Results of the week : (From the 04/02 to the 09/02)
Have a nice weekend everyone !!! 
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Notifications
Clear all
Your Poker Blogs
489
Posts
19
Users
189
Reactions
53.7 K
Views
@deralte71 Thanks man. Things are way easier when you don't struggle with rake and when you can pick-up the right opponents.
My graphs on pokerstars were quiet similar if you look back except the rake is pretty high in my opinion so not worth spending time playing higher on legit rooms.
Despite this i still experience variance between 30 to 55 BI max but it's easy to overcome it with medium volume when the rake doesn't get in the mix.
Rake when you get the best of us ... 
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Last week was pretty busy, gf had a week off and stayed in my place so no way to play or work.
Well atleast not a poker kind of game. 
I should be back at tables next week, hope everyone had a super week and is having a super weekend.
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Man you Have done Epic Job after My short hiatus! Good to Be Back now with spins and MTTs. GLHF as always!
Hello oldschoolpapa.sorry for my english im from germany.i tried managemywindows and its helpfull can even manage my console and powershell windows so thank you.I registered on the site to seek if contents were different from others but sadly not really. somehow looking into forum i found yours. ive been reading it from start and learned even few things i didnt know about pt4 and pokerstars.i also think your analysis is right about icm and nash. What you did on ICMExplorer was really funny. did you do the same for other poker tools ? you should consider moving to a forum like 2+2.
Also i write you because i work in a lab and we are looking into bigdata and clustering. we never used it before so i would enjoy if you have more tips.
interesting books to read? what kind of database do you use? how do you manage clusters? do you think raspberry pi can be reliable in a data scientist environment? i tried to add you to my friend list but seems my options are limited.do you mind if we talk in pv?
@Team0zum Thanks man ! Good to see you around, it's been awhile. Indeed when i started to get back to poker you kindly helped me out to get on the right tracks and i'm greatfull for it. Welcome Back and Wish you the best with Spins and MTTs. Happy C-Net Buddy !!!
@deralte71 Yes indeed it's way much better when it's because of my gf rather than my health. Even if my health isn't in perfect shape and probably never will be i have more freedom in my activities, work, hobbies and can manage the neuropathic pains way more smoothly wich is a relieve for me.
@BloodyLea First of all Thank You, i'm glad the little piece of tool i made could help you and obviously welcome on Pstrategy. Also don't worry about your english, it's all good and i'm an expat french so i'll definitely not judge your english skill.
Regarding the content of poker sites, well i cannot really tell since i'm not a pro poker player and don't plan to be. However most contents online are simply more or less short resume of poker books litterature and won't re-invent the wheel. For beginners or loosing players those contents and videos are way better than no strategy at all. To put it in an other way they are less loosing strategies than no strategies. Concerning videos and streams around i'm not a huge follower, when i watched videos it was more about how to make my own analysis on my own game, the different use of trackers, etc...
I don't see much interest in watching someone else playing but it mainly depends on your own way to learn. I prefer books , run my own simulations modeling opponents and threw my own practice. Still i enjoy the personality of some streamers and like to talk with them, you can find really nice and interesting peoples around.
If you refer to that thread ICMExplorer ICMExplorer should be called SCAMExplorer even if the tool was mentioned in poker book litterature LoL. But no matter the market when there is a nest of peoples with big expectations you will always find peoples who will try to take advantage of it in a way or another...
Did i try to do the same things on other poker tools ? Hmm there i must use some conditional statements in order to avoid troubles since most of the tools are supported by Pstrategy wich wasn't the case of SCAMExplorer.
IF i had to look into PT4/TableNinja/HRC/HM Licensing for intellectual purposes i would say they are more secure than SCAMExplorer but not that much.
IF i had to look deeper into PT4 then i guess i would probably find two ways to do the same. First one could be by looking wich registry keys are called when you run PT4 and i probably would notice two registry keys that could be WinLicense Keys and if i had to guess removing those keys could eventually reset the trial period. Now the other way would requiere more reverse engineering knowledges and tools and would require some patching.
IF i had to look into TN then i would also focus on the trial and instead of two registry keys i think i would look for one and only one then the job would be done without making anything complicated.
IF i had to look to HRC well then i would eventually look at the registration's window and pay attention to wich informations are requiered to get a free trial and then i imagine i would probably use a VM, make a script to change some computer's informations and a provisory mailbox. I imagine the other way could be done threw patching but then the if statements will last way more..
IF i had to look into HM it would have been many years ago when it was HM1 & HM2(first versions) and for HM2 i imagine i would create a false ".lic" file into the main folder and i imagine i would probably patch some .dll without forgetting the plugins dll files... but i simply imagine and today i assume they increased the security ... or not ...
