|
Post by krcmdc on Jul 3, 2019 14:27:06 GMT
Ok, I'm pretty stuck now. I hate the idea of posting daily results, as they are so meaningless. But, posting daily results serves another purpose also. It forces me to play, because I have to have something to post. I don't like the idea that I'm held accountable by the internet, but I've also learned quite well over the past few weeks, that holding myself accountable to myself, doesn't accomplish anything more than wishful thinking. So, starting tomorrow, I'm going to back to posting daily. And I mean posting everything again.
I won't post all the different Snowie screenshots, but I will post the main page. I will play 1000 hands with Snowie everyday. And at least 3 times per week, I will import Ignition/Bovada hands and check my progress with trying to play more in accordance with what Snowie recommends.
I will play at minimum 1300 hands at Ignition or Bovada everyday, and it should only be that low when I'm playing only regular tables. I've been getting less hands per hour recently, as they've done a few more updates to the software, and it seems to actually be a bit slower now. I get about 55 hands per hour per regular table, and about 175 hands per hour per zone table. I should average playing about 6 hours per day, and I should never play less than 5 hours per day. I'll allow myself "time off" so long as the average remains at 6 or more. DriveHUD does a terrible job of calculating "time played" because it doesn't stop the clock when the hands stop coming, unless they stop coming for more than 30 minutes. So if I take a 20m break, DriveHUD counts that as time played. With this in mind, I'll just use the hands per hour I listed here, and get to a satisfactory total by whatever combination of those numbers it takes. If I play 350 hands of zone, then that counts for 1 hour of play, regardless if it actually took 1.5 hours or 0.5 hours, it just counts as 1. Same for regular tables, playing 660 hands is 3 hours of play, if the hands came faster, lucky me, if they came slower, unlucky me. My hands per hour estimates are on the lowish side anyway, so this shouldn't really be any sort of problem. DriveHUD screenshots will be posted daily.
I will also work on my ranges for at least 30m per day, and hopefully 60m, but 30m at a minimum.
And I will spend at least 30m per day, again, hopefully 60m, but at least 30m working on GTO+ stuff. This could be several things, 1) actually inputting information into a spreadsheet, as would be the case when a sim has completed, 2) it could just be looking over the output information and noting the ranges from it, which can then be used when I'm studying ranges in the future, 3) prepping GTO+ for future sims, which would be the case after an entire group of sims has been completed as well as possible other times, 4) whatever else might need done. Of course this time doesn't include time spent actually running the sims, as that takes a decent amount of time and will be done while I sleep.
I will spend at least 1 hour per day doing something with videos, written content, or even twitch. I have several training videos that I haven't watched, so that's a good place to start. I found software that allows me to download twitch vods, so I can get some of those also. Although most of the best ones are unavailable without subscribing, so at some point, I'll probably start subscribing to one channel each month just to download all the vods, and then sub to a different channel the following month, rinse and repeat. I will also search for more videos and such regularly. Searching for videos, written content, twitch content, etc is something I can do while I also do other things, like crypto stuff, or chatting with my kids, or whatever, so there is really no reason to ever miss a day of this.
So that's it, not sure how I could ever prove the last 3 things, but I'll certainly include something about what I did in my posts daily. The first 2 things will be done and proven with daily screenshots, same as I did a few months ago. I don't know why I struggle so hard with this stuff, but I do. Once I get going, it's pretty easy for me maintain it, but getting going is a huge problem. The motivational speaker idiots often talk about this sort of thing with words like "momentum", and while it's certainly not the correct word by definition, their intent seems quite prophetic in my case.
|
|
|
Post by krcmdc on Jul 7, 2019 12:07:02 GMT
Well, obviously that plan hasn't worked out so far. I got off to a good start on the first day, well, not really, my results were abysmal, but I was doing everything that I said I would do, until I started having internet issues. These issues, combined with my foul mood caused by the abysmal results, led to an early stoppage for the day. The next day my internet problems were improved, but not actually fixed, and this combined with Ignition/Bovada being down about half the day, led to not much getting done for at least half the day. And the third day was mostly punted due to laziness. I did play a decent amount, but I didn't do much of anything else. Combine all of these things with the fact that it's 4th of July weekend, and I haven't gotten off to the start I was hoping for, obviously I'm also not doing so well today either. But I have managed to get my sleep schedule turned around to a decent starting time, I hope anyway. So the plan is to do a lot today, without worrying about the plan, and then tomorrow, get back on the plan.
