GTO PioSOLVER - postflop equilibrium solver for Holdem - Page 144 - Poker Software -- Two Plus Two Poker Forums

Click a button to show Strategy or Strategy + EV and then right click on 13x13 area. There are many options to export the results including .txt and .csv formats.

Quote

01-19-2018 , 07:16 PM

#3577

alwyscrzy4life

centurion

Join Date: Feb 2010 Posts: 189

Hi, I am debating between a few new laptops in which I am going to buy one and use PIO on it. First question is will 32gb instead of 16gb of ram make much difference in computing speed/power of sims? Second question is how much will the processor effect the speed/power of sims? I am looking at getting a i-7 8550hz processor, would that be sufficient or any recommendations?

Thanks!

Quote

01-20-2018 , 03:08 AM

#3578

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

First question is will 32gb instead of 16gb of ram make much difference in computing speed/power of sims?

Right now more RAM allows building bigger trees. You may want 32GB for some bigger ones. RAM doesn't influence speed at this point although it is likely to change in the future.

Quote:

Second question is how much will the processor effect the speed/power of sims?

Speed is proportional to number of physical cores multiplied by base frequency they run on. There is about 15% bonus for hyperthreading (so quad i7 over quad i5).

Quote:

I am looking at getting a i-7 8550hz processor, would that be sufficient or any recommendations?

I am unable to find this CPU. There are for example 8550u and 7700hq.
Let's see which one is faster:

1)8550u has 4 cores @ 1.8Ghz, 4 * 1.8 = 7.2
2)7700hq has 4 cores @ 2.8Ghz 4 * 2.6 = 11.2

The CPU in 2) is about 1.55x faster.
Sadly it's not the best time to buy a new CPU, especially from Intel. Latest exploit (Spectre/Meltdown) fiasco means that all existing CPUs are likely to get huge performance penalty with coming OS updates. It's hard to say how painful it's going to be but initial reports suggest it's going to be bad (admittedly we only got reports about performance on Windows Server which is supposedly hit the hardest).

Quote

01-20-2018 , 11:33 AM

#3579

Lanky

stranger

Join Date: Oct 2017 Posts: 8

Quote:

Originally Posted by punter11235

Sadly it's not the best time to buy a new CPU, especially from Intel. Latest exploit (Spectre/Meltdown) fiasco means that all existing CPUs are likely to get huge performance penalty with coming OS updates. It's hard to say how painful it's going to be but initial reports suggest it's going to be bad (admittedly we only got reports about performance on Windows Server which is supposedly hit the hardest).

For what it's worth, I've run the Pio benchmark with the Meltdown patch enabled/disabled and can't see a significant difference. I don't have a BIOS patch for Spectre yet so can't test that.

Windows 10 pro, E5-2650 CPU. 8.3s benchmark both with/without the patch.

Quote

01-20-2018 , 08:19 PM

#3580

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

Interesting. Maybe it's because of Windows verison (Windows 10 is supposedly affected the least and Windows Server the most) or maybe online.net does some shenanigans in addition to the patch.

Quote

01-22-2018 , 06:13 AM

#3581

supervagina

stranger

Join Date: Sep 2017 Posts: 10

Quote:

Originally Posted by punter11235

Click a button to show Strategy or Strategy + EV and then right click on 13x13 area. There are many options to export the results including .txt and .csv formats.

thanks

Quote

01-24-2018 , 02:16 PM

#3582

noxee

newbie

Join Date: May 2016 Posts: 37

Hi,

Quote:

UNFORTUNATELY it seems many Ryzen/ThreadRipper builds are unstable. Please make sure you know what you are doing when assembling those setups. A lot of our users experienced hard to overcome memory stability issues with those.

There are specific DDR4 kits for avoid this problem with a Ryzen 7 ?

Quote

01-25-2018 , 01:20 AM

#3583

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

There are specific DDR4 kits for avoid this problem with a Ryzen 7 ?

Google says that there is a list from AMD and also a list of BIOS versions for specific motherboards you need to get but I don't know enough about hardware to recommend anything.
What I do recommend is to get on our Skype group and try to copy one of the successful builds our customers assembled. There was quite a lot of discussion about it, maybe someone could chime in and make a post here as well

Quote

01-25-2018 , 02:31 AM

#3584

MU MUF MUFC OK

grinder

Join Date: Jan 2010 Posts: 438

is it possible to pause a running script?

