Podcast: Play in new window | Download
Gareth Chantler joins us from the Dominican Republic to talk about his new life as a writer (a discussion that began during his last appearance on the show) and his WCOOP so far (as of the time of this recording, that is).
Timestamps
0:30 Hello and Welcome
33:41 Strategy
Strategy
I enjoyed hearing Gareth’s adventures playing WCOOP from the bakery in Dominican Republic and especially enjoyed him talk about his time in Ireland. My ears perked up even more when he mentioned Maeve Binchey, and then Andrew mentioned me! Luck of the Irish, I suppose.
FWIW I was planning to hear the rest of the interview with Alex Welton, as I was curious as to who was taking Andrew out for his birthday in Montreal, when I saw Gareth’s interview.
Good luck to Gareth in writing and to Nate in reading about math. Follow your bliss!
Lots of interest here:
I’m a twitter fan, but it can definitely be an attention soak, along with the rest of the internet. But the best way to read twitter is with aggressive use of the ‘list’ function, and one of the apps like hootsuite, so that different themed accounts sit together, and the poker players can lie way over on the right to be visited if & when you can be bothered.
And regarding the tipping, my bet is that the guy didn’t want to feel like he was subsidising you/letting you get away with more net cash than him. He’s already given you a generous chop, why would he let you get one over on him wrt the tip.
I guess you have probably encountered this in your trawls of Irish lit, Gareth, but I just read Troubles by PG Farrel(?). It doesn’t have the fantasy element of The Third Policeman, which you mentioned before, and it’s uneven, but it’s much of the same sense of the absurd, I think.
Oh, and thanks for the thanks, Gareth, and thanks for feeding the cats. Next time come to Wales when we’re actually around.
I am in your and Emma’s debt and am therefore at your disposal, should you have need :).
Hi guys, love the podcast. Re pressure on Gareth from the staff in the Fitz to save the bubble – this is probably a cultural thing sort of, possibly related to Irish culture and certainly the culture of that club. In my experience there the bubble is basically always saved – so it’s more like there’s always an extra spot in the payouts that isn’t listed if that makes sense. So I can see where people get their noses bent out of joint when some stroppy yank comes along and doesn’t want to cooperate (I say this as a stroppy yank myself).
Stroppy as charged. Canuck though.
Gareth saying he loves the Fitz, a shoutout for Mrs Doke, Gareth utterly butchering a hand…..so much to love in this episode
Agreed Doke. My favorites were the sounds from the bakery where my fantasy future wife works and all the imagined passive aggressiveness around the 1 hour mark. I’m sure it was not intended this way, but for entertainment purposes, I decided to pretend that the needles were real and deep.
Where will the people go to get that Gareth Chantler feeling without twitter?
Deepest apologies Gareth. I’ve made that error more times! If I had called you “American” instead of “Yank” I maybe could have got off on a technicality. Irish people nearly always ask me if I’m Canadian first so they don’t risk the faux pas.
No bother at all. Funny about that in Ireland. Many a cab driver in Dublin has fallen over themselves apologizing to me for going the other way. My accent doesnt make it that easy I guess.
In the blind battle hand did I hear y’all correctly?
I think Gareth said that Villain is likely to be Nash-shoving his hard to play hands and raising with hands that are easier to play post-flop. I am confused.
Why wouldn’t he?
Because that is very exploitable.
You can’t pick and choose some hands and not others from a Nash chart (unless you are assuming that Villain will play Nash in response, which I don’t think is a reasonable assumption).
I think very is a silly word to use here. Villain has an open shoving range, a raise/call, and a raise/fold range. He can get away with the latter by including his strongest traps in the raising range. He needn’t dilute his shoving range much because some of his hands naturally realise their equity better preflop than others. What I was saying in the episode is that I have a strong belief he would raise/call AA and open shove K5o and probably, as Nate mentions, raise/fold T5o. I don’t think this is “very” exploitable on this stack size. In game he is going to look at what he can shove on the chart and if the hand sucks postflop go with that. I’d certainly wager you’d have a hard time exploiting that in game to a great degree.
Thanks. I think I’m starting to get it.
I didn’t realize a hand like K5o could shove 16 BB without the strongest hands in the shoving range.
I do still think that Villain can’t just read the Nash chart and use it directly to chose his shoving hands if he’s not shoving all of the stronger hands as well. But, I think I’ve severely over-estimated how wrong that would be. The Nash chart may actually end up being a halfway reasonable approximation.
When Villain shoves hands like K5o, but not hands like AA, then Hero can call more often than Nash but Hero’s range will have less equity against K5o than Hero’s Nash range would have. Villain’s higher equity helps balance against the lower fold equity.
Something worth looking into is the Sklansky Chubkov rankings.
Nash assumes an uncapped range as you say, whereas SC numbers assume our range is that one hand and we are face up – and he designs a perfect calling range against it. So SC is also wrong for this situation but in the opposite direction to Nash. SC says K5o is a jam up to 12BBs even face up. With the protection of the pocket pairs and some bigger Kings and Aces we can go to 16 – presumably there are also antes to win which we can add to the size of the BB.
I agree that it can be useful to know a thing or two about the SC rankings, but I think their primary utility is to give you lower bounds on depths at which you can jam certain hands. That is, they’re good for telling you something about the hot-and-cold strength of a hand, but I wouldn’t rely on them for much more than that. I think that Will Tipton’s criticisms of SC numbers (in v. 1 of his book) are accurate: why look at a certain approximation or heuristic when we know so much about the optimal ranges themselves? It’s a bit like a SNG player worrying too much about ITM% when he could just examine ROI directly.
