EQ favourite.
Dec. 21, 2020 | 11:31 a.m.
If we have a non-AI SQ range we should make it a tad smaller and definitely not have KTs in it. I guess it would be something like QQ/KK+ and maybe AKs and then you just use A9o and KTo at some frequency or something along those lines.
Dec. 5, 2020 | 2:20 p.m.
You deliberate whether the opener would jam or flat 77-99. How would you split your range? Jam TT-QQ and AKo always I suppose, mostly flat KK+, AJs-AQs and then mix AKs, KJs+. But what about the rest of our range, SC, weaker suited Ax and broadways and PPs
Nov. 26, 2020 | 9:40 p.m.
Yup, I have to agree, this Series is probably the best content I have seen of you on RIO and you did not even use any software. The quality of your play also reflects the quality of the vid.
Pretty incredible also that he found the fold with QQ ingame vs AA, just goes to show how good of a grasp of Cortez's ranges he has.
Nov. 7, 2020 | 1:29 p.m.
Once againg thanks for the gr8 vid!
36.00 AKo table 4
At that point and time ICM considerations start to factor in into your decision making. Say it was in an earlier stage of the tournament, would you still lean towards pure folding OTT? Don't you think (and in theory this might be a thing?) villain should sometimes show up with KJ here? Plus on this board he can have tons of strong draws that are kinda fine with b/calling if you jam. So all in all your EQ vs his turn betting range, where his only value is KQ and 77 (maybe a very rare 76s) should be good enough to warrant a gii, don'tyou think?
Nov. 4, 2020 | 9:51 a.m.
Thanks for the vid Luc!
You say 4b/jam is standard and I agree that most players would shove in that spot, however I could see a solver wanting to flat here, as he is not 3b/c TT and so we end up folding out hands that have ~30% EQ vs us and get called by stuff that has us crushed (+AKs).
Do you think even considering all this, shove might still just be better because of ICM considerations?
Oct. 8, 2020 | 9:45 p.m.
Mostly people are using Monker generated ranges or ranges they bought from other coaching platforms I would imagine. You could also run a DB analysis to see which opens are profitable, but for MTTs you would need a huge number of hands.
Suited Ax just perform well enough EQ wise, you can overflush and you have the best possible blocker.
KJo and 66 also clear opens, 97s can be opened if the players behind are tight/passive and you have an edge on the table in general.
Oct. 8, 2020 | 3:03 p.m.
The main difference between the 2 spots is that IP you do not want to b/f a hand like 74 here, whereas in the spot in the video you can not x/c it, so PIO wants to rather bet it than havin to x/f (for obvious reasons you can also not check all these types of hands with the plan to x/r).
Oct. 6, 2020 | 8:48 a.m.
Kinda strange question, if you have a good estimation of your ROI, you should also have an idea of how many average entrants the tournaments you play have and what the payout structure usually looks like.
Then you just see how hard you can swing over a sample you are planning to play (I would use the 95% confidence interval) to determine what the highest BI is, that you want to play with your current roll. Remember that you can and should pretty quickly drop down in stakes if you choose an agressive BRM and loose some.
Some more factors to consider: it is really hard to estimate your true ROI, so it is usually better to be rather pessimistic and conservative here. Also avg field size does not tell the whole thruth. You can get to an avg fieldsize of 500 by playing 10 MTTs with 500 entrants or 5 with 100 entrants and 5 with 900. The latter option will yield a higher variance...
Sept. 24, 2020 | 5:07 p.m.
K6o performs really well as a shove. In this spot your jamming range should be a bit wider than Nash, because the BB should call a bit tighter. However if you think the player is not ICM aware and wil call too wide you should jam tighter. I would jam K6o in any case, as it just performs too well as a jam.
Btw I would never fold, if you think he is calling way too wide, then you can still limp or openraise.
Sept. 24, 2020 | 4:56 p.m.
Your txt file should look like this:
So only flops and weights, nothing else.
Also why dont you just use the presaved flop subsets, they will show better results in aggregated reports than just randomly generated flops (although 10 is too little anyway to get any kind of proper results).
Sept. 24, 2020 | 4:50 p.m.