Quote

01-25-2018 , 02:42 AM

#3585

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

is it possible to pause a running script?

It isn't. You can kill it it and re-run it though. As long as you don't move the save files from their original location the script will skip over already solved trees. Be careful not to kill it while it's saving though (you can look at the console window, kill it when a new tree starts).

Quote

01-25-2018 , 02:55 AM

#3586

HappyLuckBox

old hand

Join Date: Feb 2013 Posts: 1,491

hey punter, still having alot of problems with the latest release of pio and sinple trees being unable to converge. i never had these problems prior to the latest release...

heres a roughly 5gb tree that gets stuck solving:

Code:

#TreeBuilding#V2
#Range0#AA,KK,QQ,JJ,TT,99:0.75,88:0.75,77:0.5,66:0.25,55:0.25,44:0.25,33:0.25,22:0.25,AK,AQ,AJs:0.75,AJo:0.25,ATs:0.75,ATo:0.25,A9s:0.5,A8s:0.5,A7s:0.25,A6s:0.25,A5s,A4s,A3s,A2s,KQs,KQo:0.5,KJs:0.75,KJo:0.25,KTs:0.5,K9s:0.5,QJs:0.75,QTs:0.5,Q9s:0.5,JTs:0.75,J9s:0.5,T9s:0.75,T8s:0.75,98s:0.75,97s:0.25,87s:0.75,86s:0.25,76s:0.5,65s:0.25,54s:0.25
#Range1#QQ:0.25,JJ,TT,99,88,77,66,55,44,33,22,AK:0.25,AQ,AJ,ATs,ATo:0.25,A9s,A8s,A7s,A6s:0.5,A5s:0.5,A4s:0.5,A3s:0.5,A2s:0.5,KQ,KJs,KJo:0.25,KTs,K9s:0.5,K8s:0.25,QJs,QTs,Q9s:0.75,Q8s:0.25,JTs,J9s,J8s:0.25,T9s,T8s,T7s:0.25,98s,97s,96s:0.25,87s,86s,85s:0.25,76s,75s:0.75,65s,64s:0.5,54s,43s:0.25
#Board#3d 3s 2c
#Pot#450
#EffectiveStacks#7500
#AllinThreshold#80
#AddAllinOnlyIfLessThanThisTimesThePot#500
#MinimumBetsize#0
#UseUnifiedBetAfterRaise#False
#UnifiedBetAfterRaise#
#UseUnifiedRaiseAfterRaise#False
#UnifiedRaiseAfterRaise#
#ForceIPBet#False
#ForceOOPBet#False
#Cap#4
#CapEnabled#True
#CapMode#NoLimit
#RemovedLines#
#ExtraLines#
#FlopConfig.BetSize#33.33 66.66
#FlopConfig.RaiseSize#60
#FlopConfig.AddAllin#False
#FlopConfig.DonkBetSize#
#TurnConfig.BetSize#33.33 66.66 100
#TurnConfig.RaiseSize#3.5x
#TurnConfig.AddAllin#False
#TurnConfig.DonkBetSize#
#RiverConfig.BetSize#33.33 66.66 150
#RiverConfig.RaiseSize#3.5x
#RiverConfig.AddAllin#True
#RiverConfig.DonkBetSize#60
#FlopConfigIP.BetSize#33.33 66.66
#FlopConfigIP.RaiseSize#3.5x
#FlopConfigIP.AddAllin#False
#FlopConfigIP.Dont3bet#False
#TurnConfigIP.BetSize#33.33 66.66 100 
#TurnConfigIP.RaiseSize#3.5x
#TurnConfigIP.AddAllin#False
#TurnConfigIP.Dont3bet#False
#RiverConfigIP.BetSize#33.33 66.66 150
#RiverConfigIP.RaiseSize#3.5x
#RiverConfigIP.AddAllin#True
#RiverConfigIP.Dont3bet#False
#Rake.Cap#0
#Rake.Fraction#0
#Rake.Enabled#False
#ICM.Enabled#False
#ICM.Stacks#0
#ICM.Payouts#0

Last edited by HappyLuckBox; 01-25-2018 at 03:08 AM.

