Cuteraven poker stuff

[CuteRaven]

800px-Luna_Park_Melbourne_scenic_railway

Today the 25€ Hexapros took my bankroll onto a long ride that reached as high as about 4200 €, but in the end dived down to 3412.27€ (featuring lots of ups and downs on the way), which is where my bankroll stands at the moment. The adrenaline junkie in me did enjoy it but I wish the bankroll had stayed a bit higher :D. During the last 50 games or so that I played this evening, I probably lost between 70% and 80% of all my all ins and I was getting mostly bad prize pools.

I did start getting some doubts that I'm a bit underrolled for the 25€ Hexas but I decided that my stop loss for the 25€ Hexas is 3000€ and I won't overthink it beyond that. But eventually I started getting quite tired and a bit tilted so I decided to call it a day. I do also have some room for improvement, especially with knowing the preflop ranges more exactly, so I will probably dedicate a bit of time to that later on. Maybe I should also play shorter sessions in the future. In some ways I'm satisfied with today anyway, the loss is not too bad and losing days like this motivate me to study and come back as a stronger player.

Tomorrow will probably be a break from poker and my next session might be a change from Hexas :p

Here are my rough stats from today and yesterday (I played 13 Hexapros yesterday and 158 of them today).

Total games=171

Wins=62

Winning %=36.25%

It's still a small sample so I can't really make conclusions about it yet. 36.25% is slightly above breakeven so I need to improve that figure if I want to make grinding Hexas worth it. I am optimistic that it can be improved/will improve if I keep playing.

Today I also got my Covid test results back (they are negative!), though I'm not feeling completely healthy yet from this flu.

[CuteRaven]

I ended up doing a bit of "studying" regarding Hexapros today. Mostly so far I have been trying to find out more about the game type itself (achievable winrates and variance), instead of looking at preflop ranges.

I thought I would share some of the results for those who might be interested!

At first I started looking for a jackpot SnG variance calculator website. I couldn't find one so I decided to make my own calculator using Python.

The calculator is not completely accurate for the following reasons:

1. It is a monte carlo simulation

2. It assumes your win percentage is the same for all prize pools. However this is not true because the achievable winrates differ based on the blind levels, and the bigger prize pools have slower blind levels

3. It assumes all prize pools have a winner takes all format, which is not true for some of the rare higher prize pools. This has very little practical effect for the calculations

Anyway, here are some results:

The following results are for a player who wins 38% of their Hexapro games and plays 1000 25€ Hexapros. A 38% winrate is probably somewhere close to the highest that is achievable at these sorts of stakes, so this would be a simulation for one of the best players in the world:

Variance calculations for player who wins 38% of 25€ Hexapros and plays 1000 games. Scenario has been simulated 10000 times.

 

Basic stats:

 

Average (mean) simulated run = 1579.8775€

Average (mean) theoretical run = 1546.895€

Average (mean) simulated ROI = 6.31951%

Average (mean) theoretical ROI = 6.18758%

Probability of loss = 16.1%

 

Net profit and ROI:

 

Bottom 2% run = -1562.5€

Bottom 5% run = -962.5€

Bottom 15% run = -87.5€

Bottom 30% run = 687.5€

Median run = 1450.0€

Top 30% run = 2225.0€

Top 15% run = 3025.0€

Top 5% run = 4012.5€

Top 2% run = 4925.0€

 

Bottom 2% simulated ROI = -6.25%

Bottom 5% simulated ROI = -3.85%

Bottom 15% simulated ROI = -0.35%

Bottom 30% simulated ROI = 2.75%

Median simulated ROI = 5.8%

Top 30% simulated ROI = 8.9%

Top 15% simulated ROI = 12.1%

Top 5% simulated ROI = 16.05%

Top 2% simulated ROI = 19.7%

 

So the probability of loss is still 16%, which is quite a lot. Probably the worlds best Hexapro player would still need about -1500€ or -2000€ stop loss in the 25€ Hexapros if they wanted to minimize the risk of moving down in stakes (so 60 to 80 buy ins). So this would be a bankroll of between maybe 4000€ and 5000€. One thing that is interesting is that the median profit is a bit smaller than the mean profit. This is because in Hexapros you win very much rarely, but win a little bit (or lose, depends on win rate) quite often. The variance is huge of course as you can see.

