compilers are mostly CPU bound. secondarily I'd say they are memory bound, that is memory bandwidth, not memory size. that ht gives as much as 70% speedup isd proof that it is not 100% cpu bound. if it was 100%, I'd expect very little speedup. 70% speedup tells you that we have exactly the kind of workload ht and other mt was invented for. in old days without multithreading, a core would do only one operation per cycle, in between the pipeline, staging, bus requests etc. every time it missed a cachge lookup,ba page fault, we had to context switch, and wait until next cycle to try an operation. with ht,bwe have always multiple pipelines lined up, and when he is out because of page faulkt another is ready to go. basically, wghat my does, is to allow a core to be 100% utilized. that's why some workload see improvement with ht, while a few even see degradation. a workload that is truly 100% CPU bound might see degradation. workloads that are mosrptly cvpu and memory, will see greatest improvermernt. in the case above, they might benefit from even more than 8 threads, if thecore is not 100% utilized yet. it depends how much cpu bound vs memory bound the process is. now the core would be idle in those cases where more than 4 threads had a page fault in same clock cycle. now you understand why processor cache vsize is so important. ram access is extrermerly slow compared to CPU clock cycles.
dead tree books are buggy whips. if you haven't noticed, a lot of book stores have closed shop. borders are gone. and it is not just because of amazxon. when is the last time you looked up in a dictionary or a lexicon?
if they are state that only 1 in 25 don't listen to radio every week, theyre lying. my car had as radio, but thastsd because I've been to busy to replaced with a digital media player. I only listened to that radio maybe 5 weeks per year, and only npr, and never music. I know lots of others who also don't have as radio in the house. radio is as retro format like vinyl.
As much as we want? If you compare world food procdution to world oil production, and keep in mind that many already do not have enough food, and also keep in mind that modern agriculture is a large fuel consumer, you would have to eat those words. There would nothing else for dinner.
Thats what the grid is for. If the grid is powerful enough, you will export your excess to the grid when your storage is full and import from the grid when your reservoir is gone. If there are some hydroplants on the grid, excess solar power could be used to pump water back to the reservoi during daytime when the turbines are running at a tricke, and crank the turbines up to full load at night when all the solars are down. Excess energy could be used to produce hydrogen or methanol for fuell cells. Maybe a generator per household would be overkill, but one per neighborhood oughta do. Energy storage is not trivial, but it is a far cry from impossible. I can imagine an energy coop would want to have a,ix of souces, including solar, wind as well as a common energy storage facility.
If the purp0se of the website is to recruit suicide bombers and bloodthirsty killers to attack innocent 3rd parties, then you can consider it part of the infrastructure of a hostile army, and a first class target for anyone wanting to interrupt such monsters. If you go all meta and pedantic, you have already lost grasp on reality, and need a brain check.
Data warehousing? Realtime realistic light an texture 3d graphics rendering? Hosting dozens of virtual machines? This on i find most likely, especially if needing tet, dev, demo env etc, or if need to simulate large complex sw running across several servers. A mid range x86 desktop can easily outrun yesteryears refrigerator sized unix servers. Single host massive denial of server attacks? Or just absolutey useless waste of time an money. Or just a pack of lies. Personally i have set up servers with more than 1 tb ram, but they werea actually needed for a specific workload.
I can say the same about my i3, or even your average cell phone chip that has ony 2 cofes, eacch being only 1 5th as powerful as an i3. Disk io, network io, mory bandwidth are sill the most common bottlenecks. If a regular desktop user has bad response times due to cpu shortage, there is probably a software bug, or misconfiguration.
The dirty little secret that is shared by everyone who has a chttp://www.aftenposten.no/nyheter/uriks/Slik-smugles-folk-til-Europa-7809354.htmllue, is that the cpu hardly matters anymore, except for a few applictions. Your regular dick jane and joe that walks into bestbuy getting outfitted for college or needing a machine to do their acconting and surf the web shoould not even be looking at the cpu model. Theyd run just fine on an atom or celeron, or a pentium Ii from the 90.s.
Bah. Not interested in the quad core stuff. The only thing i want from this family is as powerful as possibe dual core chips. Dual core xeons are pretty much extinct by now , but there are a few servers sold with core i3 and i5 instead of xeons.
You see the licensing pricing struture for software such as oracle db, tuxedo, weblogic, and others are per core. So it has become tricky to source a mainstream server model and ke ep the layered product licensing under $100ks of dollars. Since the recent x86 cores are 10-20x as strong as some old unix cores , you can replace an old refrigeratior sized unix hog that cost a million dollars, with a single x86 blade module for a couple of thousand dolllars. But you still end up paying big $$ to oracle.
And dont tell me we need to roll our own hw. We have to pick from mainstream server models that can be sourced and supported in evey country.