Quote

01-25-2018 , 03:28 AM

#3587

HappyLuckBox

old hand

Join Date: Feb 2013 Posts: 1,491

Quote:

Originally Posted by punter11235

Try to get the tridentz 14 cas latency kits, samsung b die works fine with tr

Quote

01-25-2018 , 04:03 AM

#3588

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

hey punter, still having alot of problems with the latest release of pio and sinple trees being unable to converge. i never had these problems prior to the latest release...

I am testing this and I will let you know but I just want to point out that a tree that has 2 bet sizes on the flop, 3 on the turn and 4 on the river isn't exactly simple

It again has small starting ranges which in combination with many bet sizes makes it difficult for the solver. I will check if old version handled them better and try to mitigate if possible.

Quote

01-25-2018 , 06:09 AM

#3589

HappyLuckBox

old hand

Join Date: Feb 2013 Posts: 1,491

Quote:

Originally Posted by punter11235

I am testing this and I will let you know but I just want to point out that a tree that has 2 bet sizes on the flop, 3 on the turn and 4 on the river isn't exactly simple

It again has small starting ranges which in combination with many bet sizes makes it difficult for the solver. I will check if old version handled them better and try to mitigate if possible.

Thanks punter!

I went ahead and simplified it further and reduced it to 2 bet sizes otf, 2 ott, and 4 otr, still with issues converging. Here's the code in case you wanted a more simple tree to look at!

Code:

#TreeBuilding#V2
#Range0#AA,KK,QQ,JJ,TT,99:0.75,88:0.75,77:0.5,66:0.25,55:0.25,44:0.25,33:0.25,22:0.25,AK,AQ,AJs:0.75,AJo:0.25,ATs:0.75,ATo:0.25,A9s:0.5,A8s:0.5,A7s:0.25,A6s:0.25,A5s,A4s,A3s,A2s,KQs,KQo:0.5,KJs:0.75,KJo:0.25,KTs:0.5,K9s:0.5,QJs:0.75,QTs:0.5,Q9s:0.5,JTs:0.75,J9s:0.5,T9s:0.75,T8s:0.75,98s:0.75,97s:0.25,87s:0.75,86s:0.25,76s:0.5,65s:0.25,54s:0.25
#Range1#QQ:0.25,JJ,TT,99,88,77,66,55,44,33,22,AK:0.25,AQ,AJ,ATs,ATo:0.25,A9s,A8s,A7s,A6s:0.5,A5s:0.5,A4s:0.5,A3s:0.5,A2s:0.5,KQ,KJs,KJo:0.25,KTs,K9s:0.5,K8s:0.25,QJs,QTs,Q9s:0.75,Q8s:0.25,JTs,J9s,J8s:0.25,T9s,T8s,T7s:0.25,98s,97s,96s:0.25,87s,86s,85s:0.25,76s,75s:0.75,65s,64s:0.5,54s,43s:0.25
#Board#3s 3d 2c
#Pot#450
#EffectiveStacks#7500
#AllinThreshold#80
#AddAllinOnlyIfLessThanThisTimesThePot#300
#Cap#4
#CapEnabled#True
#CapMode#NoLimit
#FlopConfig.BetSize#33.33 60
#FlopConfig.RaiseSize#50
#TurnConfig.BetSize#33.33 75
#TurnConfig.RaiseSize#50
#RiverConfig.BetSize#33.33 60 150
#RiverConfig.RaiseSize#50
#RiverConfig.AddAllin#True
#RiverConfig.DonkBetSize#60
#FlopConfigIP.BetSize#33.33 60
#FlopConfigIP.RaiseSize#50
#TurnConfigIP.BetSize#33.33 75
#TurnConfigIP.RaiseSize#50
#RiverConfigIP.BetSize#33.33 60 150
#RiverConfigIP.RaiseSize#50
#RiverConfigIP.AddAllin#True
#ICM.Stacks#0
#ICM.Payouts#0

Quote

01-25-2018 , 08:20 AM

#3590

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

I went ahead and simplified it further and reduced it to 2 bet sizes otf, 2 ott, and 4 otr, still with issues converging. Here's the code in case you wanted a more simple tree to look at!

