nice blog, both the pokerpart and the gambler part. subbed.
Notifications
Clear all
For good customers all draws are with 50% equity. They either hit or miss, and they just need to decide if they feel like it's the time they will hit it. I have also tried to explain that the more outs the draw has, then the more likely it is to hit one of them. Some people don't get it, they say that no one knows what the next card will be.
That's why there will always be players at the tables who calls with gutshots.
Then I reckon that Fusion will be a hit. Its 2 starting cards are less confusing than the 4 at PLO, while the probability of flopping a bare gutshot is high too 
I like it how the number of hole cards increases gradually, making it easier for a new player to read a hand the majority of the time.
Do you think you'd attain 50 bb/100 at the $3 (PL5) Fusion if you tried similarly to your PLO2 experiment?
The rake seems large like at PLO (18 bb/100 in my 144 test hands), and the standard deviation seems at least as large as at PLO with the same blinds, so I think I'd need to be beating PLF5 at 40 bb/100 in order to prefer it to other forms of poker.
Has anyone figured out a good strategy for that game?
Well, actually, PS-gy member jmackenzie (aka jlhmac, aka James) has been giving some strategy tips on his Op-Poker site and Twitch streams, but I can't verify that they're good 
I'm bad at the game, but I'm bold enough to have disagreements with him; in particular, I think the button VPIP range first in should be much wider than 50% because position matters a ton 
But maybe the point is that the top 50% of hands earn 90% of the profit from that position, and playing the other half of hands doubles the variance while not contributing much to the winrate.
I tried it yesterday, it's a bit weird game. You can have total crap on the flop, but turn a monster. Played the lowest limit for a while, and for some VPIP was close to 100. And rarely they folded on the flop. In that game I think the turn is the crucial point, as from that point on it's just normal PLO. Harder part is to figure out when it's +EV to peel the turn and when not. Trying to give range for opponents is close to impossible as calling range can literally be total air. And for some, the betting range is also close to almost anything. And for some, 2nd nuts is just a bluff catcher as they are afraid of the nuts.
But if we start with a preflop, should we play with holdem preflop ranges or alter them in some way?
edit. I got the OP-poker starting hand charts. And I looked pokerstars page for 5 tips for Fusion. I didn't get the part "suited starting hands are worse than offsuite" thing. That doesn't make sense. Or maybe it does, I just don't get it.
My 2cents with few hands under my belt is that on the flop your only sizing should be pot. Protection is so important.
Originally posted by Kyyberi
I didn't get the part "suited starting hands are worse than offsuite" thing.
An offsuit hand can hit 2 flushes 
It becomes suited on the flop 48% of the time (out of the 50 remaining cards of the deck, 24 are of the same suit as 1 of the preflop hole cards).
I still consider suited hands more playable because they still flop a 4-card flushdraw about twice more often than offsuit ones.
In terms of raw preflop equity, though, the difference between suited and offsuit hands is very small, though still in favour of suited ones. The explanation is that, for an offsuit preflop hand to become 4-suit on the turn, 1 out of the 26 cards of the 2 missing suits has to be dealt OTF (with 50 cards still in the deck), and then 1 of the 13 cards of the single missing suit has to be dealt OTT, so the probability of an offsuit hand becoming 4-suit is (26/50) * (13/49) ~13.8% (the flop cards are ignored in this calculation, as if they weren't dealt). So an offsuit preflop hand becomes suited by the turn over 84% of the time anyway.
The reason why an offsuit hand beat a suited one in a simulation may have been that the ProPokerTools site (that merely runs 600K Monte Carlo trials) was used instead of the more precise Odds Oracle where the simulation length can be adjusted. A simulation with 600K trials can easily give an error of a few per mille.
I set the length to 10M trials (that's still fast enough but precise within 0.01%) and got the following preflop equities vs a random hand:
AKs ~57.02% (56196788 wins, 1640420 ties)
AKo ~56.50% (55664902 wins, 1667491 ties)
For hands without an ace, the difference is even smaller because the flushes that they hit sometimes lose to higher flushes.