Solving this challengeis is where the big money is right now. Being able to scale down hw for commercial software to take advatage of these fast cores. There are solutions, but none are optimal, and often come in conflict with the client company's irrational and rigid rigid it policies. Scaling up is laughably simple. Scaling down is a real head scratcher now that eveything is getting bigger.
card sharks don't make a good living by knowing when to walk away. if they did they would have to retire after a few days. they make a good living by exercising their skills in a skill based sport against other less skilled players. sometimes evern the sharks fall victim. sometimes their deep bankroll drunk millionaire victims get lucky. some times they run into other sharks. the best profesionnal gamblers avoids gambling. they do not take bets where they hold the short end of the stick. if they best on sports, they must be better than their bookies ot betting public at predicting outcomes. the opponent is not lady luck, lady luck only defines the outcome of each bet. the difference in skill between the player and other players defines the outcome and in the long run. if playing poker, they must consistently play at tables where they are there strongest player. in theory if you only ever accept profitable bets, you will win in the long run.
your way of thinking is why operating a casino is such a lucrative business. every real poker player only plays when he has an edge, which means he has an advantage over the other players. if you enter a game hoping to break even, you are just wasting time. any bet has a given chance of winning their pot. a poker player should know what the chance is for hitting the card he needs, and based on history and bidding, what cards the other players have, and what cards he wants the other players to believe he has.
there is no wnnng strategy that says play a few hands hope to get lucky.
they can't. they had to practice before entering liver tournasments. and for all that succeeded there were 2 that failed. a tell is not necessarily a grin, a tick or some hand movement. it is really hard to control your pulse. a real shark will be able to study the pulse that is visible in you throat. which is why many poker players dress a bit over the top. but the most visiblke tell is the play itself. how often you see the flop, hoiw often you bet, rasiser, fold. what hand values you bet with etc. the visual clues are just extras. who knows, maybe there are even ubersharsk that know how to manipulate their pulse and leave fake tells now and then, when the stakes get high enough. if you hold the nuts in a half million pot, you have to do all you can to make the others believe you hold less. maybe seem worried that one of them have the nuts
wrong. if you Always play the math, you opponents also always know your strength. they might fold medium hands early instead of wasting 10 big blinds calling you down to the river. in other situations they will know that you hold a big pair or high cards, so that your small straight is safe. a human could play close to even with the bot in all the vanilla hands, and for a certain number of situations, exploit the fact that the robot is a robot. so human wins. a losing strtategy would be to be clever you n every hand. just play perfect like the not, be patient and waitfor the situations where the math is what leads the bot into a trap.
you math is wrong, you don't lose 0.5% in the long run. you lose 0.5% per game in the long run. After a few hundred games the average player will have lost it all, not just half a percent of their starting capital. To NOT loose it all, one has to be extremely lucky, plus one has to quit, or keep sloppy records, forgetting about the many losses and remember the few wins.
agreed. the onmline casinos makes large sums on the rakes and feed. they win, regardless of who wens easch hand. if they were stealing from customers that way, they would be killing their golden goose. no need to cheat the schmuck out of $500. $500 will soon enough be won via rake, and then another $500 deposited,
I had a mysterious issue on my washing machine(20 year old GE top loader) , seemingly the dial did not work right, and often, the machine would get stuck on a part of the program, usually the spin cycle.
This was going on fro a couple of years. As a rule, I never would leave the house with the machine running, in case it would get stuck, and end up burning up the motor and start a fire.
A couple of months back I figured it all out. It is that stupid safety switch. The lid of the machine is just slightly bent , you can't even see it, but it is enough so that the safety switch is only barely connecting when the lid is closed.For now I just put something heavy on the lid, maybe I'll get around to fixing it permanently one day. I got the machine dirt cheap off craigslist. I bet the original owners would have kicked themselves if they knew what ailed their machine........
At the same time they can fix fewer gadgets, as well as more gadgets. And another reason why we often don't WANT to fix something is because it is too expensive to fix.
Point in case: In the old days, if your radio or TV stopped working, you would take it to a Radio or TV repair shop, and they would fix it for much less than it cost to buy a new one. People would go to TV and Radio repair vocational schools. Now Radios hardly exist, and if the TV breaks after the warranty time, it is cheaper to toss it out and buy a a new one rather than finding someone who can repair it.