So in general small ranges are often a problem for our solver (especially if overbets are present but not only then) because we implement a very memory efficient algorithm (so you can build bigger trees). We don't do averaging nor we keep any 2nd level info. It's likely that in the future there will be an option to use more RAM so such trees converge faster. For now I am testing some simple adjustments to see if I can make those trees behave. This will take a while (things take time to solve) but I will get back to you.

Quote

01-25-2018 , 02:45 PM

#3591

noxee

newbie

Join Date: May 2016 Posts: 37

Thanks for RAM advice.

I tried to update to version 1.10 with this updater : https://www.piosolver.com/blogs/news/piosolver-1-10

But now :

Quote

01-25-2018 , 06:13 PM

#3592

noxee

newbie

Join Date: May 2016 Posts: 37

^^

Apparently, the new updater can't find my older installation :
- Deactivate Licence > install 1.9.2 version (works fine) > retry for the 1.10 version > failed
- Deactivate Licence > install 1.9.2 version (works fine) > retry for the 1.10 version with the new updater in "C:\PioSOLVER" + admin mode > failed again :'(

Quote

01-25-2018 , 11:27 PM

#3593

HappyLuckBox

old hand

Join Date: Feb 2013 Posts: 1,491

Quote:

Originally Posted by punter11235

This is a nice question. I solved the tree to pretty good accuracy (0.05%) using other algorithm in my dev version and the results match yours so it's not an antic of our algorithm or anything.

Yeah it makes sense intuitively

The equities and EVs are really close as well so there wouldn't be that much difference if you choose other combos the the 3bet bluff.
I tried doing a bit more research by cutting the river part of the tree (you can do that with "create subtree configuration" option and solving the river itself to almost exact equilibrium but that doesn't change results either. The config is here if you would like to re-run:

Code:

