Hi Phil, can you do a video on best tools for cash game full ring no limit. Want to become best player possible but I would like to use the best tools in 2023 so my live and online cash game is best possible. What tools do you recommend?
Poker is killed with on-line Poker. This is the bigest scam in the Poker history. No on-line poker- no bots, no AI, no acounts cheating etc. Just live game.
The problem is like steroids in sports... once used, the advantage has already been gained. Yes, they may be caught and banned... However, they have already monetarily gained something that was not theirs. That money or opportunity will never be refunded to the supposed original winner.
When top tier players like Galfond are spending their precious time making informative poker videos rather than grinding for millions, you know the game is tough
@nintendokings But I think even that pool is starting to dry up. I mean Jonathan Little is basically a content whore at this point. All the videos he used to charge thousands for he's just uploading for free begging for a click
Nice video Phil. But I have to comment on your "you can still win" argument. I remember when I was playing back in the days when the bots started running rampart on lower stakes. For example 1/2$ I could still beat them and they were about break even (still massive winners from rakebacks, around 60% back then). But if they were breaking even that means they were sucking a lot of money from the ecosystem from losing players. So the average skill level to break even needed to be way higher. I was playing on Ongame (network) mostly and the bots played for years before getting banned. Sickening.
It technically does get harder each year but you're likey just a losing player, it is still very very possible to make a living online. I transitioned from live to online full time at the start of covid and I'm amazed at how soft it is
Just keep in mind, it would go against Phil's incentives to say online poker is dead, even if that might be what he really thinks. This is his income and niche, ofc he will say poker is booming.
my biggest fear is the combination of AI and an RTA tool being able to utilize your tendencies over time and adapt, almost like a nodelocking bot. Probably a bit far out at this point, but definitely possible
The problem is that regulations most times take more money away from the games than bots/cheaters themselves. This is very apparent in pokerstars games liquidity. The more regulations you have on deposits/withdrawals the less games are running. Public traded company, more regulations, less money on mid and highstakes. Seems pretty clear.
The thing I’m worried about is players using RTA on separate computer therefore would be undetectable to the poker sites. When AI solvers get sophisticated and quick enough to provide answers in real team you could just input info into a separate compute as you play and get instant answers. I don’t see how this could ever get detected, what do you think Phil?
In chess, where both engines and detection systems are more developed, this kind of cheater is still detected because their performance is too suspiciously close to perfect, especially in spots where humans usually get it wrong. I think it's possible that detection systems in poker will reach a similar level of sophistication. But if you just use the RTA every once in a while, then it probably will never be detectable.
The RTAs are pretty expensive. I doubt the average player puts up accounts, multiple computers, all the hassle to beat the low-mid stakes. Some sure. 90%+ remain human. Or go play the casinos. :)
Lol @phil thinking solvers isn’t ruining poker sit at any 1/2 6max online game these days and it’s 6 Belarus cheating nits when 1/2 in 2010 used to be almost all recs and beatable for the common man. Sad that ur so delusional man
6 max NL is dying. Blind and ante games 8 max are gaining popularity as they are harder to solve as many more pots are multiway where nash equalibrium is impossible to achieve
Hey Phil, long time RIO sub and fan. You were my first coach through training sites if you will and for whatever reason the way you explained things has always clicked and improved my game more than anyone else and I will be forever thankful. I will say that I always preferred the problem-solving and read based plays you made and reasons why instead of "this is just a pot, it might not work how I want it to, but it's my strategy that's unexploitable so I'm just going to do this." I find especially at lower stakes games although playing like a solver is unexploitable and unbeatable, you are passing on bigger edges because you aren't directly exploiting their weaknesses and are playing against someone who isn't really playing close to optimal, so lots of plays a solver executes will just be bleeding money. I suppose solvers are getting better and we can just node-lock our opponents ranges and find their direct weaknesses and still get answers, but of course that comes with it's own set of difficulties and many web based solvers like Vision for example you would be unable to do this. Curious if I'm thinking about this the right way or not and what your opinion is, thanks! Edit: Want to say that I'm not saying you copy a Solver output and never deviate. I like and appreciate the spots where you're facing a river bet and say "this may be a theory call, but human opponents aren't finding the bluffs here so I can just fold." Much love
Appreciate this comment very much, thank you! I agree that the game has changed a lot, but I still feel there’s a lot of room for exploitative play, even against some of the best. The level of fundamental knowledge you need to do so, though, has grown.