I am mostly clueless about electronics, but when my old monitor broke, I was determined to fix it. Googled and found videos on youtube showing me how to go abnout it. Learned that there are things called capacitors in the power supply, and my particular model contained a batch of capacitors that were often failing. When they swell up, they are probably toast. Bought soldering iron and other useful tools. Ordered the spare capacitors, and managed to get the monitor working. The monitor was worth maybe $30 tops, and I probably spent 5 hours in total on the repair and $100 on tools. Could have bought one twice as big, lighter and faster, with modern connectors for a little over $100. Later I was however able to fix my neighbor's laser printer by following same method, looking for bulging capacitors to replace, and I tried a couple more repair project where I was not successful. I have earned the cost of the tools at least. However, I _should_ have just tossed the monitor, if I had acted in an optimal manner. If it had been 1990, I would not have learned how to do this, as there was no youtube to teach me. Such a monitor would have been worth $500 and it would have been worth it to pay a repairman $100 to fix it.
It is not just a generational ting. It is a wealth thing. In poor countries (compared to tech today, UK and a lot of Europe could be considered "poor" in 50's compared to now. Less than 1 car per household, no TV, maybe a radio, working class jobs, hand me down clothes were the norm. So it woudl make sense to repair things and learn how to repair things. Now youth spend small fortunes on brand name clothes. And for the cost of a 1950's TV or radio (in 1950's dollars), you can now buy a ton of advanced electronics, and most of the gadgets are not SUPPOSED to be repaired. If there is no repair shop for a gadget, even in a medium to large city, how are kids supposed to get the idea that they should repair it themselves. Broken items are just taken in return and replaced with new product, the bad ones shipped to china for refurbishing, and the resold. Also, a gadget made today is only supposed to last a few years. My family's black and white TV purchased in 1970, was our only TV until we upgraded to a color TV in mid-80's. I'd like to see someone still running their windows 98 PC today. Or their Nokia analog cell phones? Products today are mostly obsolete by the time they need to be repaired. You can't blame that on the kids.
In poor countries, where they can afford few gadgets, there are plenty of cell; phone repair shops. 20+ year old cars are repaired. Parts are machined. People repair their bikes, using only hammers, nails and pliers as tools. It cost under $1 to get bikes repaired by professional bike repairmen slightly more skilled in the use of hammer, nails and pliers, but in a poor country, $1 is not something you waste on things you can do yourself.
SO the professor can rest his worries. There are plenty of people that fix whatever they can fix. Some in rich countries tinker and repair for fun, but in poor countries they still do it because they can't afford not to. If you add a few more dimensions to the picture, such as economy, history and geography, nothing really has changed with the kids.
It can in fact win any hand, but it is proven to be the hand with the lowest expected value of all opening hands.
However, in a very tight game, you can win with this hand if the others are bluffing.
The pros play this kind of hand when the situation is right. Bluff to steal a pot, limp in a pot with nice pot odds, or just on a whim to mix things up.
And in heads up, since the opponent on average will hold only an average hand, you can use 7 2 for bluffing, getting the other to fold if he only has a medium hand (2/3rds chance of that). And if flp falls just right, it is now a semi bluff with chance of big payoff, because you will never be put on that hand.
Sometimes you don't care what hand you have, you are making bets just to find out what other guy holds, then make a bet that forces him to lay down exactly that kind of hand.
The difference is that teh robot will always fold this hand, humans will fold it ALMOST 100% of the time. Most of the times they don't fold it, they will lose extra money, and ocasionally they will make a lot of extra money. And by being seen to play that hand now and then, you can't be treated like a bot, which makes it harder for others to figure out what kind of hand you hold based on your betting pattern.
From what I have seen, the machine played against itself. It does try to find out what your hand is, but assume that you would be playing the same perfect game. It might not even need to find out what your hand you really had. Since it calculated all possible permutations, it knows that betting in a given situation is worth.5, folding is worth 0.33, raising is worth 0.60. So it will chose to raise. If you deceived the bot by betting on a weak hand that happened to hit the flop nicely, you can certainly use that to milk the robot for a few extra $$. Since he must follow it's heuristic perfect strategy, he is forced to fall hard into all the traps you lay.
The biggest problem with playing against this bot, is that it is not human, so you can't trick it into deviating from its perfect strategy. So the question is whether you can make enough bonus money on laying these traps to cover the cost of the money wasted by temporarily deviating from a perfect strategy. The extra money you make by knowing exactly what hand the computer is holding is however yours to keep.
You never know exactly what someone holds, but based on the bidding action you can put them on a range of hands. And further, once you see the flop and they bid again, you can narrow it down further, and to test things out, you can check, bet, call or raise yourself, and see how they react. Once you know exactly what rules the robot follows in any given situation, you know what hand kind of hand they hold. The difference between a human player, who does not claim to play perfect poker, is that you can be 100% sure the robot will follow its own rules, even when it is bluffing.
Example: when betting it will use a range of opening hands. Heads up that means above average hands. Those hands are pretty well known for both humans and robots, and there are also some hands in between where betting is more situational, In the case of a human, he will usually fold a hand like 10 3 offsuit. In the case of the robot he will always fold 10 3 offsuit, That little piece of information can be gold worth after the flop, turn and river.