I did the same simulation, but for "normal" 3 handed winner takes all SnG ( on this website https://www.primedope.com/tournament-variance-calculator/). The probability of loss was only 9% and the 70% confidence interval (equivalent to bottom 15% run and top 15% run in my sim) was between 400€ and 2600€, so smaller than the -87.5€ and 3025 € in my sim. The 95% confidence interval for normal SnG was between -500€ and 3600€. The "extreme luck/unluckyness" is where you can really see the difference in variance.

How about the same simulation, but for a slightly winning player that wins 36.5% of the time?

 

Variance calculations for player who wins 36.5% of 25€ Hexapros and plays 1000 games. Scenario has been simulated 10000 times.

 

Basic stats:

 

Average (mean) simulated run = 488.2825€

Average (mean) theoretical run = 498.99125€

Average (mean) simulated ROI = 1.95313%

Average (mean) theoretical ROI = 1.995965%

Probability of loss = 39.87%

 

Net profit and ROI:

 

Bottom 2% run = -2562.5€

Bottom 5% run = -1975.0€

Bottom 15% run = -1112.5€

Bottom 30% run = -412.5€

Median run = 375.0€

Top 30% run = 1175.0€

Top 15% run = 1987.5€

Top 5% run = 3012.5€

Top 2% run = 3787.5€

 

Bottom 2% simulated ROI = -10.25%

Bottom 5% simulated ROI = -7.9%

Bottom 15% simulated ROI = -4.45%

Bottom 30% simulated ROI = -1.65%

Median simulated ROI = 1.5%

Top 30% simulated ROI = 4.7%

Top 15% simulated ROI = 7.95%

Top 5% simulated ROI = 12.05%

Top 2% simulated ROI = 15.15%

 

LMAO. With that median profit and probability of loss, not worth the pain imo :Rofl:

 

Another thing I wanted to figure out was the weighted average of the blind levels (to compare Hexas with other game types). Basically I weighted the blind levels based on how much of your EV comes from prize pools for each blind level type. Here is the calculation:

So the weighted average blind level is 2.14 minutes. For comparison the blind levels in many other jackpot SnG game types is usually 3 minutes, so in comparison Hexapros have a slightly higher luck factor. However the variance in prize pools is not as large as with some other poker sites so this reduces the luck a bit for Hexas. One comparable game type is HU hyper turbo SnG that are found on some poker sites, they also have 25 BB starting stacks and 2 minute blinds.

This website https://upswingpoker.com/heads-up-hyper-turbo-sit-gos-strategy-hu-sng/ says that poker stars has about 2% rake for the HU hyper turbos. I have seen some websites/players on certain forums that claim a 5% ROI is possible in these games.

In 3 handed SnG with 2% rake a winrate of 35.85% will give you an ROI of 5.4% however with the rake of Hexapros (6.853% rake) this will only help you to break even on average! So rake has a huge effect. Does this mean that Hexas are unbeatable? No, because player pool and rakeback can make it more beatable. Hexapros have really soft player pools in general (seems especially true during promotions), and this is most true for lower stakes imo (10€ or 5€ and lower). Also for higher stakes (100€, 50€ and maybe 25€ if you play a lot) the rakeback is quite good so this can make playing there much more profitable. But it certainly doesn't seem easy to make a profit.

Below is the source code for the variance calculator if anyone is interested. With a few modifications it should work for other jackpot SnG types too. It might have bugs, I made it today in 2-3 hours :p. It requires Python and the numpy library to run.