Again, though, I do think that SC numbers have educational value.
Thanks Richard!
I knew about SC but thought it was only a hand ordering system. I didn’t know it included profitable stack sizes for shoving.
Nate-
I’m sure that I’m currently folding some of these hands (LOL me, I know!) and the chart is a super quick and dirty way to correct that.
Yes. Some of the shoves may not be unexploitable in a vacuum, but I think the general idea holds that Villain will have both a shoving range and a raising range, and the shoving range will consist of hands like weak Ax that have a lot of equity but don’t play well postflop.
This makes more sense.
Villain shouldn’t choose his shoving hands from a Nash chart.
He can possibly shove a smaller-than-Nash range of hands. Although my instinct is that he’ll have to include a reasonable number of his strongest hands in the shoving range if he isn’t trying to play exploitatively.
I did some math and I now see that Villain can unexploitably shove a hand like A2s for 16 BB
So, Villain can construct an unexploitable shoving range that doesn’t include any of the top hands.
Right, this makes sense to me.
I think I assumed Rant you were saying that an unexploitable strategy in some way can’t involve a splitting of ranges here. Like it seems to me that the solution to this game of it folding around to the SB with 16BB effective in BvB would involve the SB splitting their range into shoves and raise/calls and raise/folds and open folds. We can’t solve this of course but I would be contend that this is the solution against a perfect opponent in BB.
I wish I had more time to work on this stuff!
I was confused by the idea that the Nash chart could be used piecemeal. I think it can’t be used directly but I now realize that it may be close enough.
I don’t know if significant range splits like you suggest are possible or not.
I think I’m convinced that SB shouldn’t open to 14 BB with a 16 BB stack. So, there probably is some max open size if he isn’t shoving.
I think I’m convinced that there will be many opening sizes — all the way from that max size to a call.
I haven’t thought about this enough to have an idea about what happens to the hands that can’t quite call or shouldn’t quite bet the max non-all-in size.
I also don’t understand the factors that make AA a shove vs. a raise. My intuition is that AA is probably going to be a shove most of the time but my intuition is evidently often wrong. 8)
Great podcast. The second hand drove home a point that has been made on the podcast before – using unbalanced sizing against an unknown pro (i.e., Gareth) can cap your range and you wind up getting badly owned when your bluff is snapped on the river.
There needs to be some more conversation on this range splitting in shove spots stuff. I’ve heard apestyles talk about it a bit as well, but I’m still not sure I have a full grasp on it after reading these comments. I could be wrong, but what I hear is…
By not shoving the top, you make shoving the bottom bad. So you can no longer shove the bottom and instead, you raise fold it while raise calling the top. The middle is vastly unchanged. Gotta give more thought as to how villain should adjust to this.
Carlos, That’s the basic idea. If villain knows we wouldn’t shove the top of our range then we are capped and therefore he can call wider, so the bottom of the range becomes unprofitable – but some hands are still profitable anyway.
With SC numbers it’s important to know what they are and aren’t, and some of the limitations they have resulting from how they are derived. SC assumes Villain plays perfectly against one hand, so with 87s it assumes Villain will call things that crush us like 97o – totally unrealistic assumption so 87 suited’s SC number of 5.5 is nonsense – unless the wind really has just blown our cards over and we are shortstacked in the small blind.
On the other hand, if we are 20-25 BBs deep and we are minraising or folding most of our hands, we might not want to open small pocket pairs because they don’t make good c-betting hands, we have to fold them to 3-bets and we don’t have implied odds – SC gives us the clue that we can just open-jam 77-22 (depending on position). We are face up and he will call with maybe 66+, but we are still profitable because our push gets through often enough (SC actually assumes he calls with things like T9s that we are flipping against because they have odds but that doesn’t change our equity that much from reality). Also it’s important to understand that in this situation 22 performs more like 44 or 55, because Villain’s real calling range is more like what he would call against those single hands so you can put 22 in a range that the SC number wouldn’t necessarily allow.
Rather than using SC numbers, the actual proper way to really construct a capped range like this would be to have two instances of Equilab (Pokerstove was before my time) running. In one you have Villain’s calling range, which you test individual jamming hands against. In the other you have your jamming range which you use to test candidates for Villain’s calling range. Start with the best hand in the jamming range (77 in my example) and put all the hands in calling range that beat it (or have odds). Then test another candidate jamming hand (66 in my example) – if it has enough fold equity (look at the total percent Villain(s) is/are calling*) and all in equity then you add the new hand to the jamming range. You may now find new candidate calling hands have enough equity to go in the calling range. Keep going till you can’t add hands to either range. (If that sounds complicated you can test you are doing it right by using the same method to derive an uncapped nash range – e.g. what you can open jam 30BB from the SB – and compare your answer with a web calculator)
One of the main use of SC numbers though is settling discussions in a situation like this one, which can otherwise get hand-wavy. We know K5o would be part of a 16BB Nash pushing range, but is it still pushable when SB is capped and it doesn’t have the protection of other hands? Well it has an SC number of 12BBs – means 8 times the total dead money in a tournament situation. So as long as the total antes add up to at least one small blind, then there is enough dead money to push it for 16BBs even if our range is totally face up as just that one hand – end of discussion. SC numbers shouldn’t be used to say particular hands can’t be jammed though, because of the limitations of how they are derived.
* How to account for the effect of blockers with this method is left as an exercise for the reader 🙂