#TreeBuilding#V2
#Range0#AdAc:0.0219419003,AhAc:0.0070840558,AsAc:0.660483897,AhAd:0.0073260814,AsAd:0.660220802,AsAh:0.675516725,KdKc:0.880723238,KhKc:0.723196805,KsKc:0.780178428,KhKd:0.830456674,KsKd:0.851661086,KsKh:0.779956222,QdQc:0.924114227,QhQc:0.898507893,QsQc:0.964862883,QhQd:0.925129056,QsQd:0.966405392,QsQh:0.963846266,JdJc:0.909975171,JhJc:0.91582042,JsJc:0.728330135,JhJd:0.910799205,JsJd:0.712088525,JsJh:0.729773879,TdTc:0.775788963,ThTc:0.811564088,TsTc:0.728380919,ThTd:0.770951807,TsTd:0.771416247,TsTh:0.735465467,9d9c:0.715767741,9h9c:0.6375615,9h9d:0.717788756,8d8c:0.512037694,8h8c:0.397080541,8h8d:0.503825366,7h7c:0.256605595,7s7c:0.684454322,7s7h:0.688322961,6s6c:0.0485383198,6s6d:0.0318844467,6s6h:0.0502031185,3s3d:0.3062796,AsKs:0.262039632,AcKs:0.0776879415,AdKs:0.110365152,AhKs:0.144671917,AsQs:0.457934678,AcQd:0.087693803,AcQh:0.194576994,AcQs:0.303382009,AdQc:0.189767286,AdQh:0.195012957,AdQs:0.307962418,AhQc:0.193566829,AhQd:0.192249998,AhQs:0.225977659,AsQc:0.601935804,AsQd:0.683414221,AsQh:0.592403591,AcJc:0.504177392,AdJd:0.656130195,AhJh:0.513765454,AsJs:0.929394066,AcJd:0.458968312,AcJh:0.509694219,AcJs:0.511149466,AdJc:0.368438423,AdJh:0.370998681,AdJs:0.650786281,AhJc:0.52207166,AhJd:0.463068247,AhJs:0.502039731,AsJc:0.849502981,AsJd:0.853805244,AsJh:0.848157823,AcTc:0.189572379,AdTd:0.160405576,AhTh:0.158684969,AsTs:0.583479285,Ac9c:0.807374239,Ad9d:0.870411694,Ah9h:0.783534646,Ac8c:0.357970029,Ad8d:0.377508879,Ah8h:0.338545591,Ac7c:0.00126288924,As7s:0.233469725,Ac6c:0.265112817,Ad6d:0.29027912,Ah6h:0.262129992,As6s:0.248190403,Ac5c:0.105496109,Ad5d:0.110499047,Ah5h:0.0913464278,As5s:0.708200991,As4s:0.511106312,As3s:0.150297612,KsQs:0.932142138,KcQs:0.734622419,KdQc:0.0178974755,KdQh:0.021556396,KdQs:0.868325889,KhQd:0.0500954166,KhQs:0.781119525,KsQc:0.830636978,KsQd:0.812657475,KsQh:0.825406432,KcJc:0.407478809,KdJd:0.654949009,KhJh:0.442079276,KsJs:0.650965929,KcJd:0.288416654,KcJh:0.309941977,KcJs:0.329607219,KdJc:0.180284858,KdJh:0.167445391,KdJs:0.384658694,KhJc:0.328122437,KhJd:0.312596947,KhJs:0.329607517,KsJc:0.369938135,KsJd:0.379353136,KsJh:0.369674295,KcTc:0.899984837,KdTd:0.91113919,KhTh:0.915904284,KsTs:0.651771903,Kc9c:0.691018939,Kd9d:0.734485984,Kh9h:0.694270015,Kc8c:0.0255194269,Kd8d:0.0401706845,Kh8h:0.0225690603,Kc6c:0.0951631889,Kd6d:0.0747317374,Kh6h:0.0981293768,Ks6s:0.0783317536,QcJc:0.899931788,QdJd:0.88927418,QhJh:0.89573282,QsJs:0.459242582,QcJd:0.166781634,QcJh:0.19323121,QcJs:0.118113592,QdJc:0.187390223,QdJh:0.192950368,QdJs:0.131056845,QhJc:0.193230033,QhJd:0.168853283,QhJs:0.115756646,QsJc:0.190919399,QsJd:0.194362909,QsJh:0.190921172,QcTc:0.752147794,QdTd:0.765352309,QhTh:0.753096998,QsTs:0.785680592,Qc9c:0.170938432,Qd9d:0.16846697,Qh9h:0.17671068,JcTc:0.825696886,JdTd:0.806214213,JhTh:0.826547205,JsTs:0.611718476,Jc9c:0.452726424,Jd9d:0.493313789,Jh9h:0.446383297,Tc9c:0.621301234,Td9d:0.616867423,Th9h:0.619497657,Tc8c:0.0307270754,Td8d:0.0458396636,Th8h:0.0325724892,9c8c:0.399849266,9d8d:0.419664353,9h8h:0.402422786,8c7c:0.351563573,8h7h:0.355463862,7c6c:0.445228219,7h6h:0.445012242,7s6s:0.614300489,6c5c:0.176719964,6d5d:0.203265235,6h5h:0.175940439,6s5s:0.109716542,5s4s:0.110275984
