yea I have always thought about streaming on twitch. Not sure if I want to play my usual stakes at Twitch though, would be -EV to allow people to see my cards.
Notifications
Clear all
So I played my first "real" tournament today, the Macau Poker Cup Main event $12 000 HKD Day 1c. Somehow I managed to survive with a 29k stack. The average stack was at 44k.
The previous day, I had dinner with a friend. And I started hearing stories about a Mega HK Boss who regularly pops up and drops 300k-1M each time, even at 100/200. Everyone was going on about how he was doing stupid shit like getting it in for 500bb with top and bottom pair on a mono Flop. So I set my alarm at 9am and made my way to the Wynn. A single table was running, and I started another Must-Move table with 3 other guys. The other regs were camping like Hyenas.
When the boss made his way to the table, the table filled up within 2-3 min, it was as though they were all using the live version of a seat script. The 3 of us who started the game were pretty upset when we were moved to the Main game. The boss was 4th on the list, and we had 3 players who wanted to leave the game. All of us paid them 1k HKD each to fuck off. Then the rest of us flipped for 1k each (9k) in the pot. Guess who won the hand? 
I had QT, board runs out Q75Q6.
I saw another friend and he mentioned that he was heading to City of Dreams to play MPC. I initially wanted to play that over the weekend but I was exhausted after Thursday's 12 hour monster session. Since I was mega-tilted, I went along with him to chase my losses and hopefully bink a tourney for once.
My table was pretty decent. We had a retard Jap reg who was leading and raising with random crap. He was soon busted by the White guy soon after (I think he's American). After that, Mysters_Y joined the table and sat on my right. She looks a lot younger than in her old Team Online photos. Apparently she's living in Bali and playing mostly on FTP and Ipoker now. She was the 3rd to bust.
I managed to build a decent stack for myself but was crippled by some spewy chinese reg.
Effective stack sizes: 24k, 300/600/25a
BTN raises to 1500, I 3b in SB with TT to 4000, he calls.
Flop (9.5k?) - AXX - I bet 3300, he raises to 8k, I call
Turn - He jams on a blank and I fold.
The table broke soon after. Here are some more all-in hands
1. I jam 16k (M = 6-7) in CO with AK, BB calls with 77 and I win the flip
2. I minraise to 2500 with AA (40k stack) from UTG, zcedrick (strong spanish reg) flats with AQ, 77 guy jams 10k with TT. He sucks out with a set.
3. I flat a CO or BTN minraise with K2hh. Fat Chinese lady bets 3k into 4k, I checkraise allin with my 18k stack on a J7Xhh board and she calls with AT high. I guess I almost always have a FD there but I could very easily take this line with a pair+FD or KJ/QJ. I was actually considering whether to raise smaller or not to rep a stronger hand but stack sizes would be pretty awkward on the Turn. I bink the King to double up.
4. MP minraises to 2500 at 600/1200/200, Fat lady calls in HJ, CO pushes for 6k+, I have JJ in the SB and I wasn't sure whether to jam or call. I ended up calling with a 40k stack which kind of seems like a mistake? Makes more sense if the 3b was to 10k but 6k is effectively a min-raise. Zcedrick jams over and he covers me. I had to call 33k more to win a pot of ~90k. I still have no clue how ICM affects this spot. But anyway Zcedrick heard me saying that I wanted to jam instead of call (he knew I made the mistake) and he jams anyway which made me feel like he has QQ+ or AK almost always. It's kind of break-evenish based on pot odds, maybe with ICM factored in it's a fold?
He shows AK, other guy shows AQ and the King hits on the board. I suppose I was lucky that I folded.
After bleeding for a while, I end the day with 29k. First decent showing at a major tournament I suppose.
Zcedrick also pulled out a pretty crazy move which led me to believe that there's still a huge amount of room to push edges. He made 2 really great plays.
1. BTN (Zcedrick) raised J9s, SB (Chinese reg with a mega stack, 3X+ the average stack size) called AJ.
Board, K:d Q:s 7:d T:c K:h
Flop: BTN bets, SB calls
Turn: SB makes a mega huge check-raise
River: SB bets 21 into 38k and BTN open folds J9s.