#Hexapro variance calc, needs the numpy library to work#This tool is somewhat innaccurate because it assumes that your winning % is same for all prize pools.#However, this is not true because bigger prize pools have slower blind levels, which affects ROI#For simplicity this calculator ignores the second and third place payouts for prize pools of 25€ and above (in practice the effect of this is very small in Hexapros)#This is meant more as a simple and rough sim to get an idea of Hexapro variance#Edit the variables below for your simulationWinpercent=36.5Buyin=25#Buyin includes rake/feesAmountofgames=100Amountofsims=1000#Keep amount of sims at least 100 for good results#Keep amount of sims multiplied by amount of games at most at 1 million for good resultsStatspercentiles=[0.02,0.05,0.15,0.3,0.5,0.7,0.85,0.95,0.98]#Statspercentiles values: For example 0.05 means stats will display data about bottom 5% run and 0.7 means data about top 70% run is displayed#Don't use fractions of % for statspercentiles#Do not change the values/code belowfrom numpy.random import choicePayoutslist=[[1000,1],[100,5],[25,100],[10,4600],[5,14000],[3,25000],[1.5,56294]]def payoutz(payoutslist):    plist=[]    for a in range(len(payoutslist)):        plist.append(payoutslist[a][0])    return plist    def payoutprobs(payoutslist):    totalprob=0    problist=[]    for a in range(len(payoutslist)):        totalprob+=payoutslist[a][1]      for b in range(len(payoutslist)):        problist.append(1.0*payoutslist[b][1]/totalprob)    return problistdef EVRoi(winpercent,buyin,payouts,problist):    windecimal = winpercent*1.0/100    EVsum=0    for a in range(len(payouts)):        EVsum += payouts[a]*problist[a]    Roi=EVsum*windecimal-1    return Roidef simulator(winpercent,buyin,amountofgames,amountofsims,payouts,problist):    resultslist=[]    windecimal=winpercent*1.0/100    for a in range(amountofsims):        netprofit=0        for b in range(amountofgames):            win=choice([True,False],1,p=[windecimal,1-windecimal])            if win==True:                netprofit += choice(payouts,1,p=problist).item()*buyin - buyin            else:                netprofit -= buyin        resultslist.append(netprofit)    resultslist.sort()    return resultslistdef stats(winpercent,results,amountofgames,roi,buyin,statspercentiles):    statslist=[]    average=1.0*sum(results)/len(results)    theoreticalaverage=roi*buyin*amountofgames    print("Variance calculations for player who wins "+str(winpercent)+"% of "+str(buyin)+"€ Hexapros and plays "+str(amountofgames)+" games. Scenario has been simulated "+str(len(results))+" times.")    print ""    print("Basic stats:")    print ""    print("Average (mean) simulated run = "+str(average)+"€")    print("Average (mean) theoretical run = "+str(theoreticalaverage)+"€")    print("Average (mean) simulated ROI = "+str((100.0*average/amountofgames)/buyin)+"%")     print("Average (mean) theoretical ROI = "+str(roi*100)+"%")    print("Probability of loss = "+str(100.0*(str(results).count("-"))/len(results))+"%")        print ""    print("Net profit and ROI:")    print ""    for a in range(len(statspercentiles)):        statslist.append(results[int(statspercentiles[a]*len(results))-1])    for b in range(len(statspercentiles)):        if statspercentiles[b]<0.5:            print("Bottom "+str(100*statspercentiles[b])[:-2]+"% run = "+str(statslist[b])+"€")        elif statspercentiles[b]>0.5:            print("Top "+str(100*(1-statspercentiles[b]))[:-2]+"% run = "+str(statslist[b])+"€")         elif statspercentiles[b]==0.5:            print("Median run = "+str(statslist[b])+"€")    print ""    for c in range(len(statspercentiles)):        if statspercentiles[c]<0.5:            print("Bottom "+str(100*statspercentiles[c])[:-2]+"% simulated ROI = "+str((100.0*statslist[c]/amountofgames)/buyin)+"%")        elif statspercentiles[c]>0.5:            print("Top "+str(100*(1-statspercentiles[c]))[:-2]+"% simulated ROI = "+str((100.0*statslist[c]/amountofgames)/buyin)+"%")          elif statspercentiles[c]==0.5:            print("Median simulated ROI = "+str((100.0*statslist[c]/amountofgames)/buyin)+"%")      Payouts=payoutz(Payoutslist)Problist=payoutprobs(Payoutslist)Results=simulator(Winpercent,Buyin,Amountofgames,Amountofsims,Payouts,Problist)ROI=EVRoi(Winpercent,Buyin,Payouts,Problist)stats(Winpercent,Results,Amountofgames,ROI,Buyin,Statspercentiles)

 

[Dennis__]

very interesting stuff

yeah, variance is amazing. I have played close to 20.000 SNGs this year (mostly hyper turbo) and I realize how little of a sample even that is. 