#Range1#QdQc:0.0799035281,QhQc:0.0353561565,QsQc:0.0498377867,QhQd:0.0799040422,QsQd:0.0521241426,QsQh:0.049837783,JsJc:0.0163502507,JsJd:0.0926097259,JsJh:0.0190844629,TdTc:0.465497673,ThTc:0.374448806,TsTc:0.362318307,ThTd:0.468162715,TsTd:0.318042099,TsTh:0.365964472,9d9c:0.19500497,9h9c:0.029605072,9h9d:0.194743216,8d8c:0.412894875,8h8c:0.122235924,8h8d:0.412893206,7h7c:0.127850324,7s7c:0.467168957,7s7h:0.467169881,6d6c:0.913999915,6h6c:0.913999975,6s6c:0.913947046,6h6d:0.913999975,6s6d:0.913955748,6s6h:0.913949788,5d5c:0.0302826688,5s5c:0.65058291,5h5d:0.140395358,5s5d:0.920912862,5s5h:0.711987913,2h2d:0.125510737,2s2d:0.47782743,2s2h:0.0646906719,AcKd:0.449999988,AcKh:0.449999958,AcKs:0.449999988,AdKc:0.449999988,AdKh:0.449999988,AdKs:0.449999958,AhKc:0.449999988,AhKd:0.449999988,AhKs:0.449999928,AsKc:0.449999988,AsKd:0.449999988,AsKh:0.449999988,AdQd:0.0255604293,AsQs:0.199921489,AcQd:0.239652783,AcQh:0.0442856103,AcQs:0.167088702,AdQc:0.0422793254,AdQs:0.141014844,AhQc:0.000478416972,AhQd:0.089952454,AhQs:0.137463391,AcJc:0.13208802,AdJd:0.486430645,AhJh:0.111672454,AsJs:0.406383604,AcJd:0.157404616,AcJh:0.143110394,AcJs:0.453653961,AdJc:0.0056119943,AdJs:0.353675455,AhJc:0.139644742,AhJd:0.139719307,AhJs:0.441163957,AsJc:0.232482284,AsJd:0.255033076,AsJh:0.234325171,AcTc:0.166619211,AdTd:0.144794971,AhTh:0.164131582,AcTd:0.386920989,AcTh:0.69599998,AcTs:0.125647515,AdTc:0.477006018,AdTh:0.477006733,AdTs:0.121452175,AhTc:0.69599998,AhTd:0.386919469,AhTs:0.133509278,AsTc:0.0774853751,AsTd:0.0803400874,AsTh:0.0705890357,Ac9c:0.792409778,Ad9d:0.731598377,Ah9h:0.792413831,Ac9d:0.500296772,Ac9h:0.484103173,Ad9c:0.509986877,Ad9h:0.509986877,Ah9c:0.484100074,Ah9d:0.500296652,As9c:0.476928473,As9d:0.478250444,As9h:0.476931959,Ac8c:0.850000024,Ad8d:0.849999964,Ah8h:0.850000024,Ac7c:0.75999999,Ah7h:0.75999999,As7s:0.759963036,Ad6d:0.0504573733,As5s:0.0795775428,As4s:0.0963536203,As3s:0.165015325,Ad2d:0.153490245,Ah2h:0.830964684,As2s:0.390399903,KsQs:0.34248817,KsJs:0.0486603826,KcTc:0.35139069,KdTd:0.149680719,KhTh:0.347628504,KcTd:0.269612551,KcTh:0.273412913,KcTs:0.107111275,KdTc:0.261020392,KdTh:0.261022419,KdTs:0.189073443,KhTc:0.273412764,KhTd:0.269612163,KhTs:0.112977147,KsTc:0.107420214,KsTd:0.0329438001,KsTh:0.111828856,Kc9c:0.931999981,Kd9d:0.930833519,Kh9h:0.931999981,Kc8c:0.885999918,Kd8d:0.886000097,Kh8h:0.885999978,Kc7c:0.850999951,Kh7h:0.851000011,Ks7s:0.716454566,Ks5s:0.132136598,Ks4s:0.089714475,Ks3s:0.267931223,Kh2h:0.629636288,Ks2s:0.631691396,QcJc:0.260306358,QdJd:0.16884245,QhJh:0.264923722,QsJs:0.271768421,QcJd:0.193138987,QcJh:0.150717378,QcJs:0.142024234,QdJc:0.210995466,QdJh:0.208187655,QdJs:0.165160611,QhJc:0.153890565,QhJd:0.184554666,QhJs:0.143815413,QsJc:0.