Online is so easy.. playing with all the bots makes playing your average reg that much easier. Then playing live is just ridiculously easy but boring as hell when you’re used to playing 20+ tables. Nothing beats online except maybe deep runs in live tournaments. But yes people should definitely work a real job.. unless you’ve been crushing the game for years online don’t try to become a pro now. Try the occasional weekend warrior route instead.. maybe you can bink a tournament or have a nice run in cash. If you want to learn tournament style poker quickly though I’d suggest SNGs.. better bang for your buck and can quickly teach nearly every level of tournament experience.
What job? Poker isn't the only thing getting destroyed by AI, real jobs will be finished too. Imagine studying CS for years and paying lots of money for it, and then you realize that chatgpt solves interview questions much better than you. And it will only get better, much better, better than humans can even imagine in their wildest fantasies. And then what job will be left for us? Yeah, playing poker, chess, league of legends - these are the harmless and useless activities humans will still be able to do. AI won't need our 'help' in real work where it matters.
@@Dzaandry If you think fields like computer science are anywhere close to being overtaken by AI, you don't know enough about AI. There is still so much that we as a species need to discover and do before we reach that point and it's not going to happen all at once
Lots of sites claim that perfect gto bots can’t be beaten and this isn’t true - they are exploitable and I proved this by beating pokersnowies bot after it had already beaten a few of the best known heads up players. You can’t beat it by trying to match its style you have to play incorrectly in effect which it doesn’t understand. Play any of the bots over a large sample size and you win by simply muscling every pot- a gto bot struggles against 2x pot bets or raises on every street on every hand where a human will quickly realise you are a lunatic and adjust accordingly a bot still tries to play optimally
Its good that youre talking about how secure the poker apps are against bots, but also giving away so much information like "they take screenshot" is also helping bot/rta users to circumvent detection.
Online Poker is just beyond rigged anymore. Tournament Big money hand in the money and I have someone call my preflop AA all-in with Queen Five offsuit and turns/rivers the Q and the 5.. c'mon man in no way in no world in the history of forever would anyone do that... ever. Rigged completely 100% and anyone that replies other is simply in bed with one of the sites (or agrees with this all-in move, and is one of the worst poker players in the entire history of the world. haha. There's always going to be way more players giving these sites way more money than I will and they will always give them these hands to keep them playing vs. me. Period.
The answer is YES! AI is ruining online poker. They have an event where pros play against AI bots and guess who won. The AI bots smoked all of the pros.
1. One of the main mechanisms bots cheat is via sharing information between them. Bot collusion shifts the game window such that what might be optimal assuming no-collusion changes. 2. I also don't think (non-cheating) bots are at 95%, playing a perfectly optimized GTO strategy is just sort of an entry-point, the ideal agent would do things like detect sub optimizations , classify opponents, and build an "exploitative" decision tree automatically based on hand-histories/prior knowledge to maximize profit in any game.
Nice video - I have a couple of points in regards to bots - you mentioned about timings being the same etc, could these people developing these bots not just randomize everything making it so that its even harder to detect? Also in regards to sites taking screenshots couldn't RTA users just be running the RTA on a different computer with a mirror screen making it impossible to detect?
Great video. How do you think country-wide regulated online poker in the US would affect live poker? I've been having this conversation at the table a lot and am curious to hear your take
Solver is bullshit. If you study solver outputs for 1/4, 1/3, 2/3, pot, 1.5 pot and 2x pot, and I develop a strategy for 10% pot, 3x pot, 5x pot and 10x pot, then you will simply not be prepared in any way for my strategy.