Human to human, if a hand like 10 3 is the only one you are afraid of after seeing the river, then you can still hit the brakes and avoid 3-betting, because you know mr XYZ will and can sometimes play that kind of hand.
Human to Robot, the human knows the robot would have folded that kind of hand pre flop, so when we get to the river, and that hand is the only one we are afraid of, we can bet with confidence, as we know the the robot is not holding that hand. Maybe something harmless like AA or AK.
Every single hand of poker is a game of finding out as much possible about the opponents hand while revealing as little as possible about your own hand. For the most part, both will have to play it straight in order to not throw money away, and now and then the human can get an edge by taking advantage of such information, and at other times the human can take advantage by spreading misinformation.
And while the number of possible actions (bet call - raise check etc.) give a mindboggling numebr of permutations, a human will synthethize those to a much smaller problem set, Strong weak, medium opening hands, what hands to fold, check, bet with after the flop given different flops etc. A human spending some time playing this robot will remember how it acts in different situations. And if it is 100% predictable, the human can likely play only slightly worse by trying to follow same "perfect": rules, and try to cash in big in those spots where the human gets an informational edge. Since the human knows the computer's brain and not vice versa, the computer can't get away fro those situations, it will fall into the same trap every time.
The only problem, is that if your opponent KNOWS this is your strategy, he can adjust. He will let you bet your slightly above average hand, and call only when he has a much better than average hand. And it is not like you will get to steal every blind, only the ones where you chose to jam the pot. And that has to be more than 50% of the hands if you want to do more than averaqe number of pot steals.
So the best way to beat such a guy is to play natural with medium as well as big hands, and wait for the opportunity to call one of the jammed pots with a top hand. Then your strategy is no longer perfect. And since you decided to jam the pot with anything better than pair of 6 or whatever it is, you can't get away when the other player calls your pair of sixes with pair of kings.
The jam or fold strategy is however a very effective strategy for the novice, even in earlier phases of a tournament(earlyu on they can use the strategy only with the top hands such as AA,KK,QQ,AK) , because that eliminates any chance of being outfoxed by a better player. But again, if they know this is your strategy, they can take advantage of it, and will not call you except when they hold aces.
Yep. Or the angry drunk player, who bets every hand aggressively. I was once playing limit poker online at a table (not high just $1-2, or $2-4), and this one guy became aggressive after a bad beat. There were 3-4 other players. To make it big on that table was as easy as betting hard on every hand that was a bit better than average, because teh drunk woudl always pay you off. Shows to prove that sometimes one has to adapt th1e strategy to the situation, and play profitable poker is far away from perfect poker.
OK if that was some rich asshole blowing off some steam with his luch money, but since I quit poker, I always keep wondering if that was some college kid gambling away his tuition, or some dad gambling away his family's food budget.
It is called misrepresentation. You are trying to get the opponent to believe you are holding a certain hand. Ultimately you want the opponent to make the wrong decisions, either by underestimating (calling your honest bets) or overestimating (folding your bluffs). And within the action of the same hand you want to remain somewhat credible, so he will believe your story. But you want to keep it also as random as possible, so they can't be too sure what you have. You also are projecting a certain playing style, the careful, tight, the aggressive tight, the lose, the crazy etc. And during the same game, you might have to switch personalities. You can even have a reputation (deserved or not) that precedes you. If you are known as a crazy player, let them think you are, and play tight, while acting crazy. Online, players will even pick handles to project a certain style. You will take losses on purpose if necessary, to help build a perception, but you can't overdo it, as doing it too openly would be like bad acting.
Then you have semi bluffing. You bet hard on a hand that isn't much yet. The bluff works and you get the pot. And if it doesnt work, there is still the odd change that the flop gives you something good. And when semi bluffs turn out good hands you often get paid off really big, because the opponents have you pegged on completely different hand than what you have.
OK, it is true, that there are optimal plays. And heads up, it is possible that the computer might have an edge simply by not being human, because in heads up limit poker, you have to play and bid so aggressively to play optimally, that it is it counter to most humans risk acceptance levels. And the other problems with playing the computer, is that you can't teach him a lesson to get more respect. If a human is getting too aggressive, you can knock them back down by check raising a few times, and bleeding and wounded, they will back down, and respect you next time you swing the whip. If the opponent is getting too timid in limit poker, consider yourself lucky, and steal blinds fro as long as it lasts. A computer will forget about those incidents, and won't be impacted by those mind games. But if it doesn't get involved, and insist on playing like a robot, it will lose some extra big hands because the human can be pretty sure what hands he has based on its optimal betting patterns.