102475621,QsJd:0.0936532542,QsJh:0.10074833,QcTc:0.585057318,QdTd:0.326690316,QhTh:0.584542215,QsTs:0.0367327295,QcTd:0.268105119,QcTh:0.0985839292,QcTs:0.138103843,QdTc:0.328086764,QdTh:0.329853058,QdTs:0.109219618,QhTc:0.0985834077,QhTd:0.268104166,QhTs:0.137203202,QsTc:0.172776133,QsTd:0.125106305,QsTh:0.175871,Qc9c:0.949999928,Qd9d:0.944416583,Qh9h:0.949999988,Qc8c:0.699999988,Qd8d:0.699999988,Qh8h:0.699999988,Qc7c:0.349999994,Qh7h:0.349999964,Qs7s:0.249374837,Qs6s:0.036803294,Qs5s:0.271187156,Qs4s:0.0784261823,JcTc:0.0361940227,JdTd:0.0139926337,JhTh:0.0356750973,JcTd:0.0277524926,JcTh:0.0422813632,JcTs:0.00534590939,JdTc:0.0542858727,JdTh:0.052935306,JhTc:0.0432710946,JhTd:0.0282778721,JhTs:0.00401936471,JsTc:0.0903524235,JsTd:0.040920224,JsTh:0.0947995633,Jc9c:0.787126303,Jd9d:0.616536736,Jh9h:0.78522712,Jc8c:0.800435901,Jd8d:0.716161072,Jh8h:0.801263988,Jc7c:0.719347119,Jh7h:0.718079448,Js7s:0.652256131,Tc9c:0.586380482,Td9d:0.300904632,Th9h:0.58689487,Tc9d:0.243050769,Tc9h:0.243581295,Td9c:0.25,Td9h:0.248699352,Th9c:0.243578225,Th9d:0.243281305,Ts9c:0.065957509,Ts9d:0.127940089,Ts9h:0.0659576505,Tc8c:0.79643029,Td8d:0.621307254,Th8h:0.796434522,Tc7c:0.829999983,Th7h:0.829999983,Ts7s:0.0351964422,9c8c:0.459295541,9d8d:0.384960681,9h8h:0.459704459,9c8d:0.075000003,9c8h:0.0218962803,9d8c:0.0587586425,9d8h:0.0596787073,9h8c:0.0218964648,9h8d:0.075000003,9c7c:0.66700995,9h7h:0.668178737,9c6c:0.570212662,9d6d:0.389172256,9h6h:0.57021153,8c7c:0.688345253,8h7h:0.688887775,8c6c:0.548152149,8d6d:0.285221964,8h6h:0.55090034,8c5c:0.590128362,8d5d:0.630100429,8h5h:0.58896023,7c6c:0.514640272,7h6h:0.514899433,7c5c:0.733046293,7h5h:0.731464744,7s5s:0.322755754,7c4c:0.749997616,7h4h:0.749997675,7s4s:0.428861856,6s4s:0.2501643,6s3s:0.198001832,5s4s:0.80847919,5s3s:0.247475281,5s2s:0.200000003,4s3s:0.561779618,4s2s:0.200000003,3s2s:0.5
#Board#9s 8s 7d 2c Jd
#Pot#1150
#EffectiveStacks#4435
#AllinThreshold#80
#AddAllinOnlyIfLessThanThisTimesThePot#500
#MinimumBetsize#0
#UseUnifiedBetAfterRaise#False
#UnifiedBetAfterRaise#
#UseUnifiedRaiseAfterRaise#False
#UnifiedRaiseAfterRaise#
#ForceIPBet#False
#ForceOOPBet#False
#Cap#4
#CapEnabled#True
#CapMode#NoLimit
#RemovedLines#
#ExtraLines#
#FlopConfig.BetSize#33 65
#FlopConfig.RaiseSize#60
#FlopConfig.AddAllin#False
#FlopConfig.DonkBetSize#
#TurnConfig.BetSize#33 65
#TurnConfig.RaiseSize#60
#TurnConfig.AddAllin#True
#TurnConfig.DonkBetSize#50
#RiverConfig.BetSize#33 65 100
#RiverConfig.RaiseSize#60
#RiverConfig.AddAllin#True
#RiverConfig.DonkBetSize#50
#FlopConfigIP.BetSize#33 65
#FlopConfigIP.RaiseSize#60
#FlopConfigIP.AddAllin#False
#FlopConfigIP.Dont3bet#False
#TurnConfigIP.BetSize#33 65
#TurnConfigIP.RaiseSize#60
#TurnConfigIP.AddAllin#True
#TurnConfigIP.Dont3bet#False
#RiverConfigIP.BetSize#33 65 100
#RiverConfigIP.RaiseSize#60
#RiverConfigIP.AddAllin#True
#RiverConfigIP.Dont3bet#False
#Rake.Cap#0
#Rake.Fraction#0
#Rake.Enabled#False
#ICM.Enabled#False
#ICM.Stacks#0
#ICM.Payouts#0