Hey Galfond. Nicely video indeed. As a coder/programmer Myself, I can tell you from personal experience It takes a lot of work and coming up with " useful " ideas when we are talking about creating A.I bot with implemented RTA combined with game theory. You see now days is much " easier " to develop anything from scratch or from " used ideas " and go from there, as technology and information are way much quicker to obtain than it was 10 years ago. Regarding Bots or in general " cheaters " in Poker, they have been exciting from day 1 and nothing will change that fact. But you also forget to tell people that all these big poker names on the market, did also cheat regular players, and giving a certain % to companies who invested and used " GODMODE " not to mention DDOSING players on WSOP or Major High roller Tournaments, etc. You are also right, those who use certain " bots " have a ring between them, and is not a secret, to be honest, especially same those rings are highly targeting " apps ". Hopefully to see more videos like this :>
I quit online poker 3 days ago for good. It’s become so blatantly obvious that bots are in every game I’ve played in. Also we’re ultimately not playing the player but the algorithm. It’s come to a point where I can pretty much know when the all in opponent will river me. It’s become laughable. These online sites, gg, pokerbros have obvious algorithms that are rigged for action. The bigger the pot the bigger the rake. I’m done. I’m going to stick to live cash only.
Has Poker looked at what chess does? Chess has had this problem for a very long time and is still going strong. The key thing is that the websites track if you're using other windows and for a general player using other windows means you're getting distracted and your play will generally be worse. But for a cheater using RTA their play is going to improve when they're using other windows. There's a whole bunch of other algorithms too. But the main takeaway is that you don't catch people when they do something smart, you catch them when they never do something stupid.
Point of clarification: My understanding of solvers is that they are typically implemented using _brute force,_ analyzing a combinatorial stream of factors and playing through the corresponding poker hands. Technically speaking, this is not AI (more akin to "regurgitating memories"). AI -- and especially neural network-based AI, now the state of the art thanks to special-purpose hardware -- is completely different, using _inference_ corresponding to a combination of training samples plus a model which weights different aspects found in the samples (i.e. able to respond to _new situations_ using patterns recognized from previous). An AI which was training itself in real time could _theoretically_ adapt itself to exploitative deviation from GTO play by its opponents (scary, albeit difficult to amass a sufficient amount of training data in-game for said adjustment).
i have just subscribed to your channel, the content is soo good..and i watch a lot of other poker channels…btw you have a very likable personality, too bad the high stakes (the video when you speak about having been invited) back then didnt work for you
Why Jake and Ali would cheat more, being already winning players shows me how terrible humans they really are! Real colostomy bags of humans right there!
Great vídeo Phill, i was one of the players who caught bots at wpn, when i played there, there was like 30 accounts with the same stats, i was playing 24 tables at the same time there back then, when they got banned after my "exposed" i got 500$ back, wich was nothing compared to How much i Lost, a good thing to know would be what are the GTO stats, like you Said on tour vídeo, there are stats that are "hard to get", i dont know If these guys where GTO, but the stats hard to get that they had was 66% cbet Flop 62 turn and 60 River, without a pre calculated range, this is really hard, mostly you Will end UP over bluffing .
the lesson here is clear. playing online poker is worse much much worse than playing cards in the jungle. Rife with cheaters and the main difference is theyre all invisible.
There is a youtube channel called RTA poker that has a video showing him or her using their RTA on GGpoker, i guess as an ad for the software. Pretty blatant lol
can you do a hand history review video of some of the numerous PLO high stakes hh in the february and march 2+2 high stakes thread involving berri- amsogood-sanita-linus? Interested on your take on the top level HU PLO these days
I lost so many showdowns on one online site with premium hands that i started to be afraid of AA or KK cause i knew if i go all in i will loose.... was playing just micro and i m not a good player but have notice that almost always same players won showdowns no mather the cards and almost alway they where the big stack....... when u loos i dont know like 95% of showdowns there is something fishy going on....
The four Phils(now 3) have always been some of my favorite players, and this is some of the best poker content by one of the best players in the world.
Bots will eventually be able to take other players actions as input and begin playing exploitatively. At that point it will become very difficult to detect.
If anyone also follows chess, it's the same thing and should make people feel pretty optimistic. Chess is less complex than poker in terms of decision trees, and even there the bots are so deviant from human play they are easily detected on websites. Detecting bots as Phil said likely is actually quite easy, so it's not something to be super worried about. As long as the sites have integrity, that's probably the only real concern.