Slashdot Mirror
compilers are mostly CPU bound. secondarily I'd say they are memory bound, that is memory bandwidth, not memory size.
that ht gives as much as 70% speedup isd proof that it is not 100% cpu bound. if it was 100%, I'd expect very little speedup. 70% speedup tells you that we have exactly the kind of workload ht and other mt was invented for. in old days without multithreading, a core would do only one operation per cycle, in between the pipeline, staging, bus requests etc. every time it missed a cachge lookup,ba page fault, we had to context switch, and wait until next cycle to try an operation. with ht,bwe have always multiple pipelines lined up, and when he is out because of page faulkt another is ready to go. basically, wghat my does, is to allow a core to be 100% utilized. that's why some workload see improvement with ht, while a few even see degradation. a workload that is truly 100% CPU bound might see degradation. workloads that are mosrptly cvpu and memory, will see greatest improvermernt. in the case above, they might benefit from even more than 8 threads, if thecore is not 100% utilized yet. it depends how much cpu bound vs memory bound the process is. now the core would be idle in those cases where more than 4 threads had a page fault in same clock cycle. now you understand why processor cache vsize is so important. ram access is extrermerly slow compared to CPU clock cycles.
dead tree books are buggy whips. if you haven't noticed, a lot of book stores have closed shop. borders are gone. and it is not just because of amazxon. when is the last time you looked up in a dictionary or a lexicon?
if they are state that only 1 in 25 don't listen to radio every week, theyre lying. my car had as radio, but thastsd because I've been to busy to replaced with a digital media player. I only listened to that radio maybe 5 weeks per year, and only npr, and never music. I know lots of others who also don't have as radio in the house. radio is as retro format like vinyl.
As much as we want?
If you compare world food procdution to world oil production, and keep in mind that many already do not have enough food, and also keep in mind that modern agriculture is a large fuel consumer, you would have to eat those words. There would nothing else for dinner.
Thats what the grid is for. ,ix of souces, including solar, wind as well as a common energy storage facility.
If the grid is powerful enough, you will export your excess to the grid when your storage is full and import from the grid when your reservoir is gone.
If there are some hydroplants on the grid, excess solar power could be used to pump water back to the reservoi during daytime when the turbines are running at a tricke, and crank the turbines up to full load at night when all the solars are down.
Excess energy could be used to produce hydrogen or methanol for fuell cells. Maybe a generator per household would be overkill, but one per neighborhood oughta do.
Energy storage is not trivial, but it is a far cry from impossible.
I can imagine an energy coop would want to have a
If the purp0se of the website is to recruit suicide bombers and bloodthirsty killers to attack innocent 3rd parties, then you can consider it part of the infrastructure of a hostile army, and a first class target for anyone wanting to interrupt such monsters. If you go all meta and pedantic, you have already lost grasp on reality, and need a brain check.
Data warehousing?
Realtime realistic light an texture 3d graphics rendering?
Hosting dozens of virtual machines? This on i find most likely, especially if needing tet, dev, demo env etc, or if need to simulate large complex sw running across several servers. A mid range x86 desktop can easily outrun yesteryears refrigerator sized unix servers.
Single host massive denial of server attacks?
Or just absolutey useless waste of time an money.
Or just a pack of lies.
Personally i have set up servers with more than 1 tb ram, but they werea actually needed for a specific workload.
I can say the same about my i3, or even your average cell phone chip that has ony 2 cofes, eacch being only 1 5th as powerful as an i3.
Disk io, network io, mory bandwidth are sill the most common bottlenecks. If a regular desktop user has bad response times due to cpu shortage, there is probably a software bug, or misconfiguration.
The dirty little secret that is shared by everyone who has a chttp://www.aftenposten.no/nyheter/uriks/Slik-smugles-folk-til-Europa-7809354.htmllue, is that the cpu hardly matters anymore, except for a few applictions.
Your regular dick jane and joe that walks into bestbuy getting outfitted for college or needing a machine to do their acconting and surf the web shoould not even be looking at the cpu model. Theyd run just fine on an atom or celeron, or a pentium Ii from the 90.s.
All that proves is that hyperthreading helps. It would still be just 4 cores.
Bah.
Not interested in the quad core stuff.
The only thing i want from this family is as powerful as possibe dual core chips.
Dual core xeons are pretty much extinct by now , but there are a few servers sold with core i3 and i5 instead of xeons.
You see the licensing pricing struture for software such as oracle db, tuxedo, weblogic, and others are per core. So it has become tricky to source a mainstream server model and ke ep the layered product licensing under $100ks of dollars. Since the recent x86 cores are 10-20x as strong as some old unix cores , you can replace an old refrigeratior sized unix hog that cost a million dollars, with a single x86 blade module for a couple of thousand dolllars. But you still end up paying big $$ to oracle.
And dont tell me we need to roll our own hw. We have to pick from mainstream server models that can be sourced and supported in evey country.