After combo-locking TT to always raise the solver in fact shows IP can adjust to that to get more EV.
It looks like suit distribution is important. If you look at the EVs then you will see raise with ThTd and TdTc is better than a call but we don't have those in our range (they don't bet 380 on the river, they make a bigger bet).

This is difficult to make sense of. I tried but wasn't able to find any intuitive pattern unfortuntely.

Hey punter, sorry for late response, I appreciate the effort though! Yes it seems very counter intuitive to me as well haha.

I was wondering if you had any networking or suggestions for me if I was looking to get involved in a weekly skype pio study group? I am willing to put my computer with 128gb ram and threadripper to work, also I'm a live cash 10-25 player, just looking to find a group who I can discuss and analyze pio strategies with.

Quote

01-25-2018 , 11:52 PM

#3594

HappyLuckBox

old hand

Join Date: Feb 2013 Posts: 1,491

Quote:

Originally Posted by Lanky

I think (not got this completely straight in my head) this has to do with the fact that villain's calling range *only* includes hands with at least one ten. Therefore having the extra blocker doesn't skew villains calling range away from QT - it actually skews towards hands with a single ten because your extra ten blocks a greater % of villain TT hands (2/3) than Tx (4/12).

Essentially you want villain to have a greater % of TT in their calling range and, when calling, villain wants hero to have a greater % of TT in the bluffing range. It's a tiny adjustment but I think that might be at least something to do with it.

That is interesting! I think I get it! I guess this makes the most sense, but intuitively, it still feels like TT is a better bluff and bluffcatcher. Since most villians rarely 3bet "bluff" with Tx here the optimal amount, I wonder if that skews it back towards TT as being the better bluff and bluff catcher. I guess I could simulate that.

Quote

01-26-2018 , 04:17 AM

#3595

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

Apparently, the new updater can't find my older installation :

When you click "install/update" there should be a field with a path to the installation. You can change it to your older one.

Quote:

Deactivate Licence > install 1.9.2 version (works fine) > retry for the 1.10 version with the new updater in "C:\PioSOLVER" + admin mode > failed again :'(

What do you mean by "failed" here? Does it not download the files? Does it download them to some location you can't find? Does the updater produce an error? I need a bit more information to help you.

Quote

01-26-2018 , 08:00 AM

#3596

noxee

newbie

Join Date: May 2016 Posts: 37

Quote:

Originally Posted by punter11235

When you click "install/update" there should be a field with a path to the installation. You can change it to your older one.

What do you mean by "failed" here? Does it not download the files? Does it download them to some location you can't find? Does the updater produce an error? I need a bit more information to help you.

Quote

01-26-2018 , 08:48 AM

#3597

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

The installation Directory is C:\PioSOLVER, are you starting the program from there?
If some files are missing you can always check "Force update" checkbox (visible on your screenshot) so it will download all the files again. Please do that and then go to C:\PioSOLVER and start PioViewer from there.

Quote

01-27-2018 , 10:40 AM

#3598

Ingenol

centurion

Join Date: Apr 2010 Posts: 155

Threadripper owners: are you guys using distributed (DIE/AUTO/UMA) or local (CHANNEL/NUMA) memory mode? I don't know too much about this but the benchmarks seem slightly faster with local memory mode, however this must be a small tree since it completes in ~2.2s.

Punter: do you have a recommendation on NUMA vs UMA memory access modes (just in general--I know you don't personally use a Threadripper)?

Quote

01-27-2018 , 02:06 PM

#3599

Ingenol

centurion

Join Date: Apr 2010 Posts: 155

Sorry can't edit but I flipped the last post: distributed seems slightly faster (~2.10s bench vs ~2.15s).

Quote

01-27-2018 , 04:07 PM

#3600

punter11235

Carpal \'Tunnel

Join Date: Mar 2005 Posts: 8,210

Quote:

I don't know too much about this but the benchmarks seem slightly faster with local memory mode, however this must be a small tree since it completes in ~2.2s.

The benchmark runs 6 iterations (6 for both players) on a simple 100bb tree with one bet size.

Quote:

Punter: do you have a recommendation on NUMA vs UMA memory access modes (just in general--I know you don't personally use a Threadripper)?

No recommendations. I don't know anything about it sadly.