The issue with chess is that a good player, say a 2700 level, would only need assistance twice in a game to be virtually unbeatable. However, the mindset required to become a 2700 player is unlikely to be drawn to cheating.
i dont think its comparable in terms of decision trees. in poker, every hand is essentially a new "game", if we are talking cash game especially. in chess, every move changes the complexion of the game, pieces affect each other in different manner. i would argue its way more complex, but what do i know
no, this is completely wrong and an insane level of cope. there is nothing at stake in online chess. there is real money at stake in online poker. honestly, this video is some pretty insane cope too, although i appreciate the level-headed analysis (although keep in mind that the video creator is incentivized to downplay the dire situation for online poker). but anybody who is still playing online poker for money is totally nuts. think about it: if you were playing chess online for real money, you'd be completely insane. chess engines are strictly better than humans. yeah, cheaters usually eventually get caught, but that's after you already lost your money to them. same thing in poker. poker engines are strictly better than humans. if you are playing poker online, you are losing money to cheaters who are playing 100% optimally, period the future of poker is similar to the future of chess: IRL events where you can't cheat, and fast time controls online which make it much harder to cheat due to time constraints. you'll see these changes happen over time naturally as more and more people understand the situation with AI
@@xNOTMYREALNAMEx prestigious classical chess events are being held digitally as we speak.. like literally today for example. and the monetary prices are not negligible
the fact that 90% of people are loosing in online poker is the reason not to play online poker. You have better chance win a lottery for billion, than to be winner online
@@losyart not if you’re getting dealt pocket aces 5 times in 30 minutes and lose all 5 times to hands like 2 5 offsuit to runner runner straight. It’s become quite laughable for people to say online poker isn’t rigged
Thanks Phil, legend! Two quick thoughts 1) RTA seems more problematic to me as I see it as less detectable (players can use it selectively, only versus certain players, or in certain spots e.g. over a certain size) 2) Wouldn't want you to give out detection methods to bot users who don't already know about them (I assume they already do). Thanks for the video.
The Mersenne Twister is a dead algorithm and should not be used anymore. A.I. Bots have reverse engineered it for the last 10 years almost. GG Poker is horrible for bots right now. They are targetting too which makes me think it's actually part of GG Poker's tools to gleam profits. Nobody audits their players except when collusion is discovered, and although they have AI detection it's obviously too prevalent to not be in use by the company themselves. The RNG audit is the only thing they get tested for outside of financial ins and outs. We need a live audit system in place to prevent the company using illegal methods.
While at a meeting for an MLM I met this guy who called himself an entrepreneur. We didn't really talk much but he once invited me over to play online poker. I had no gambling knowledge then like I do now. He showed he was playing multiple screens at the same time. He looked at me and said nothing, explained nothing. I got bored and left almost fell asleep actually. Looking back that day I realize he was trying to put me on how to take the fish with technology but I was too oblivious. I'm kinda glad i never got into it then. I may start a little poker now that I got some know how but not like him, fk that.
The biggest Canadain player on GG uses a solver and sits infront of windows, you can make out the solver in the reflection. Ill slap him when I see him at the casino again but then again he can't play without one so is a "mostly online" player.
Poker is doomed, bots will improve dramatically to a point that they could be tuned to play like anybody and on any style. It will be impossible to distinguish play in that case
OmgClayAiken, you are awesome! I used to play in decent high stakes game (fixed limit though) but obviously nothing along your calibre but it always made strange sense to go GTO as you would normally lose against rake even if that becomes quite negible below 10/20 then that still means 99.999% of all players. It would seem best to go highly exploative, non GTO as long as you can figure out your opponents. Do you disagree?
i suspect my winrate (ssnl,1.5mio hands@ >12bb)is barely ~30% made of playing solid. the major part of my edge came from timing tells (~20) and most important (50%) is putting people on tilt. which for the most were bad agressive plays, but also "sympathy" calls and stuff. upon today that feels shady. that said id really be "interested" in poker bots that focus on those exploitative strategies against human players.