2. UTG (Zcedrick) opened T8s and 3-barrels (45%, 65%, 150%) on a KQX ♠ ♠ XX runout. The chinese woman tanked for really long and folded (showing a King). Why bother? Most obvious KJ/KTs ever 
Zcedrick opened his hand and made some cocky comment. He also called the Taiwanese guy wearing a patch (I think he was playing on some site which Raymond Wu was promoting) and the other fat Chinese lady fishes and that the woman should fold vs those 2 people. It was hilarious and honestly I don't blame him for being arrogant. Just that I would probably be less confrontational about laughing at live fishes ♦
Originally posted by mbml
yea I have always thought about streaming on twitch. Not sure if I want to play my usual stakes at Twitch though, would be -EV to allow people to see my cards.
hi mbml, you can set the stream to show after 3 minutes the hand is played. Everybody is doing so.
Originally posted by dreamtime
Originally posted by mbml
yea I have always thought about streaming on twitch. Not sure if I want to play my usual stakes at Twitch though, would be -EV to allow people to see my cards.hi mbml, you can set the stream to show after 3 minutes the hand is played. Everybody is doing so.
You can also cover up the cards, like this german coach does: http://www.twitch.tv/likandoo/b/634393845
(I think he reveals them in certain spots)
I tried streaming a few days ago but it didn't work. Will consider streaming in April when I return to Macau. I flew back to Singapore yesterday and I'm spending a lot of time thinking what I want to do here. Just bought GTORB (Finally!) as well as some packages.
Getting a headache looking at all the ranges and mixed frequencies. The interface is certainly way superior to the other alternatives but it is still a little laggy which I dislike.
I'm looking at all the popular solutions in the Flop Library for now. Some things make sense, and others seem kind of counter-intuitive.
5h2s2h board, BB vs BTN, 65% Flop Cbet size
As BTN
1. BTN always bets 5X and 66-88. That makes sense, these hands are strong enough to bet for value and don't really want to give a free card.
2. BTN never slowplays 2X. However, he does check back 99-AA sometimes. In fact, he checks back KK and QQ more than half the time, which makes some sense. These hands are strong enough to go for 3 streets of value, and are not too vulnerable. AA is checked less than KK/QQ, I find this part counter intuitive. Would imagine that KK/QQ are better for betting because they are slightly more vulnerable and they have more value against Ace highs, which AA blocks
3. BTN checks back most Ace highs. He usually cbets AK for thin value, and also cbets AQ without a heart most of the time. Most of the mid-strength AX hands are checked back, A6 and A4 are bet slightly more often.
4. We see a similar pattern for the King highs. Medium strength Kings are almost always checked, KQ is cbetted sometimes, K7 as well. The algorithm is using a polarized cbet strategy. Which I suppose makes sense, you still want to have some AX and KX in your cbetting range (which is what old school players would have, but guys who are checking back Ace and King highs too often will find it hard to rep those cards on the Turn)
5. For A and K highs, we are often checking back the Ah, Kh or KsXs/AsXs hands. I would typically bet these myself but I think GTORB wants to allow BTN to continue vs turn leads more often and be able to show up with more backdoor stuff.
6. Flush Draws: Nut flush draws are rarely slowplayed (<5%), AhQh is checked 17%, KhQh is checked 70% (?). QhXh and KhXh are bet most of the time as well. Small flush draws are checked back between 30-70% of the time. That makes sense, because it's hard to bet/call the Flop with a baby flush draw profitably when your opponent is check raising some higher flush draws a lot of the time.