I have been managing to continue with the GTO+ stuff. Including one huge mistake, lol. I always save the sims, and then start a new one with the same settings, and then save the new one with a different name. This eliminates the need to for me to completely "set up" the sim each and every time, as well as guaranteeing that every sim is exactly the same set of circumstances. However, if you forget to rename one, you end up either not saving it, or saving it over the top of a previously saved result. Which is what I did, and to make matters worse, I seem to have saved one with results so obviously poor that I'll never look at it, over the top of what seems to be the very best option so far. I saved a sim where I allowed only a pot sized flop cbet or a flop check, over the top of a sim where I allowed only a 2/3 pot sized flop cbet, or a flop check. And clearly thus far, betting 2/3 or checking is the best option. So I may end up having to re-run that entire sim. But before I start that, I thought that I might as well run one with a 3/4 flop cbet sizing and see if maybe that does just as well, or hopefully even better than the 2/3 sizing. I thought to do this simply because betting 3/4 pot is simple as can be, there's a button for that already built into the software. So if it's the same or better, then I don't have to re-run that sim, lol. If it's better, I'll probably run one more with something like 7/8 flop cbet sizing, trying to locate the point where it stops being better to bet bigger. I know 2/3 is better than 1/2, and I know 2/3 is better than full pot. So if 3/4 is better than 2/3, then there has to be an optimal size somewhere between 3/4 and full pot, and anything bigger than this starts to become -ev. I'll also run another with 1.5x pot sized cbets, just to get an idea of how that works. As it is perfectly possible that a point also exists where extremely large, becomes better than the size I locate between 3/4 and full pot. I don't expect this to be true, but it is possible, so I might as well run the sim and find out for certain while I'm here.
But anyway, I'm getting near the end of the sims that run from the flop. And everything that begins on the turn happens much faster, almost instantly in fact. I can just pick a turn card, click a button, and about 2 seconds later, I have the entire gto solution for that turn card. I can lock in everything from the flop, including the flop cbet strategy that I derive from these current sims, and fly through a bunch of different turn cards to locate the optimal turn strategy somewhat quickly. The river is the same way, except it takes about 1 second, instead of 2. At which point, I will be able to actually write down ranges, exact specific ranges for every possible postflop decision. I will obviously be searching for a way to simplify it, as it would be a lot of writing if I actually have to do every single possible postflop scenario, but I'm hoping that things can be somewhat combined together, depending on the flop texture, and make it a manageable task to write it all out. And if not, then I'll just write a lot.
I've played some with Snowie as well, but not as much as I need to, I have done virtually nothing in terms of video's or other content. And I have done literally nothing in terms of memorizing ranges. Tomorrow will be different, and I'll get back to doing everything I'm suppose to be doing. I'm running out of options, so soon enough laziness won't be an option, and shit will get done.
|
|
|
Post by krcmdc on Jul 8, 2019 9:52:44 GMT
Ok, finished the flop round of sims. Concluded that I shall cbet flops at a 3/4 sizing. This is not what I expected, I expected a smaller sizing, but this sizing outperforms every other sizing that I tried by a good distance. Of course 2/3 was also good, but not as good, and 7/8 was okay, but worse than 2/3. Nothing else was in the same ball park. Full pot lost a ton of ev, all the small sizing lost a ton of ev, so 3/4 pot it is.
Fwiw, 3/4 pot actually outperformed the initial baseline sim where I used 5 different sizes. I was limited to only 5 sizes, so I tried to pick "in between" sizes sometimes, such as choosing 70%, rather than 2/3 or 3/4. So 3/4 pot wasn't an option in the baseline sim, but 70% was an option. Given that 3/4 outperformed, and 2/3 performed equally as well, I can't imagine that 70% would not also perform at least as well as all 5 sizes, if not outperform them. This seems like a contradiction to me. If using only 70% is better than using all 5 sizes, then why didn't it just use 70% and ignore the other options?
Looking at the sim results, there are 3 boards where cbetting 1/3 netted a higher ev than cbetting 3/4. But there was only 1 board where 1/3 outperformed 2/3. Still that means it's possible to net a higher ev by using multiple sizes. Which is what I would expect, but it isn't the result I've gotten. Not sure why this is the case, someday, if I ever get a high end computer, I'll run more of these multiple size sims, trying to choose 5 perfect sizes and see how that goes. But until then, if I'm cbetting, I'll be cbetting 3/4 pot from the BU vs BB in a single raised pot.