Im still confused on where the profit comes from in playing perfect GTO against someone not playing perfectly themselves, like in your example if someone is calling too many hands on the flop, wouldn't that make our value bets more money but our bluffs lose more money, and this would net 0? I assume this isn't correct but I dont understand how or why
Hands which are not mixed frequency plays at equilibrium but pure instead (you take X action with X hand 100% of the time) will cost you EV if played differently.
You should earn your profits from recreationals (pure value bet) and unbalanced bad regs (exploit poker ) . Vs balanced regs Gto strategy is an even strategy maybe slightly minus because of the rake.
You are wrong at 1:50 Any 1/3rd "strategy" on RPS example is an average over time but humans including you cannot avoid statistical bad patterns. 'AI' software will find your patterns and very slowly will beat you. Same as the bookmakers. We are talking fractions of any % advantage over time.
Would it help if poker sites randomized names of players? Yes there are plenty of cons to this. But what happens when you play against AI or Bots over an extended period of time that keeps track of how you play? Bots could create a customized strategy against certain players once they have a big enough sample size? Would almost be nice if there was customized hardware (like a Xbox console) that u could buy that could not be altered - something along those lines. I want to battle humans one vs one. Just me vs you. Study the session and come back for round two.
Back in my day, we were never afraid to play SkyNet HU4Rollz. #OldManPhilG. The irony of you and the other "old school" internet poker players being seen as the "older" guys on the block is humorous. Especially since I found the game in 03 in my mid 20's. Keep on making top-level thinking understandable, it's really a gift.
bs pokerstars is trash. I'm done there for good. Totally rigged. If by playing good you mean sucking out constantly with 6 or less cards to hit then yeah you definitely need to do that. not only that but the same 5ish or so people keep on dominating every time they play. hmmmm no not coincidence their betting and running skill is just far superior lmfao BULLSH*T.
I dont know man i dont really know what to say about online poker, i just ran online in 2 big tournaments ( bought in 30 to 50 dollars each. i get to the money i am very deep stacked, someone shoves i call with my higher pocket pair.. kk vs jj .. he hits a j on the flop , and my other table i got 10 10 vs 2 2 and the 2 hits the flop, making me go out only to take a bit of money .. but always on key situation i get all the chips in, like 85 % and higher to win on flop, then turn river comes and voila he hits 1 of those 6 outs or less. I dont wanna say the rng is rigged, but it's quite bs when u play 20 years of online poker u notice stuff...
Online Poker surely is dead as of this moment. It will probably devolve into a zero-sum game. Nobody Wins, Nobody Looses, except the house who takes stack.
I played on party poker for years. My last cash out they just kept the money. Zeroed my account then stopped responding to mail. Even though it was almost nothing a few hundred at most. If I was a bad man I could easily cheat any poker site and never get caught BTW. In my mind online poker is over unless they start booths you have to go sit in at the casino XD
I've been playing on global poker since it's inception. And I've noticed over the last 2 yrs how the site has changed. If you're ever played pac man and figured out the routes that's how it feels with global.
Anything competitive online is done. Tech is out of human hands. Its was a nice honeymoon but all this is dead. Cant even believe anything on here anymore. This could be all fake. No way to know.
they are ruining online poker. if solvers exists cheating is more likely which creates fear in the ecosystem. Ive been playing online poker for 20 years and since solvers came around i stopped playing. Id love to play live poker thought if there was a game in my circle of friends
what if you just stream the game to another monitor where an AI program will be evaluating the game input and make recommendations that you the human will execute . This way it wont be detected.
its like in other areas for instance hackers vs admins neverending battle between two polarities security will improve significantly as well as the capability of cheating, so I think we should now worry about it
It is very sad that bots and other automatic and RTA tools are destroying online poker. Many people don't want to try it again because many places are full of it.
Artificial intelligence has no emotions, The math in poker is not 100%. It is very important but not the only skill needed to be top 1. The answer to the question in the title is negative
It seems that the optimal strategy for poker AI on poker sites is to be indistinguishable from strong human players. All that is needed is a dataset to be trained on; the "game" for AI is to win and not be caught. It will be interesting to see how far this technological arms race will go, this is just the beginning.