Solving this challengeis is where the big money is right now. Being able to scale down hw for commercial software to take advatage of these fast cores. There are solutions, but none are optimal, and often come in conflict with the client company's irrational and rigid rigid it policies. Scaling up is laughably simple. Scaling down is a real head scratcher now that eveything is getting bigger.
card sharks don't make a good living by knowing when to walk away. if they did they would have to retire after a few days.
they make a good living by exercising their skills in a skill based sport against other less skilled players. sometimes evern the sharks fall victim. sometimes their deep bankroll drunk millionaire victims get lucky. some times they run into other sharks. the best profesionnal gamblers avoids gambling. they do not take bets where they hold the short end of the stick. if they best on sports, they must be better than their bookies ot betting public at predicting outcomes. the opponent is not lady luck, lady luck only defines the outcome of each bet. the difference in skill between the player and other players defines the outcome and in the long run. if playing poker, they must consistently play at tables where they are there strongest player. in theory if you only ever accept profitable bets, you will win in the long run.
your way of thinking is why operating a casino is such a lucrative business.
every real poker player only plays when he has an edge, which means he has an advantage over the other players.
if you enter a game hoping to break even, you are just wasting time.
any bet has a given chance of winning their pot. a poker player should know what the chance is for hitting the card he needs, and based on history and bidding, what cards the other players have, and what cards he wants the other players to believe he has.
there is no wnnng strategy that says play a few hands hope to get lucky.
they can't. they had to practice before entering liver tournasments. and for all that succeeded there were 2 that failed. a tell is not necessarily a grin, a tick or some hand movement. it is really hard to control your pulse. a real shark will be able to study the pulse that is visible in you throat. which is why many poker players dress a bit over the top. but the most visiblke tell is the play itself. how often you see the flop, hoiw often you bet, rasiser, fold. what hand values you bet with etc. the visual clues are just extras. who knows, maybe there are even ubersharsk that know how to manipulate their pulse and leave fake tells now and then, when the stakes get high enough. if you hold the nuts in a half million pot, you have to do all you can to make the others believe you hold less. maybe seem worried that one of them have the nuts
wrong.
if you Always play the math, you opponents also always know your strength.
they might fold medium hands early instead of wasting 10 big blinds calling you down to the river.
in other situations they will know that you hold a big pair or high cards, so that your small straight is safe.
a human could play close to even with the bot in all the vanilla hands, and for a certain number of situations, exploit the fact that the robot is a robot. so human wins. a losing strtategy would be to be clever you n every hand. just play perfect like the not, be patient and waitfor the situations where the math is what leads the bot into a trap.
you math is wrong, you don't lose 0.5% in the long run. you lose 0.5% per game in the long run. After a few hundred games the average player will have lost it all, not just half a percent of their starting capital. To NOT loose it all, one has to be extremely lucky, plus one has to quit, or keep sloppy records, forgetting about the many losses and remember the few wins.
agreed. the onmline casinos makes large sums on the rakes and feed.
they win, regardless of who wens easch hand. if they were stealing from customers that way, they would be killing their golden goose. no need to cheat the schmuck out of $500. $500 will soon enough be won via rake, and then another $500 deposited,
I had a mysterious issue on my washing machine(20 year old GE top loader) , seemingly the dial did not work right, and often, the machine would get stuck on a part of the program, usually the spin cycle.
This was going on fro a couple of years. As a rule, I never would leave the house with the machine running, in case it would get stuck, and end up burning up the motor and start a fire.
A couple of months back I figured it all out. It is that stupid safety switch. The lid of the machine is just slightly bent , you can't even see it, but it is enough so that the safety switch is only barely connecting when the lid is closed.For now I just put something heavy on the lid, maybe I'll get around to fixing it permanently one day. I got the machine dirt cheap off craigslist. I bet the original owners would have kicked themselves if they knew what ailed their machine........
At the same time they can fix fewer gadgets, as well as more gadgets.
And another reason why we often don't WANT to fix something is because it is too expensive to fix.
Point in case: In the old days, if your radio or TV stopped working, you would take it to a Radio or TV repair shop, and they would fix it for much less than it cost to buy a new one. People would go to TV and Radio repair vocational schools. Now Radios hardly exist, and if the TV breaks after the warranty time, it is cheaper to toss it out and buy a a new one rather than finding someone who can repair it.