If we assume that UTG is a somewhat competent reg, we can also assume that he will be opening quite wide. Our hand is a 3b/gii for chips vs a normal CO range and vs a presumably wider than normal one we definitely want to 3b. Also the payout structure is really top heavy, so ICM is not that big of a factor and whe should just try to pick up chips.
Sept. 11, 2020 | 11:08 a.m.
Hi Daniel, great vid as usual!
I am currently renewing my DB of PIO sims, therefore running a lot of scripts and have mixed feelings after this vid. On the one hand I see the very valid points you make, on the other hand it is a) a pretty big investment of time of running a "pre-sim" on a huge flop subset, then try to create groups of boards that use approximately similiar sizes, then rerun it, validate it.... and then also implementing it into our learning routine, esp if one were to use simple gto trainer and b) as Mathias has pointed out, when using many different sizings it is much easier for us to make mistakes (ofc this is relativized a bit by our opps making more mistakes as well, when they encounter unusual sizings). So what is your take on that?
I know these are very broad questions and they boil down to some extent to how much time, energy and effort one wants to invest into studying, but this is not me being purely lazy but also trying to find the proper balance in increasing effeciency when learning.
Sept. 3, 2020 | 3:43 p.m.
On two tone vs rb boards sizes tend to get smaller because a) even with a large size you do not fold out FDs (esp OOP) and b) with the smaller size you get more folds on two tone boards as villain has less BDFDs.
Sept. 3, 2020 | 3:32 p.m.
Great vid Sam!
7.13 BB strategy after x/x x/x
We see that BB is giving up with a lot of hands that have 0 potshare (42s+, 52s+) on a runout that is rather favorable for the BB range. Would you vs most human opponents still take these hands and use them in a block sizing, hoping IP has checked back maybe too many A highs and to fold him off PPs <6 (I guess PIO is mostly giving up with these hands, as they are blocking a lot of the folding range)?
Sept. 3, 2020 | 1:32 p.m.
Kyriakos Papadopoulos I did not run this spot in a MW solver, but HU we definitely want to bet for a large size here. As you have said, our probing range is very narrow and consists of only 8x and bluffs. It being so polar means we are going for a big/huge size. Us not probing many of our draws is not a reason to go small here. I would also guess this dynamic does not change much 3way, as the flatters range is pretty capped and by going big bet, jam river we are getting Kx indifferent.
Sept. 1, 2020 | 11:58 a.m.
thanks for the video!
Some notes about your intro. Doubling up in the first hand of a 100$ MTT is worth around 90$-75$, rather than 10$. And yes, decisions at FTs are worth a lot more money obv, but they also come up far less often than spots in the beginning of MTTs. So while a typical error in the beginning of a 100$ MTT might cost you 5$ (just shooting random number here) and an error of the same magnitude at an FT will cost you 500$, the error in the beginning will be happening so much more often, that the EV loss might actually be much closer to the FT spot than one might think.
Aug. 11, 2020 | 2:37 p.m.
Thank you for the always insightful vids and funky lines Ben!
32.40 You keep talking about OOP using an overbet, while it actually is a PS bet. Pretty interesting that IP overfolds by quite a bit vs it (61% folds vs what "should be" ~50%) and is also folding some really strong hands i.e 2pairs. I do get the folds from Q7 and T7, as OOP is predominantly using 7x to bluff here, however I do not understand why Q8 is being folded at a frequency, not blocking bluffs really? If this hand is a 0EV calls shouldn't we be looking to call it at a 100% frequency in order to reach MDF?
Aug. 9, 2020 | 12:29 p.m.
thanks for the great video as usual :)
16.30 Do you even float unpaired hands OTF here, like AJdd? If not, I guess you have to turn 6x into a bluff and thus the only sizing here for you should be 2x pot jam? Meaning for value you only get to use FHs and maybe KQ?
Aug. 8, 2020 | 3:15 p.m.
Also here is BB strat vs open, I only allowed for call or jam. As one can see, BB is overfolding vs cEV (continuing ~52% vs what should be ~66%), but not as much as one could expect.
Vs jam BB is calling 66+,A7s+,KTs,ATo+, mixing KQo which is a 0EV call. Vs an ATC jam BB was calling 66+ A9s+ AJo+ KQs.