7. Medium Strength hands: 44/33 are bet half the time
8. Overcards: Bet ~half the time
Comparing GTORB ranges with my own
1. My checking range is way too weak, and too highly weighted towards Ace highs. Not enough slowplays
2. I'm probably betting X:s X:s too often and GTORB recommends checking back these hands a lot. Same goes for X:hX
As BB
1. BB leads 17% of the time on the Flop. He does so with some flush draws, some backdoor stuff, and some top pair type of hands which bets mostly for protection
After checking,
2. BB folds 53%, raises 20% and calls 27%
This is an interesting result I must say. I'm very accustomed to raising a very low frequency and I think it makes a lot of sense to fight hard for such volatile boards with thin value checkraises as well as fastplaying most 2X hands. I'm kind of surprised at the extremely high fold to F cbet frequency of 53%. 1-a% tells us that the min defense frequency is at 60% vs a 65% pot size bet, obviously not too applicable for Flop play but it always serves as a reasonable guideline. I guess we can defend less vs a larger Flop bet size, and we are at a significant range disadvantage here with far fewer 2X and over pairs compared to BTN. It's also fine to defend far less when half your defending range is going to raise the Flop. I already had the feeling that 1-a% was wrong about half a year ago because I noticed that most of the strong regs were already starting to tighten up their Fold to F cbet frequencies back to ~45%.
3. Trips: 2X is raised ~half the time
4. Gutshots are raised half the time, with a slight preference for AhX combos
5. KhX and QhX are raised 20% of the time
6. Flush Draws: 7 high to J high FDs are raised between 20-50% of the time
7. Top Pairs: A5/K5 is raised half the time, weaker 5X hands are rarely raised (20%), 54o is also raised half the time. 77/66/44/33 is raised 20% of the time
8. Ace highs: Some Ace highs are raised, some are called, some are folded (some combos of A6/A7). I pretty much never fold an Ace high here.
9. King highs: King highs are folded if they do not contain a heart.
10. Queen highs: Queen highs are often folded. They are defended more often with Qh, and sometimes with Xh, but usually folded without a heart.
Note: The Flop Library is using snowie ranges for this, which I feel is way too tight. Most guys open 55%+ in my games and therefore are defending much wider as well in the BB.
Main differences:
1. I'm calling too much with weak overs and am very susceptible to multi-barrels on low volatile boards.
2. Not really raising much
nice post
yeah the snowie ranges threw me off on that one
BB cold calling preflop all that 2xs (whilst BTN doesn't) gives him a deceptively strong, polarized range. So that leading range (and aggressive x/r range) is an anomaly, if guys aren't defending all 2xs.
re point 2:
In most solutions I've found that AA is checked back more like you suggest. I'm thinking this is probably just a solution specific issue where something like QQ and KK block some hands that x/r or something, but AA doesn't. BB's aggressive x/r may add weight to this idea.
Re As BB point 2:
I've never really liked 1-a. I think if we weight x/r higher (I don't know maybe 2 x/r to each x/c?!) then defending 1-a starts to get a bit better.
There's a new stat for me to sell. I'll call it the Jules Continuing Stat (JCT) :f_cool:. It's x/f flop, but gives a 2-2.5 weighting to x/r. So in this scenario of: BB folds 53%, raises 20% and calls 27%, JCT would be ~33%.
@jules: what do you mean by JCT? 100 - 20(2) - 27 = 33%?
I bought the 3b pot strategy pack too. And I see that the optimal solutions all involve betting OOP super often. Before Alex Sutherland started showing all the various Cbet frequencies on various boards, he asked the viewer to guess the EV and Flop Cbet frequency. I assumed that EV and Flop Cbet frequency are directly proportional to each other but it seems that there were many other factors in play.
I didn't really analyzed the PF ranges assigned to CO and BB but I assumed that K86s > JJ4r > AT6s > QT4s > 965s
3bettor has AA/KK/AK so KXX boards should be the best. JJ4r is pretty much a blankish board, so it favours guy with more overpair combos. AT6s felt pretty neutral but it turns out that the PFC was given a ton of AXs combos. I thought that QXX boards were traditionally better for the caller but KQo wasn't assigned to CO's range. And of course 965s is better for the caller because he can have more set combos and maybe 2-pairs and straights if the BB is using a more polarized range. But in this case, BB was 3betting a lot of AXs and 65s type of hands.
K86s: 98%
QT4s: 79%
JJ4r: 77%
965s: 64%
AT6s: 63%
AT6s (150BB): 50%
So yeah, the GTO solution feels like a step backwards to 2009, where everyone was just C-betting with a really high frequency. And I would imagine that those frequencies make a lot more sense if the 3bettor was using more of a small C-bet size if he were to bet Flop with a wide merged range but the bet size given was 60% pot which is kind of huge for today's high/mid stakes standards.