Too many numbers for me. I just can't see how ATo is better than ATs. Core preflop equity is good, IF you can get to the showdown. But to get there you need often to hit the board. ATs can hit nutflush draw on the flop roughly 11% of times. I have no idea how often ATo picks a flush draw on the flop, but it becomes suited around 50% of times and then hitting a flop with two of that suit is 11%. So it might be around 5,5%? Hitting the nutflush draw is half of that. And as it's PLO for turn&river, you really want to have the nutflush instead of a smaller one.
What I saw yesterday was that most players are probably from Holdem. They draw to small flushes and then gave max value when they hit it, not folding pocket QQ+ postflop to any boards etc. I am quite confident that on the turn&river PLO players have HUGE edge. Tricky part is to get to the turn with good strategy. And one key is to have some kind of an idea about how often backdoors and backdoors of backdoors come.
Like when you have Kxxx on J73r board, and see a free turn. Turn is K and your 4th hole card is K. From air to nuts. Chance for that to happen is small (less than 1%), but if you have multiple miracle outcomes and few good ones, the overall chance to get one of them might be a difference maker.
I see that everytime in PLO-CG, even on PLO20. Many players come from Holdem (ok, me too) but do not adapt to PLO. They overplay AKxx (xx= any crap), they 3-bet or 4-bet QQxx (xx = any crap like 72o), often they totally overplay a hand with only two good cards like KTsxx or AKsxx, they overplay bad kings, bad queens and especially bad acies on any board. My edge as a PLO-beginner is to make the nut-str8, the nut-flush or the nut-full and then milk those who overplay any flush, any str8 or any FH or even any two pair. I watch out for those who even think one-pair-hand is good got get it in. So at the moment I wait for good PLO-hands like AAKKds, AAKKss, AAQQds/ss, AAJTds/ss, ATKQds/ss, .... suited aces with rundowns, suited aces with pairs and rundown, double suited broadway-rundowns, double suited rundowns (in late pos), double suited double pairs, single suited double pairs ...etc. If the flop hits hard and gives me a set with a draw, a str8 with a draw to the nut flush, top set, the overfull and such stuff, I get it in. There´s often someone who pays (and sometimes gets lucky but that happens).
Originally posted by Kyyberi
And as it's PLO for turn&river, you really want to have the nutflush instead of a smaller one.
Vs the loose population of 0.02/0.05, likely yes. But as you move higher and start encountering more TAGs, a K-high FD has a decent equity vs their ranges, as they fold the worst Ax hands pre. However, I agree that a T-high FD is too weak at large SPRs (over 5?). A way of mitigating that weakness is 3-betting wider (and for much less than pot) pre IP 
(again, it can only work at higher stakes where they respect your raises).
That said, I'd rather 3b 98s than ATo on BTN vs UTG because I'm happier to call a 4b with a non-dominated hand.
In PLO-MTTs Villains are even worse. Many play Holdem (and even that very bad) and just try PLO for fun without having any idea about it. I either get a huge pot with the nuts or get outdrawn by some lucky random stuff like gutshot or some random-ignorant-str8 or some runner-runner-whatever.....
Hmm. I had an idea. As the starting two cards are first two of a PLO hand (used in turn and river), could we use PPT rankings to determine starting hands?
Like J9s. In PPT Oracle range explorer we can get PPT ranking for a hand, and if we put just two cards in it the spectrum is wide. But then we could use the average rank to determine the average value of the hand?
Spoiler
As far as I know, the solvers give similar RFI frequencies for holdem and plo. So if we use my standard micro plo ranges based on those, we might have decent preflop strategy? 15,15-20,30,40 from UTG to BTN.