Now I just need to figure out a way to memorize which hands I'm cbetting and which I'm not. I haven't really looked, but I assume that GTO+ will be using some sort of mixed strategy where it's betting hands at a certain frequency, but not always. This is almost impossible for me to memorize. So I'll need to pick and choose certain hands that will cbet at a 100% frequency, and certain hands that will cbet at a 0% frequency, in some combination that loses the least amount of EV. Thinking about this now, if my EV loss is greater than the EV difference between the sizings I've already compared, then I might need to do this with other sizings as well, until I'm convinced that they remain inferior. This probably isn't necessary, but it's required for absolute precision, and that's why I'm striving for at the beginning of this, later, I'll adapt, but until I'm sure that it's okay to adapt, I'll do what I can to remain absolutely precise.
Once I figure out a cbetting strategy, then I figure out how to deal with flop raises, and everything that might occur on turns and rivers. Sounds like a lot of shit, but this part should go faster, I think anyway. Sims that begin on the turn can be run in just a few seconds, rather than several minutes, or even hours, as is required when starting on the flop. I hope that I can finish all of the "optimal" strategy (yes, I am aware that it isn't actually optimal, but it's as close as I can get it, for now at least), within the next week or so. I won't need to be doing any of these sims while I sleep, so I can go ahead and get started on the next batch of long sims, which I haven't thought about at all until just now. I suppose the easiest way to do this would be to just start with BU v BB and move on to CO v BB, then HJ v BB, then UTG v BB. Probably not the optimal order, but it's a logical order, and that's at least equally important to me at this point in time.
So then, a week more to optimize my baseline strategy, and then another few days to tweak that strategy based on villain types. That should mean that I have about 10 nights worth of time to run sims on the CO v BB, which is almost enough time to finish, but not quite. But I'm hoping that this means I'll always have things to do from this point forward, and I'll never be stuck waiting for sim results again, like I have been all these previous days. I hadn't considered this before just now, I find it somewhat uplifting and motivational. Knowing that I should be getting real answers to meaningful questions everyday as I move forward. Tiny little things that I can implement into my actual play to become slightly better, so that each new day, my skill set is better than the previous day, even if only by a tiny amount, at least it's finally going in that direction, at least that's what I'm hoping is true, lol.
|
|
|
Post by krcmdc on Jul 8, 2019 16:58:34 GMT
Well, so much for getting off to a flying start today. Just when I was getting ready to start my day of pokering, I lost internet service. I hope this doesn't become a habit, it wasn't the same this time as it was the other day, as this time it was just completely out, whereas last time it kept coming and going. I guess I'll just pick up where I'd be at this part of my day as if nothing had happened. Not sure what else to do, but absolutely sure that I need to be doing something.
|
|
|
Post by krcmdc on Jul 10, 2019 11:38:47 GMT
Finally, a bit of good news on the software front. Not Ignition/Bovada, but rather DriveHUD. It seems that they've finally released an update that will allow us to import hands from Ignition/Bovada directly, and overwrite the existing hand histories that were created by DriveHUD during play, thereby allowing us to see all the hole cards from every player. This combined with the population tendency information that DriveHUD already accumulated should become a quite powerful tool, if I can figure out how to use it correctly, lol.
|
|
|
Post by krcmdc on Jul 18, 2019 6:11:39 GMT
Well, clearly nothing has been working very well in terms of motivating me to get back on track with poker. As such, it occurred to me today that perhaps I should just go back to doing what I was doing previously, when I was actually making progress.
So I went back and looked at all of the images I have saved from back in January, February, and March. And at the beginning, I was clearly dividing things into Mon-Thu and then Fri-Sun. Where I would play 2k hands against Snowie Mon-Thu, and 1k hands against Snowie Fri-Sun. I would also play 1k actual hands Mon-Thu, and 1.5k actual hands Fri-Sun. This resulted in 11k Snowie hands every week, and 7.5k actual hands every week. I'd like more volume than that, but with that said, it's still way more volume than I've been averaging lately, so let's start there and perhaps work our way up to where we think we should be.
At some point in late Feb or early March, I began playing half my Snowie hands only from the BU. I'm not going back to doing that, I'm just going to play all 6 positions through all the hands.
I'll run GTO+ sims while I sleep, and I'll log GTO+ data into a spreadsheet sometime during the day. That's pretty easy, I will also find time a few times each week to do what I can to condense the GTO+ data to make it useful for study purposes. I started this already, but with the way I've started, it's going to be a long, long, long, longer than that, and also quite complicated process. So I really need to find something a bit better.