This seems really different from the GTO Cbet frequencies of 40-50% in Single-raised pots but the main takeaway here is that when you have a significant range advantage and are OOP with low SPRs, betting becomes much higher in EV (relative to other options such as check-calling or check-raising).
Turn Play: Low SPRs with High Flop Cbet frequencies give room for wide Turn check-raising ranges (20%+), that's a lot higher than I expect so I think that's a pretty useful insight to have.
IP vs OOP: Now this part got me confused a little. So some time last year, I felt strongly that betting more OOP makes sense, because it's better just to take down the pot and you don't have the option of checking back to take a free turn card because of position. Then I started checking OOP a lot more often because I was protecting my range more. This was mostly in SRPs though, did not really think about the differences between SRPs and 3BPs. And Cbetting is more effective because the OOP guy will probably find it hard to float wide. It turns out that IP ends up cbetting a lot lesser IP.
Some other note-worthy points
1. Top set is now a lot better for slowplaying as compared to in SRPs. I guess since you always have the option to pump in the money on later streets anyway and blocker effects play a much larger effect when dealing with narrower ranges.
2. Betting frequencies fall by around 30% when ranges become wider (vs BTN as compared to vs CO). Why is that the case? Is it because the 3bettor's range advantage is reduced when the amount of nuts (QQ+/AK) as a % of his entire range is lower?
3. The Caller has a leading range on 965s (14% range), which was what I expected 8 months ago when I was using CREV to analyze these spots in Amsterdam.
I'm also still using kind of the old school polarized 3b range still with a little "semi-bluffs" like SCs and AXs mixed, maybe I should start mixing in more of these hands and reduce the QXs KXs garbage. Will revisit this topic tomorrow, I think spending 1.5-2hr per day is a little too much
Great updates from both of you, keep it up guys!
Galvin, can you explain more this part "2. Betting frequencies fall by around 30% when ranges become wider (vs BTN as compared to vs CO). Why is that the case? Is it because the 3bettor's range advantage is reduced when the amount of nuts (QQ+/AK) as a % of his entire range is lower?"
Everything else makes total sense. You want to x more behind IP as you can raise turns and rivers more efficiently vs OOP, also play more aggressively if opponent x to you flop and turn. Higher the PSR the more difficult it becomes to play the hand by the river. All the TP type of hands become weaker in strength now which is why frequency in barreling is decreased.
Those are some of my assumptions and happy to see that they fit with GTO stuff a lot 
Please provide more examples and elaborate on point 2 you posted with some range and example.
Originally posted by IvicaIliev77
Galvin, can you explain more this part "2. Betting frequencies fall by around 30% when ranges become wider (vs BTN as compared to vs CO). Why is that the case? Is it because the 3bettor's range advantage is reduced when the amount of nuts (QQ+/AK) as a % of his entire range is lower?"
Please provide more examples and elaborate on point 2 you posted with some range and example.
Well u can just measure the equity difference of the 3bettor vs the caller in BB vs CO/BTN and see the difference, then you'll understand what I mean
Originally posted by mbml
Well u can just measure the equity difference of the 3bettor vs the caller in BB vs CO/BTN and see the difference, then you'll understand what I mean
So the basic idea is because somebody opens tighter CO then BTN, his defend range will have stronger hands naturally therefore when we 3 bet CO we should have 30% less frequency OOP vs CO instead of BTN for example?
This sounds very logical, the way you wrote it was thinking it was some brutal discovery or something ♦
The ranges of both players will be wider in btn compared to co. One would expect the increase in ranges for both players to be approximately proportional but thats probably not the case since the bb cbets a lot more aggressively vs co.
Its not that co is stronger here, in fact its the opposite. The 3bettor has more nutted hands (QQ KK AA AK) as a fraction of his entire range as compared to if he were 3betting vs the btn.
Originally posted by mbml
@jules: what do you mean by JCT? 100 - 20(2) - 27 = 33%?