I am mostly clueless about electronics, but when my old monitor broke, I was determined to fix it. Googled and found videos on youtube showing me how to go abnout it. Learned that there are things called capacitors in the power supply, and my particular model contained a batch of capacitors that were often failing. When they swell up, they are probably toast. Bought soldering iron and other useful tools. Ordered the spare capacitors, and managed to get the monitor working. The monitor was worth maybe $30 tops, and I probably spent 5 hours in total on the repair and $100 on tools. Could have bought one twice as big, lighter and faster, with modern connectors for a little over $100. Later I was however able to fix my neighbor's laser printer by following same method, looking for bulging capacitors to replace, and I tried a couple more repair project where I was not successful. I have earned the cost of the tools at least. However, I _should_ have just tossed the monitor, if I had acted in an optimal manner. If it had been 1990, I would not have learned how to do this, as there was no youtube to teach me. Such a monitor would have been worth $500 and it would have been worth it to pay a repairman $100 to fix it.
It is not just a generational ting. It is a wealth thing. In poor countries (compared to tech today, UK and a lot of Europe could be considered "poor" in 50's compared to now. Less than 1 car per household, no TV, maybe a radio, working class jobs, hand me down clothes were the norm. So it woudl make sense to repair things and learn how to repair things. Now youth spend small fortunes on brand name clothes. And for the cost of a 1950's TV or radio (in 1950's dollars), you can now buy a ton of advanced electronics, and most of the gadgets are not SUPPOSED to be repaired. If there is no repair shop for a gadget, even in a medium to large city, how are kids supposed to get the idea that they should repair it themselves. Broken items are just taken in return and replaced with new product, the bad ones shipped to china for refurbishing, and the resold. Also, a gadget made today is only supposed to last a few years. My family's black and white TV purchased in 1970, was our only TV until we upgraded to a color TV in mid-80's. I'd like to see someone still running their windows 98 PC today. Or their Nokia analog cell phones? Products today are mostly obsolete by the time they need to be repaired. You can't blame that on the kids.
In poor countries, where they can afford few gadgets, there are plenty of cell; phone repair shops. 20+ year old cars are repaired. Parts are machined. People repair their bikes, using only hammers, nails and pliers as tools. It cost under $1 to get bikes repaired by professional bike repairmen slightly more skilled in the use of hammer, nails and pliers, but in a poor country, $1 is not something you waste on things you can do yourself.
SO the professor can rest his worries. There are plenty of people that fix whatever they can fix. Some in rich countries tinker and repair for fun, but in poor countries they still do it because they can't afford not to. If you add a few more dimensions to the picture, such as economy, history and geography, nothing really has changed with the kids.
It can in fact win any hand, but it is proven to be the hand with the lowest expected value of all opening hands.
However, in a very tight game, you can win with this hand if the others are bluffing.
The pros play this kind of hand when the situation is right. Bluff to steal a pot, limp in a pot with nice pot odds, or just on a whim to mix things up.
And in heads up, since the opponent on average will hold only an average hand, you can use 7 2 for bluffing, getting the other to fold if he only has a medium hand (2/3rds chance of that). And if flp falls just right, it is now a semi bluff with chance of big payoff, because you will never be put on that hand.
Sometimes you don't care what hand you have, you are making bets just to find out what other guy holds, then make a bet that forces him to lay down exactly that kind of hand.
The difference is that teh robot will always fold this hand, humans will fold it ALMOST 100% of the time. Most of the times they don't fold it, they will lose extra money, and ocasionally they will make a lot of extra money. And by being seen to play that hand now and then, you can't be treated like a bot, which makes it harder for others to figure out what kind of hand you hold based on your betting pattern.
From what I have seen, the machine played against itself. It does try to find out what your hand is, but assume that you would be playing the same perfect game. .5, folding is worth 0.33, raising is worth 0.60. So it will chose to raise. If you deceived the bot by betting on a weak hand that happened to hit the flop nicely, you can certainly use that to milk the robot for a few extra $$. Since he must follow it's heuristic perfect strategy, he is forced to fall hard into all the traps you lay.
It might not even need to find out what your hand you really had. Since it calculated all possible permutations, it knows that betting in a given situation is worth
The biggest problem with playing against this bot, is that it is not human, so you can't trick it into deviating from its perfect strategy. So the question is whether you can make enough bonus money on laying these traps to cover the cost of the money wasted by temporarily deviating from a perfect strategy. The extra money you make by knowing exactly what hand the computer is holding is however yours to keep.
You never know exactly what someone holds, but based on the bidding action you can put them on a range of hands.
And further, once you see the flop and they bid again, you can narrow it down further, and to test things out, you can check, bet, call or raise yourself, and see how they react. Once you know exactly what rules the robot follows in any given situation, you know what hand kind of hand they hold.
The difference between a human player, who does not claim to play perfect poker, is that you can be 100% sure the robot will follow its own rules, even when it is bluffing.
Example: when betting it will use a range of opening hands. Heads up that means above average hands.
Those hands are pretty well known for both humans and robots, and there are also some hands in between where betting is more situational,
In the case of a human, he will usually fold a hand like 10 3 offsuit.