[CuteRaven]

Thanks @Dennis__ ! Yeah it's a pretty weird thought that you can play for such a long time and still not quite know your true winrate. I think one of the most difficult things in poker is that it's hard to properly evaluate how well you are playing since variance masks the effectiveness of strategies so well (compared to e.g chess where you can visually see cause/effect more directly)

[GR1ZZL3R]

Glad you're back and playing @CuteRaven a very interesting analysis of variation, though in a format I only play if absolutely necessary (very rarely  🤣)

I tend to watch more videos on overall strategy these days rather than how to play, and what comes out from the likes of Matt Berkey and others is that a "live" player will probably never know there own true variance, never playing enough hands, and possibly 1,000,000 hands online might not even be enough to reflect a true figure. It's still thought provoking though and thanks for the work you've put in. (I still won't be playing Hexanonos though 🤣)

[CuteRaven]

Thanks @GR1ZZL3R !

Yeah all else being equal Hexas have big rake and big variance so not a bad idea to avoid them. Though I can see why they are so popular. They are at least as exiting as tournaments in a fraction of the time :p

I think both overall broader strategy and the the more specific "how to play certain spots" strategy (whats the term, tactics?) are important.

Maybe in the future we will have some artificial intelligence that can evaluate/rate our strategy more objectively? Seems like that techbology is developing fast.

[CuteRaven]

So, quick update. Bankroll is now 3679.77 €.

I played some more 25 € Hexas today. I played 74 of them and won 28, which means my win percentage was 37.83%. Despite the decent winning percentage, I was initially running quite bad in prize pools and I was only able to get back in the green towards the end of the session.

I haven't had much time for poker on the previous days, and the time that I have had, I have used mostly for studying/constructing preflop ranges for Hexapro. I also got a subscription for Icmizer. I am aware that button ranges at certain stack depths like 25 BB involve limping, and I have been trying to construct some sort of Frankenstein button range (with mixing between limping and minraising). I have been trying to find info from all kinds of sources to construct the button range. Some sources that I have used include Icmizer, Jonathan Littles' Youtube videos (showing Pokersnowie's and other preflop ranges), the ranges on this website https://www.pokersites.io/category/strategy/ and some forum posts. It's probably not perfect but I'm still pretty happy with the result :D

Of course in these sorts of games it's probably good not to follow ranges too religiously (but it's nice to have a starting point and feel like I know what I'm doing). For example I need to stop calling preflop all ins with a GTO range, because most players seem to go all in with a tighter range.

Also I might need to stop calling river lead jams :p. Here is an interesting hand:

It is the first hand of a 125€ prize pool Hexapro, so every player has 25 BB (500 chips).

I had JcJd on the button. I decide to minraise it this time, both players call (including a tightish but not ultra tight multitabler reg in the big blind).

Flop comes Js Ah 6h. It checks to me and I bet 40 chips, which is 1/3 pot. Both players call. Maybe my flop sizing should be bigger, more like 1/2 pot or something.

Turn comes 3c. I bet 120 chips or half pot. The small blind folds and the big blind calls.

River comes Qh and the big blind jams 300 chips into a pot of 480. It's the first time I have seen them lead like this on any street. I'm quite tilted at this point thinking that this mf almost certainly hit a flush on the river :Rofl:. I let my clock go almost to zero, and call it off only to see Kh7h.

Idk if I should've found a fold there, feels pretty hard to get away from but maybe just maybe...

 

[Dennis__]

I call the JJ. GTO is a call for sure. But do villains have enough bluffs here? Quite possibly not. However I am not folding without a history.

That being said, if you had a history with that player and knew he never bluffs in that spot you could make an exploitive fold. The problem with that is that you in return become ultra exploitive, although that might not be of any importance.

[P0kerM0nk]

@CuteRaven  Maybe a bit late but read some post back and good to see you got your life back on track 😃
Also wondering what do you do beside poker? I see you used Python to create a 'variance calculator' so are you a programmer of some sort or did you just learn that yourself out of interest? 

[CuteRaven]

This weekend I have been on a fun but tiring university organization event in nature, so I haven't played any poker.