Exactly. I said the same thing with a lot more words above then read your comment and learned a little lesson in brevity. BTW the poker sites will just adapt by throwing off all winning players. That's what sports book do. They don't care how you are winning and they have that right. If a small % of players are winning a lot of the money and you can't distinguish the machines from the legit players then it just makes business sense to toss them all off and cater to the bad players who will have more fun playing against other bad players. That's what I would do :)
So I had a few years away from the game, years ago it was really common place to see videos where people were using apps that tracked players stats. I was waiting for you to mention those in this video, so I'm still not certain coming back in to this space, are stat trackers considered RTA?
what a blast from the past! Great to see Phil's still around the game, I spent a lot of time watching Phil's duecescracked vids and big full tilt games.
GTO only works against people/machines who have a clue about poker and are attempting to execute a coherent strategy. If humans are making "suboptimal" (not GTO) plays such as being massively LAG/NIT, but they are schizophrenic in their execution, only a human can pick up on this, mostly because we've made these mistakes before and can use our empathy. For example the AI GTO bot is holding AQoff at a 9-player table in the BB and is facing a raise from Player X with slightly less chips and a shove from Player Z with slightly more chips. The GTO bot is in that difficult to solve zone of 20-30BBs before the bubble of the FT or on the FT, how is the GTO bot supposed to react to the shove from Player Z, who is schizophrenic? The table as a whole might only be shoving with AKoff or better in that situation, depending on the blind structure and pay-outs it may be looser. Player Z is schizo and is actually shoving with AJoff. We as humans will have a better instinct about what to do in that situation than the GTO bot. If the GTO bot assumes the other players are playing GTO and has observed them making several strong GTO plays in the recent past, then it should fold the AQoff to a raise and a shove, right? But if we as humans know Player Z is LAG and Player X is nit, then we might reasonably assume the nit is folding and the LAG is shoving with a worse hand. Of course there's a risk the nit is only raising with hands they intend to play but if we are talking suboptimal nit, not TAG, if the nit is holding the best of it (say AKoff), there's a higher than usual chance they fold on the bubble or at the FT So calling with AQoff is not the GTO play and might make us look and feel pretty stupid against some players/fields but we'll be feeling pretty happy about life if we call and the schizo shows AJoff IMHO the GTO bots and AI enjoy their biggest advantage in deep stack cash and in the early stages of MTTs, as well as in short-handed turbo games like 6-max hypers. If you don't make any mistakes in those scenarios/games then you're guaranteed to be profitable. To actually win though against a strong field takes more than just GTO. The optimal play does not depend solely on standard maths but is tailored to an individual player. The AI bot won't be able to feel another player tilting or be able to read their minds as they have no concept of emotions, motivations and goals. IMHO the optimal strategy is mostly standard GTO with a liberal sprinkling of instinctual plays, all the best players exhibit instincts and not just computation. There's an even bigger problem than AI GTO bots and that is organised groups sharing cards and info. You're talking about regulated sites and I used to play a lot on PokerStars. I lost count of the number of times I got emails saying I was playing against people they proved were colluding or against a botnet they took down. Sometimes these were weeks or months behind. The colluders, even if they are only average players, have such a big edge they will smash any AI or the best players, especially in non-holdem games. The big thing is the change in demographics. I'm just talking about PokerStars, which used to be North America and Western Europe. Now it is majority eastern euro and South America. People in these regions are fully satisfied with making $20K/year and are more than happy to collude in order to make that happen. This is organised and sophisticated and they have enough players at enough sites on enough tables to entrench their edge. There's also the spectre of an AI GTO botnet with multiple entrants in the same tourney/game. That would be unbeatable and it almost certainly exists. You've probably had this feeling before - you tagged some players and have determined with a high degree of confidence that they are not a top 1% player, yet they have top 0.1% stats. You wonder how did they pull that off? Just lucky? Variance? Then you notice they are always playing at tables with players from outside western Europe/Canada. These players also suck but mysteriously have killer stats. Some such players I've described have been banned but so many slip through the cracks. The sites themselves like Stars can do nothing to stop proper sophisticated and determined people who have VMs or separate machines for cheating and playing. It's trivial for most people to use encrypted comms to pass hole card and other info These kind of players have killed the low-mid stakes scene and traffic has fallen off a cliff. Players are suspicious and just not having fun. Some people don't mind consistently losing as long as the game is packed with action and plenty of fun but playing grinders and colluders? It's not fun at all for most people, certainly not over the long-term ** I sound bitter but that's just because I hate PS and online sites, not because I lost money there. I actually won some sharkscope leaderboards for profit and best 500 streaks, etc, had an ROI in MTTs of about 80% and in SnGs about 15% which includes 6-max and 18-player games. I'll make the claim I was one of the best in low-mid stakes. I no longer play cards and am happier without it. If you are capable of beating poker there are many more useful things you can do with your life than play cards.