I won't be posting so many of the Snowie images that I used to post, but I will post the "overview" from Snowie and the "daily results" from DriveHUD each day. I still don't think that this is the most optimal way to do things, but I know it worked, and nothing else has worked, so I'm going back to that. I only stopped because they changed the software and I couldn't play so much anymore. But the software has been improved, and although it's not as fast at it once was, it's fast enough that I can play 1k to 1.5k hands per day without any real issue.
So, here we go again.
|
|
|
Post by krcmdc on Jul 19, 2019 2:10:35 GMT
1 day done, lol...
|
|
|
Post by krcmdc on Jul 20, 2019 2:39:09 GMT
2 days done...
|
|
|
Post by krcmdc on Jul 21, 2019 2:02:31 GMT
3 days...
|
|
|
Post by krcmdc on Jul 22, 2019 2:04:05 GMT
4 days... Week results include a few hands played prior to going back to the old schedule... Results since they changed the software back in April... Results prior to the software change, going back to last July... Something is different, but I'm not sure exactly what it is. I keep fighting the urge to think that something is up with the software. But up until today I was slowly becoming convinced that something was indeed up with the software. Not that the site was directly cheating me or anything, but something was different. I had many thoughts as to what it could be, but could never shake the rigtard feeling in any of them. And now, looking at all these results, and thinking about everything, I'm becoming more and more convinced that the simple truth is that I'm not playing as well as I did before. I considered this before, but my Snowie scores had remained pretty good, so I felt that meant that I was playing at least somewhere close to the same as I had before. But today I realized that my thought process isn't anywhere near as clear as it once was. I struggle to remember simple ranges in preflop situations, which isn't the worst thing in the world, as it is always something close, but not knowing always puts doubt in the thought process, and the time wasted pondering the fact that I don't know, means I have less time to ponder the things that do matter, especially in close situations. Snowie scores don't really reflect this problem, as I have infinite time to think everything through, including looking at ranges if I need to do so. I know for absolute certain that I'm getting way fewer hands per hour playing Snowie than I used to get, down like 40% or so, and I now attribute this to the fact that my thought process is not clear, and I'm spending way more time thinking and looking at ranges. Playing yesterday, I noticed that I developed headaches while playing, nothing extreme, and not really even "painful", just like a constant pressure inside the head. Not sure exactly what causes this, but I hope that it goes away when I find a way to clear up the thought process. Not exactly sure what I need to do in order to clear up the thought process, but I do know that remembering the ranges would be a giant help, so I'm going to focus some time to memorizing all the most basic ranges once again. And then see what happens.
|
|
|
Post by krcmdc on Jul 23, 2019 4:07:32 GMT
5 days...
|
|
|
Post by krcmdc on Jul 23, 2019 12:44:23 GMT
Something is different, but I'm not sure exactly what it is. I keep fighting the urge to think that something is up with the software. But up until today I was slowly becoming convinced that something was indeed up with the software. Ok, I'm going on record saying that something isn't right in these games. I know that ~50k hands isn't a big enough sample to prove anything, and I know how the mathematics works. And with this knowledge, I am still certain that after the software change, someone/something is now somehow cheating in these games. I'm going to keep playing, and keep telling myself that it's just a weird distribution, blah blah blah. But in the future, when the truth comes out, I'll be able to point to this post and say that I told you so.
|
|
|
Post by krcmdc on Jul 24, 2019 0:47:50 GMT
On a completely unrelated note, I'm finding that running these sims is showing me that my idea of the way I thought would be the easiest to implement, probably isn't true. I thought it would likely be best/easiest to just figure out the best one single cbet size for each position v position scenario, and use that one size.
But as I go through these, there are obvious patterns forming, where certain boards just prefer to use smaller cbet sizings with a high frequency, and other boards prefer to use a larger sizing with a low frequency. I'm not anywhere near done running sims, so maybe this will change, but so far, it's holding true.
It might be nice if instead of memorizing every single position v position scenario across a variety of different flop textures, if I could just memorize "in position" and "out of position" ranges, for 2 different cbet sizes/frequencies. It's still a lot, but it's also a lot less.
|
|
|
Post by krcmdc on Jul 25, 2019 7:05:05 GMT
6 days...
|
|
|
Post by krcmdc on Jul 25, 2019 7:06:02 GMT
day 7...
|
|