With that and solid turn/river strategy (just as in PLO) should give decent edge. The only dark area is from flop to turn. How wide we should peel? I had a nice hand with pretty good turn cards today:
Spoiler
Looks like PT4 doesnt support Fusion in any way.
Install the latest HM2 beta, it imports Fusion hands as PLHE. (Afaik, earlier versions import them as PLO but don't display a HUD at the tables.)
The average of the PPT hand rankings is one of the good ways to rank Fusion starting hands, yes.
Thanks, got HM2 working. It does import the hands, but it doesn't show the 3rd and 4th cards. So there won't be an EV line.
I have played a LOT of hands at smallest stakes online PLO. But even in those, people don't stack off this wide. In both of these hands, the opponent had at least pot sized stack behind on the turn when the money went in. I raised or shoved, and opponent called.
Spoiler
What about 3bet range? AA/KK only?
Originally posted by Kyyberi
So there won't be an EV line.
I read raw HH files and type hands manually into Odds Oracle 
What about 3bet range? AA/KK only?
OOP, yes, there's not much value in 3betting anything else, the equities are too close pre. IP, it depends on the opponent; I make it wide vs a TAG, as I said before. Generally, there are different schools of thought on Fusion, though.
3bet game is tricky. Don't think we have much FE pre, so 3bet is only for value. In PLO when we know all 4 cards it easy to pick up hands that work in that situation. But in Fusion, we have only half of a hand. Have to think that a lot.
Originally posted by Kyyberi
Don't think we have much FE pre, so 3bet is only for value.
Don't we have FE OTF and the ability to take 2 free cards if checked to?
Yes. As in PLO. But if the main plan with a hand is to 3bet, get called and then bluff the flop, I don't think it will be much +EV in the long run. FE on the flop is smaller than in Holdem, and I would think that at least at smaller stakes it's even smaller than in PLO. People love to chase, and in this game you can chase a lot. Even with total air you can chase, and catch it every now and then. I am really interested to see what kind of stats the opponents will have when I get more hands.
Using PPT rankings doesn't work. If I put all hands that have avg ranking around 15 or so (or better) our UTG range is QQ+,AK. Talked with Mackenzie today in his stream, and those preflop charts at OP-poker are a bit old and not based on any deeper strategy. I suppose Fusion is at the same stage as PLO some years (10 or so) ago. You can play it in many ways and make profit. Reason is that no one knows the "optimal" way, so no one can heavily exploit anyone's tendencies
Equities run really close, as in PLO. But in PLO you know your hole cards so you can select hands that dominate your opponent's and/or has good playability. In Fusion, Holdem monsters (KK+) are big hands in Fusion. AK is 55% against random hand, but creates flop situations where we can shove with small equity edge. Rest are in a big bunch of hands that no one knows which ones are good or not. Like which one is better, 75 or K4 in fusion? Or are they the same?
As preflop equities and possibilities are even closer than in PLO, I think you can limp with almost anything from the BTN. Position is really important as general population (at lowest limit) doesn't raise flops without really nutty hits. And even then they go for c-raise. It means that position gives you a lot of options. You can bluff certain boards (decent PLO players should have a good idea on what kind of boards we have enough FE), or get a free peel for the turn to suckout with your flopped air.
In one way it's like PLO long time ago. No one understands the value of draws or medium strength hands, so their bet or raise range is limited close to the nuts. But their calling range is wide, as they play it like Holdem. At the moment I think you need to protect your flopped hits a lot, which means cbetting and donking when you hit the board. Free turns are even worse in Fusion than in PLO. Don't give them unless you want to get one.
Here you can see Holdem players. In PLO when I re-raise on the river, most players would put me on nut straight. I did pot flop and turn, and on the river I bet half a pot and SB minraised. BB overcalled and I shoved. SB calls instantly and BB went for a tank. And then in the end called. After I re-raise and SB calls, BB should realize that his 3rd nuts can't be good. I don't know about Holdem ranges there, but in PLO BB should fold for sure.
Spoiler