Well i am playing since 2008 and made 100k + in 10 years by playing just on free days when not working only used an tracker, played long time tourneys then after birth off kids played mostly spins. made good profit in spins, but suddenly in 2017-2018 not possible to win anymore especially on the spins off 100 dlrs on stars. Off course i was playing players from cartels who used collusion and solvers, Had my account banned for more then a year cause i appointed this to stars thay they should investigate and give me back the money if there is proof. Never received anything back and some (not all) are still active in high spins. I always loved the game, but i know now when i play a certain limit off buy ins the game isn't fair aymore. So i quitted online poker
I believe that GG Poker is full of Bots because this happened to me twice. I was playing Tournaments then suddenly we had some technical problems. They paused all tournaments, at this point there was a traffic over 200 k players. After a while, i think 30 to 40 minutes they cancelled all running tournaments and all tables getting closed. Within seconds the traffic fell down from over 280 k to 25 k. Running tournaments were cancelled but you could still play cash games or rush and cash or other ring games. No one can tell me that all those people left the game because the tournaments they were playing was cancelled. This happened to me twice, very suspicious or not ?
I think Phil’s argument, ultimately breaks down, and in fact, AI and solvers and other tools, will eventually destroy online poker. But not in the way one might expect. I think the sites will be forced to throw off ALL winning players and only allow bad players. Here’s why: 1. There are tools, that can teach people, how to play an unbeatable GTO strategy. 2. And adversary can use these tools at runtime (i.e., during a game) to play perfectly. 3. The only reliable way, for sites to catch an adversary, is by analysing the play of winning players, and seeing how far their play deviates from perfect GTO strategy. If it’s too close they are suspected or banned or shamed or punished in some way. With just these three truths (Which Phil agrees with), we have a BIG problem. The sites will be forced to punish all the players that are closest to playing optimally. So as time goes on, and some players get closer and closer to GTO, (and not by memorizing every scenario, but by heuristics and real time algorithms running in their heads), these very players will need to defend their play to integrity teams. Ah, you might say (as Phil did), even the best human players are nowhere near as good as someone using a solver in real time, so the sites will just use some sort of AI to account for this when doing their analysis. All you must do then, is find the best players, and use their hand histories to see how far they deviate from perfect play. Then you play at the same level or just a bit better. If they are playing GTO 68% of the time, then you can play 65% or 73% of the time and still play as good or better then the best player on the site. And there is no real way the site can distinguish between these two players unless the adversary makes some sort of other stupid mistake. But if you read your poker client from another machine or device and plug that into a solver on another machine that then spits out the optimal play, and then you decide when to and when not to obey it, there is no way they can distinguish you from Phil.
Powerful advice. I quit poker back in 2011 just when game theory and trackers etc where about to be take off. I watch it all the time now but never play, as I think I would just have no edge with all the Solver stuff doing the rounds these days. But listening to you about limitations in human beings input limitations (I would respectfully disagree with you in the cases of Chidwick & LLinusLLove though sorry) I think I am going to give it another go this winter. Thank you Phil, you really are one of lifes good people sir.
I'm a bit confused about how you can play a strategy that your opponent knows and he still can't exploit you. If, for example, you decide to hero call, say, 33% of the time on the river and I know that that is your fixed strategy. Couldn't I just make a large bet on the river and pick up the pot 66% of the time? (well slightly less then 66% of the time since occassionally you'll be able to justify the call but you know what I mean).