Well I meant JCS, JCT is a typo.
It was just a silly way of me saying that I think (and could be wrong) that one should defend less than 1-a when one has a raising range and not just a calling range. And a better rule of thumb than 1-a, might be to give the percentage of hands that are raised twice the value of those that are called.
So in that example 1-a is 60%. If BB folds 53%, raises 20% and calls 27%,
So here it's defending 47%. But if we weight raises at 2, it's defending 67% (2*20% + 27%)
And JCS would be (100-67%). I just made it it 100-x to make it similar to a 'fold to x' stat.
@Jules: That's pretty reasonable, though I think coming up with the actual defense frequency is pretty pointless now (was really useful when we had to use CREV and input the ranges manually) as GTORB provides the exact ranges for us.
Today I met up with a friend and we spent 3hrs + playing around with GTORB at Starbucks. Gained quite a few interesting insights. We were analyzing 3 main spots. Disclaimer: I can't remember everything, many details might be wrong. Just trying to refresh my memory here and internalize the stuff learnt today.
1. 6max CO vs BB 3bet on 964r
2. HU SRP K972m
3. HU SRP K72r7
964r
I think GTORB solves one of the most important questions of them all. How should we play AQ/AK on such board textures? Back in the day (and currently in live games at Macau), most of the bad/mediocre players would bet AK 100% of the time. These days I tend to check-call AK here a little too often in live games. The ideal frequency is to bet ~50% of the time.
Another interesting insight is that CO should be raising the flop at a very high frequency (20%+), with some backdoor + overs, some 9X/TT, some straight draws.
I didn't analyze too many board runouts but on coordinating boards which bring a 4-straight and 3-flush on the River, the 3-bettor has to check the majority of his range, while shoving some flushes and some flush blocker hands.
HU K972m
This was the most interesting spot of the day. First of all, let's talk about the Flop. The BB is supposed to lead the Flop ~8% of the time IIRC, with some strong made hands, some straight draws without a suit, and some nut and second nut flush draws. The SB is supposed to bet only 30% of the time, with a good mix of random air (30% betting frequency), top pairs (~50%), nut flush draws (80%), second nut flush draws. He ALWAYS checks back 9X and 7X here, and the rest of the hands are played with a mixed frequency.
The PFC calls with any 8:sX FD and above more than half the time, he folds high card with a 6:s and below. He calls any pair+FD and any straight draw + FD. He calls JT always, but folds the weaker straight draws. He calls most second pairs (9X) and folds bottom pair without a flush draw.
Turn: The 4th flush card is a pretty interesting one. The PFR fires all his 99/77/22 combos, checks back his 2-pairs, and plays a mixed strategy with KK. He bets the Turn with A:s and Q:s about 80% of the time, checks the smaller flushes 70-80% of the time, and bluffs with all the other random air about 30% of the time. I honestly cannot remember what the ideal defending frequency for the PFC should look like, but I think his range is mostly 1-card flushes and some 2-pairs/sets.
River: I was analyzing a 5th flush but I realized that there wasn't really any split-pot effects on such a board so it didn't really matter if the board was a complete blank or not. There would be a slight difference on a pairing card like the 9:h but sets just comprise such a small portion of both players' ranges that it probably doesn't matter.
I remember Internet telling me that the ideal type of board texture where having multiple bet sizes would make the most sense would be on a 4-flush type of board so I attempted to confirm this idea. I inserted the option of a All-in bet, 75% bet, 50% bet and 25% bet. The 25% bet was hardly used, the 50% bet was used 6% of the time, 75% bet was used 25% of the time and the all-in bet was used 30%+ of the time. I figured that the 25% bet was probably pretty useless, and the 50% bet could be slightly useful but the overall EV didn't change too much when I compared it to a 2 betsize option scenario of 75% and All-in. So I think it makes more sense to have a normal bet and a shove.
The shove was done obviously with the nuts (A:s) and with random air from all over 25% of the time. Q:s will bet almost always with the 75% pot bet, and J:s will bet some of the time. All other hands will check, and the bluffs will bet smaller with a slightly lower frequency.