@Dennis__Thanks for the advice! Yeah, I agree that GTO is a call for sure. The thing is that I have been in this spot a couple of times in the past (including HU against you long ago, if I don't remember wrong :p), where someone has lead the river on a flush card and I don't remember ever being surprised by seeing a bluff (it has always been the flush), though I have always pretty much ended up calling. The only time that I have seen something like this as a bluff is when playing against overly aggro/bluffy fish that use a lot of donk bets on various streets. Maybe I just have selective memory. But solid/tightish players seem to just always have it here. It's a hard bluff to make for most players if they are not thinking about it actively in the situation. I had played many (10-20) Hexas vs this player in that session but I hadn't seen them lead before on any street (which makes it a bit more likely that they never bluff this way and only play monsters like this).

"The problem with that is that you in return become ultra exploitive, although that might not be of any importance."

That it very true, this bluff would probably be very profitable vs me (maybe I should do this as a bluff too haha). But on a site like this when I can often change my screen-name and they don't see what I folded I don't think it's much of a practical problem :D

@P0kerM0nkThanks :) ! I'm really glad about it too, anxiety is by far the hardest thing that I have had to overcome so far in life. I'm currently an engineering student (though quite lazy and especially have been now during corona). I know some Python and Scala and would like to learn more programming too, I quite like the creative/problem solving aspects of programming at least. I wouldn't go as far as to say I'm a programmer though. I have learnt some in Uni, some in high school/college (idk what to call it) and some myself in my free time.

[CuteRaven]

Quick update:

Because I was kinda tired this evening, I decided to pass time by listening to some relaxing house music and get my feet wet at 5 mans (played 9 of them) to see what it's like to play at higher stakes (10€,25€,50€) and to figure out if I want to invest any more time to learn them (I still haven't really made up my mind :D). It was pretty fun, the heads up match feels a bit more epic when you have to work for it initially! I still probably prefer HU SnG though. I also played a few 5€ and 1€ Hexas with the promotion that is going on and because of tickets.

I broke even pretty much in the end (BR 3672.84€). I ran bad in the 50€ and ran good in the rest. I also happened to bubble @Dennis__ once and he stacked me once in another SnG.

[Dennis__]

@CuteRaven 

how dare you just bust me like that:happy:

[CuteRaven]

Haha, I was thinking there is some chance I might end up on some sort of highlight reel of either soft players, or bad beats, or both 😉

[Dennis__]

it´s usually both xD

your screenname + avatar gave me a strange dejavu but couldn´t figure out it was you

[CuteRaven]

Bankroll is now 3132.84€. Today was a day when variance decided to really ❤️♥️❤️ me in the 25€ Hexapros! A way was always found in which I could lose a hand. There were a lot of bad river cards. I also started tilting and playing on autopilot and with too much of a fast pace, and I made some bad plays and stupid mistakes. For example, I jammed small blind vs big blind (who had a stack of about 6 BB) with a pretty standard hand. Then I noticed that a big stack on the button had gone all in before me and I lost a little over 20 BB in a pot against A2 of all hands (this was in a 5x prize pool too lol).

I noticed my tilt and forced myself to go slower, but it didn't really help with the results even though I started playing better :p

I played 61 games and won 18 (of which all but 4 were 1.5x buy in games). That makes for a terrible winrate of 29.5% today. Oh well, better luck next time I guess, next time I will probably play HU SnG though, and study some more preflop too.

Btw, I haven't mentioned yet, I quite like the client redesign (on the browser version)! The hand history search is quite nice for example, and I haven't found bugs yet.

[CuteRaven]

Bankroll is now 3205.43€. Since the last post I have played the following:

42 games of 50€ HU SnG with 22 wins (profit 34€)

20 games of 25€ Hexapro with 7 wins (loss 162.50€)

42 games of 25€ HU SnG with 24 wins (profit 114€)

8 games of 5 man SnG with 3 wins and 1 2nd place finish (profit 77.09€)

It's funny how you can win over 1/3 of the Hexa games you play but still lose 162.5€ lol. I played some 50€ HU yesterday evening. Today I also started off by playing the 50€ HU SnG. The games dried up there after a while. I don't know if it was a coincidence or if it was partially because I was playing 3 tables there, but I guess it's a good sign that my opponents might not be super enthusiastic about playing me (even though I kinda prefer getting action). The 50€ tables seemed to be pretty reg heavy at the time I was playing at least.