About Pstars i won't say anything more even the conditional statements won't be good.
However i want to insist on the fact that it would be only out of intellect curiosity and to self-challenge myself so no need to message me asking for patch i gave enough clues.
About 2+2 forum or forums in general, i'm not so interested in. I'm only around for a momentum. In something like 6-9 Months i won't be much around so no need to register to many forums since it won't teach me much more. However i met nice peoples here and will stay in touch with them.
Well Concerning BigData. As i said previously BigData is somewhat a wide world grouping many different skills (Networking, Electronic, DB management, Coding on a wide scale,...). For the books to read i have many on a GDrive so i can read them from anywhere, anytime when bored so if you interested i'll send you the link.
Won't post it here to avoid copyright troubles, warnings, ... and i also encourage you to by them if you enjoy the books after reading.
For the DB i use MongoDB but it doesn't mean you should use the same, it simply match my needs.
As for managing the clusters in short i use a Connectivity model based on a customised Agglomerative Hierarchical Clustering algorithm so i don't need to specify a number of clusters and it's not sensitive to distance metrics. Again the reason is because it does match my needs.
Rasberry Pi in a datascientist environment... That's an other wide question, giving you an answer without knowing the dataset and application won't help you much unfortunately, however Rasberry Pi allows us to do tests in a cheap way especially if your lab never worked on BigData it can help.
If your work involves heavy calculations and requieres you to keep the clusters running 24/24 it won't be a good investment. Heavy Calculations and a 24/24 work will increase the heat of the CPUs as result you will experience lags , system failures and after few months the boards will die. On the other hand with a normal use and radiators+fans or watercooling it will work smoothly.
Also if you think about matching one of the TOP500 supercomputers in the world, clearly forget about it.
The metrics generaly used to rank them are the FLOPs, "floating-point operations per second"... Gflops/Tflops/Pflops
official ranking of the TOP500 is made using HPL benchmark ("Highly-Parallel LINPACK benchmark") and use the Rmax, Nmax, N1/2 & Rpeak to define the rank.
You also have the GREEN500 wich use the FLOPS per Watt to rank them by energetic efficiency.
Before getting into BigData i suggest to plan well ahead your project and to start with the cheapest hardware, except concerning the network hardware because cheap means you might end up with loss of packets, bottleneck troubles on your network.
Also use the best quality Ethernet Cables matching the hardware speed and the shortest length possible in order to reduce latency as much as possible and to avoid interferencies.
To make little tests before to invest in expensive hardware you can use Raspberry but not only, you have alternatives like BeagleBone Boards, ESP SoCs and boards (Even cheaper than Pi), Parallela Boards.
Personally if i didn't expatriate myself i would use Parallela Boards 16-core Epiphany RISC SOC + Zynq SOC (FGPA + ARM A9) , Gb ethernet ...
Now there's also the coding side, knowing algorithms doesn't mean you have the skills to code algorithms ...
To get a taste of it you should try using Python or R, many sources and libraries are available. Data Scientists usually use Python & R.
Concerning the DB you have many choices, for instance big corps with large datasets use Cassandra (e.g: Twitter, Netflix, Reddit, Cisco, ...).
You also have HBase with modular and linear scalabitlity, auto failover support , strictly consistend reads and writes...
MongoDB i use is a NoSQL database, stores data in flexible, JSON-like documents so large field and data structure is scalable, ad-hoc queries, real time aggregation, well documented ...
Neo4j/OrientDB/CouchDB/Terrsore/FlockDB/Oracle.I won't make the pros/cons for each coz you can find it online.
Roughly speaking you have the hard and soft clustering. i leave it to you to look into it else the post will expand endlessly.
In short you have density / distribution / centroid / connectivity models types of algorithms.
Easier to implement and most used is K-Means Clustering wich is an iterative clustering algorithm
The DBSCAN or Density-Based Spacial Clustering of Applications with Noise, in shorts you starting with a random unvisited starting data point and all points in Epsilon are classified as neighbourhood points.
the Mean-Shift Clustering wich finds a dense areas of data points with for objective to locate the center points of each group ... selecting the right window size or radius t is painfull ...
the GMM wich is the Gaussian Mixture Model, similar but more flexible than K-Means
The Agglomerative Hierarchical Clustering top-down or bottom up
and we can go on since it's non exhaustive but let's say those are the most often used with few variations.
Now if we go to machine learning, let's say we have two big families of learning algorithms.
The Supervised Learning wich includes variables as predictors and variables to predict wich are tied and trained to make the futur prediction.
and two sub categories Regression / Classification.(feel free to look into it...)
The Unsupervised Learning with no variable to predict. data have only inputs with multiple variables to describe the data and that's where clustering comes into account. Let's try to take a shortcut, clustering is an unsupervised machine learning method that tries to find natural groupings and statistical distribution of data.