While analyzing this particular spot, my friend showed me a hand history between Rui Cao and a Swedish friend involving QQo on a 4-suit club runout where Rui Cao 3X pot jammed the River and he ended up check-calling down and winning the 600k HKD pot. Pretty sick huh
K72r7
Couldn't really find a nice example in the Flop library and the net was lagging so we quickly ran some sims on this Turn spot. So we know that the Turn pairing the 7 favours the PFC a lot more than the PFR. But in the scenario we analyzed (think it was CO vs BB), both players didn't really have too many 7X in their ranges. The PFC ends up leading the Turn with ~5% of his range, which was in line with my expectations for this spot. I would imagine the Turn leading range to be much wider (maybe 15-20%) if the second card was higher, like a Queen or Jack.
Re betsizes not effecting ev.
I think this is the biggest mistake that is going around currently that these GTO solutions will eliminate. Namely bet sizing (apart from not having a shove option to end the hand) doesn't matter. So if someone is using a weird betsize, they're doing it because they're trying to exploit (like we all do against fish). Which is transparent and easy to abuse.
I recently looked at a few of the regs I play with regularly that have better winrates than me. In pots against other regs, their winrates are worse than mine, but against weaker players theirs are better. And it's no surprise to anyone that winrate as a % from weaker players is far > than winrate as a % from pots with regs. So it's time for me to up the spoilts and put the GTO away for a while.
Takes insane reads (like we have vs the fish population) to alter our bet sizing in any spot and make it more +EV then with our regular GTO strategy.
If you do, go ahead, but be extremely careful before making such big assumptions.
I also believe time spent trying to study and make balanced ranges in a given spot is way higher EV then time spent trying to find spots to alter our bet sizing to exploit some weaker regs (especially in 6max).
@dooleslovs: it's a 22% polarized 3b range BB vs BTN, containing K4o, Q6o, a ton of offsuit mid connectors and suited trash like T4s-T2s, 94s-92s, J2s, 85s, 72s type of stuff.
@jules: What games are you playing in at the moment?
@WillyD: Yep you are probably right
I chanced upon this site earlier today. Think it's really interesting how this guy adds drawings to his presentations to make them a lot more engaging. I was kind of doing that with "The Art of Poker" but it was getting way too time-consuming. And I have a drawing tablet now, so maybe I can make nicer stick-man illustrations this time for my new mini-courses.
Anyway, my plan was always to figure out how to do Poker Education well, and then transition out back to Academic Education. Spent almost 20 years studying and I was always really good at it so I figured it would be a waste to not put that knowledge to use.
Some other plans: I'm really starting to consider streaming regularly on Twitch.TV. It's probably a good way to keep online poker interesting for me and also could help to boost my poker business. I'll probably start watching some Poker streams from now on (Jason Somerville completely dominates the Poker category at twitch). Pretty sure I could produce top quality education stuff but the entertainment value is probably way more important when it comes to streaming. I'm no StickyRice or SingSing.
Now to fix that monotone voice ...
A part of me wants to continue playing Online poker regularly. Was actually intending to play live:online at a 4:1 ratio but maybe I'll shift it back towards 1:1. Never really enjoyed Live Poker all that much since it's so slow, but it's always nice to alternate between the 2 options to keep things fresh. I got completely crushed in my second trip to Macau, mainly because I lost 300bb at NL10000 USD and was tilting a little in some later sessions.
Have been working very hard for the past 2 days. I created 2 new mini-courses, which can be purchased separately on my site.
1. Donk Betting on the Flop ($20)
2. Increasing Your Win-rate through Self-Analysis ($60)
I decided to include these 2 mini-courses as bonuses. Students who purchase Yolosaurus Goes All-in will receive these mini-courses, Free of Charge.
The Facebook group is also pretty active at the moment and is proving to be a conducive environment for students to come together and learn together. So get your copy of Yolosaurus Goes All-in today! 
*end promotional message* 
hey,
really interesting posts recently, enjoying reading about the GTORB stuff,
but how can you give away so much information/insight for free? you're a midstakes grinder, and hundreds of players reading this extremely valuable poker strategy/theory stuff will definitely have a negative impact on online poker for you...