Once the action there dried up I made the mistake of playing a few Hexas 🤣. After the loss there I moved down to 25€ HU and was able to make my losses back. I also played a couple of 5 man SnG games at the end and those went well today.

Here are the two most fun hands of the day (with a similar theme):

During my 5 man session one of my friends called me and asked if I want to meet. We ended up going to a bar 🤣. I'm back home now after having a beer. Probably soon the bars will close again because of Covid...

 

[Dennis__]

nice

what 5man stakes did you play? Didn´t see you at the tables

[CuteRaven]

@Dennis__Just 25€ this time, I kinda like keeping it to one stake if there's the player pool for it. Edit: Played around 21:30 Finland time

[Dennis__]

@CuteRaven well before my grind then

[CuteRaven]

Bankroll is now 3436.30€.

I was a bit tired today. Somehow my body doesn't like even 1 beer and a few hours of delaying bed time, and it decided to up the anxiety for the night, so I didn't sleep very well. I decided to play some cash games for a change, earlier today (mid afternoon). I figured that if I played cash when the player pool was so small, the tables would be more short handed and 4 tables would not feel like too few. The stakes that had players playing at the time were 100€ NL and 25€ NL. I chose the 25€ NL because of my fatigue.

I ran quite well and I thought I also played reasonably well. In the end I made a profit of 155.86€ after 4 tabling for 90 minutes.

This evening I decided to 4 table some 25€ 5 man SnGs for about 90 mins. I thought I played well enough for the most part, though perhaps a bit overaggressive 🤣. The aggression was working for the most part though (with the exception of 1 or 2 punts that ended up being called by the other players). Running well in all ins helped!

In the end these were my results:

16 played, 5 wins, 2 second place finishes (profit 75.01€)

It's funny that I played the same amount of tables of cash and SnG - for the same amount of time and same "buy in" - and I ran about as well in both (or maybe slightly better in the cash). But I made double the profit in cash games (and the variance in 25€ SnG is much bigger than in 25€ cash games). I can kind of see why the 5 or 6 man SnG format is dying in online poker. It's not as exiting for recs as the Hexapros/spins, and it's not as profitable for pros as cash games (or maybe I'm wrong?). Still, a shame that it's losing it's popularity like this.

 

[Dennis__]

nice you ran good. I saw you at the tables once. I was playing under the alias "Schwyzer", so you wouldn´t have recognized me. 

I don´t know if cashgame is more profitable than SNGs, it´s certainly more available though

[CuteRaven]

@Dennis__  Ok :D. I think I remeber the screenname now that you mention it. 2 seats to my left or something? I don't think we played many hands vs each other.

Yeah I guess it's a bit difficult to compare cash and 5/6 man SnG (and also to find the most comparable stake level, maybe comparable stake levels would be the ones with the same bankroll requirements?), but my impression so far is that cash tends to be more profitable per table played at comparable stake levels, though either can of course be quite profitable if you are good at the format and are playing high enough volume/stakes.

Since poker is a zero sum game I guess that would mean that the worse players also lose more in cash games (if my opinion happens to be "correct" by some metric).

 

[Dennis__]

@CuteRaven one to your left I believe and we didn´t really clash

it is indeed difficult to compare stakes. Nl2 doesn´t feel the same as 1€ SNGs for example, but maybe it is comparable? I don´t know.

 

[CuteRaven]

@Dennis__ Yeah I guess my way of comparing doesn't make that much sense anyway. 1 more SnG win and I make double the profit. 1 less win and I break even.

Would be interesting to know if any data on some sort of winrate divided by variance has been published on either cash games or SnG. I guess that could be roughly figured out by variance calculators and researching the highest attainable winrates (in bb/100 or ROI) in different game formats.

[CuteRaven]

Bankroll is now 3407.70€. Played some 100€ NL cash and 50€ HU, I broke even in both. Really not happy with how I played today. I had some random anxiety in the morning again and I slept badly. I started tilting especially in 50€ HU, I made some calls that I knew were bad and that's when I knew that I should stop playing for the evening.

I probably should know not to play when I'm this tired. Would be so much ❤️♥️❤️ easier if my brain didn't decide to get ❤️♥️❤️ anxious for no reason when I'm trying to sleep in the morning. Just makes me angry that the anxiety is not under control yet.