Tough doesn't mean Supervised and Unsupervised learning cannot co-operate or interact.
i guess i tried to make a very short approach even if i took many shortcuts, i hope it can help atleast to have a bigger picture in mind.
To conclude, if you work on a project with a short dead line your lab should consider hiring someone with the right skills to work with you.
And yes if you just registered on Pstrategy you cannot invite peoples or post pictures so i'll add you on my side.
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Thank you OldSchoolPapa.very informative post. you give me a lot to think about. sent you a message.
Best Regards.
Since traffic of players dropped down on SNGs had to register to an other room to keep on some volum so this week i played on ACR + 2 Non Legit Rooms.
ACR client is a bit cheap and buggy but well i don't care much. played 5.50/5.45$ (Regular/Turbo) by day since low traffic and added few 11/10.90$ by night.
In few weeks i'll add some 16.50/16.40$ and maybe a 22/21.75$ Here and there if i get the right timing to register on them.
If anyone heard of some poker rooms with SNGs entry fees under 9% i'm interested legit or not legit.
Result of the week on ACR: ( From the 18/02 to 23/02)
Results of the 2 non legits rooms are also positive but i'm too lazy to do screeshots for each of them and upload
Happy Weekend Everyone !!!!
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Hi OldSchoolPapa!If you come around i would like more informations concerning the answer you sent me and also how did you get a report from an other lab doing similar researches?Did you get time to look into the matlab simulation?if you need a license for matlab just send me an email. why did you say it wont be profitable for our lab?on our sim and the report you sent it looks like it will be good.
Regards
@BloodyLea Break Time before next session. Had a busy week so didn't connect much.
i gave a look into your matlab script and it looks good atleast in theory except that it's for an other kind of application and 1/3 of the script has been copied from an other source. Thanks for the license offer but i use Matlab , Maple and tools alike since a very long time so no need. Also i didn't say it won't be profitable but it might not be profitable. Seems you take for granted that a positive match found by using a generative adversarial networks "GAN" is a true positive match however in those positive matches you will have false positive matches, some true positive matches that will end up on a dead market and few true positive matches that might lead to a profitable exploitation.I assume in your range of application checking if a positive match is a true match does have a cost.
So without running your sim i can tell no matter how good the numbers might be they are far from beeing accurate in your range of application.
If i have a little time during the weekend i'll make a sim you can run to have a more accurate overview.
About other labs doing similar reseaches, well it's not that difficult. The method you want to apply is used since a long time maybe a decade if not more, i just had to use some social engineering skills and ask nicely a russian lab that dropped the same project because of profitability.
Will send you more data and informations during the weekend.
@Team0zum Thanks man, i'll look into it during the weekend. What matters to me more than the softness of the game is to have the lowest SNGs entry fees. I prefer games with push fold lovers (good or bad ones) since a moment i got tired to be raked by big poker rooms.
Best of Skills & Luck @ Tables,
OldSchoolPapa.
ACR Overall Results: (Only 5 Days weekly)
Overall results on other crypto rooms are good. One almost breakeven but only 40 games played due to traffic and on the other one +112$ for 87 games played.
Happy Weekend Guys, Ladies
Best of Skills & Luck @ Tables,
OldSchoolPapa.
nice graph, mine has looked like an alpine skislope the last couple of days, but have in general got my money as a favourite, PLO is tricky.
Well like I run at the moment, mine probably looks like a huge big slide - I do not even wanna look at it 
Winamax and Unibet are full of totally clueless players. You can see My latest Post on My spinblog for few examples. Also My signature has link for My Instagram which has Prime examples on hands i come across there
@deralte71 Thanks man. Actually i mix a bit stakes even if i don't like much to mix because it can brings some false impressions of crushing a game when we just get lucky taking the up waves at the right moment. About PLO can't say i'm a PLO Noob didn't try it yet but once i'll only play for fun i'll give it a go.
@Prim Don't worry about your graph, turn your screen upside/down and you gonna see an amazing graph
Positive Attitude !!!! 
@Team0zum Thanks Buddy, i'll look into it but i think asian poker rooms might be more interesting to me. must look at winamax and unibet fees for SNGs T. if it's higher than 9%, i will definitely add an asian poker room even if i'm not supposed to have the right to play there. 
Best of Skills & Luck @ Tables,
OldSchoolPapa.
Aaaaaw, THAT IS IT, Papa! Problem solved thank you very much! I just turning my screen upside down and there´s no downswing anymore 
Thank you for cheering me up 
@Prim Always happy to help
Go Go Go Sharky Grindette Girl !!!
Best of Skills & Luck @ Tables,
OldSchoolPapa.