In the case of the robot he will always fold 10 3 offsuit, That little piece of information can be gold worth after the flop, turn and river.
Human to human, if a hand like 10 3 is the only one you are afraid of after seeing the river, then you can still hit the brakes and avoid 3-betting, because you know mr XYZ will and can sometimes play that kind of hand.
Human to Robot, the human knows the robot would have folded that kind of hand pre flop, so when we get to the river, and that hand is the only one we are afraid of, we can bet with confidence, as we know the the robot is not holding that hand. Maybe something harmless like AA or AK.
Every single hand of poker is a game of finding out as much possible about the opponents hand while revealing as little as possible about your own hand.
For the most part, both will have to play it straight in order to not throw money away, and now and then the human can get an edge by taking advantage of such information, and at other times the human can take advantage by spreading misinformation.
And while the number of possible actions (bet call - raise check etc.) give a mindboggling numebr of permutations, a human will synthethize those to a much smaller problem set, Strong weak, medium opening hands, what hands to fold, check, bet with after the flop given different flops etc. A human spending some time playing this robot will remember how it acts in different situations. And if it is 100% predictable, the human can likely play only slightly worse by trying to follow same "perfect": rules, and try to cash in big in those spots where the human gets an informational edge. Since the human knows the computer's brain and not vice versa, the computer can't get away fro those situations, it will fall into the same trap every time.
The only problem, is that if your opponent KNOWS this is your strategy, he can adjust. He will let you bet your slightly above average hand, and call only when he has a much better than average hand. And it is not like you will get to steal every blind, only the ones where you chose to jam the pot. And that has to be more than 50% of the hands if you want to do more than averaqe number of pot steals.
So the best way to beat such a guy is to play natural with medium as well as big hands, and wait for the opportunity to call one of the jammed pots with a top hand. Then your strategy is no longer perfect. And since you decided to jam the pot with anything better than pair of 6 or whatever it is, you can't get away when the other player calls your pair of sixes with pair of kings.
The jam or fold strategy is however a very effective strategy for the novice, even in earlier phases of a tournament(earlyu on they can use the strategy only with the top hands such as AA,KK,QQ,AK) , because that eliminates any chance of being outfoxed by a better player. But again, if they know this is your strategy, they can take advantage of it, and will not call you except when they hold aces.
Yep. Or the angry drunk player, who bets every hand aggressively.
I was once playing limit poker online at a table (not high just $1-2, or $2-4), and this one guy became aggressive after a bad beat.
There were 3-4 other players. To make it big on that table was as easy as betting hard on every hand that was a bit better than average, because teh drunk woudl always pay you off. Shows to prove that sometimes one has to adapt th1e strategy to the situation, and play profitable poker is far away from perfect poker.
OK if that was some rich asshole blowing off some steam with his luch money, but since I quit poker, I always keep wondering if that was some college kid gambling away his tuition, or some dad gambling away his family's food budget.
Bluffing is only small part of the deception.
It is called misrepresentation.
You are trying to get the opponent to believe you are holding a certain hand.
Ultimately you want the opponent to make the wrong decisions, either by underestimating (calling your honest bets) or overestimating (folding your bluffs).
And within the action of the same hand you want to remain somewhat credible, so he will believe your story.
But you want to keep it also as random as possible, so they can't be too sure what you have.
You also are projecting a certain playing style, the careful, tight, the aggressive tight, the lose, the crazy etc. And during the same game, you might have to switch personalities. You can even have a reputation (deserved or not) that precedes you. If you are known as a crazy player, let them think you are, and play tight, while acting crazy. Online, players will even pick handles to project a certain style.
You will take losses on purpose if necessary, to help build a perception, but you can't overdo it, as doing it too openly would be like bad acting.
Then you have semi bluffing.
You bet hard on a hand that isn't much yet.
The bluff works and you get the pot.
And if it doesnt work, there is still the odd change that the flop gives you something good. And when semi bluffs turn out good hands you often get paid off really big, because the opponents have you pegged on completely different hand than what you have.
OK, it is true, that there are optimal plays. And heads up, it is possible that the computer might have an edge simply by not being human, because in heads up limit poker, you have to play and bid so aggressively to play optimally, that it is it counter to most humans risk acceptance levels. And the other problems with playing the computer, is that you can't teach him a lesson to get more respect. If a human is getting too aggressive, you can knock them back down by check raising a few times, and bleeding and wounded, they will back down, and respect you next time you swing the whip. If the opponent is getting too timid in limit poker, consider yourself lucky, and steal blinds fro as long as it lasts. A computer will forget about those incidents, and won't be impacted by those mind games. But if it doesn't get involved, and insist on playing like a robot, it will lose some extra big hands because the human can be pretty sure what hands he has based on its optimal betting patterns.