CuteRaven

Thankyou! These changes are great!

"We believe that the rake at NL4 is almost too good today, in the sense that it gives little incentive to move up in stakes, and the stake becomes much tougher and not as great a place for new players to start out, as we want."

This is a weird explenation if I understood it correctly. Sound like the classic "more rake is better". I don't think the NL 4 games are at all tough. But anyway I think it is fine to increase the rake there because it used to be so low. 

What is the table cap on other stakes than NL 4?

I played a pretty long session of 4€ SnG today. My results were ok I guess. Competition is quite soft at those stakes and I'm very happy with how I played today. BR is 346,40€ :

"You played 156 games and won 83 of them. You paid 18.72 in rake and your ROI is 3.22%"

400€ here I come!

Ooh that makes sense. I think you are right, my bad :D. 

What I mean is, without knowing the gametype that is being played (including blind levels and stack sizes), it is completely impossible for any calculator to accurately calculate variance. It is the same difference as trying to calculate the variance for pot limit omaha cash games with a standard deviation that you use for calculating variance of texas holdem cash games.

Or maybe I am wrong somehow but in that case I don't understand what I am missing :)

Yeah @psrquack, you could definately be right, my nunbers aren't really based on any real data beyond my own intuition. 4 months of breakeven sounds pretty brutal :o

@FreedoM Thanks for the link :). I like the answer, I do prefer more aggressive bankroll management straregy nowadays. Of course winrate and tolerance for risk affects how big a bankroll should be also, so I guess finding a fixed number on the internet isn't so useful. But I think it's an interesting way to (try to) compare variance of HU SnG to a gametype that many are familiar with.

"Starting stacks and Blind levels don't influence the variance if you use your estimated ROI for the calculation. "

I'm not sure why this would be the case? To quote Wikipedia: "Informally, variance measures how far a set of (random) numbers are spread out from their average value." As far as I have understood it, your ROI has nothing to do with the variance, because the variance is a measure of how far your results are spread from your (real) ROI. So variance is a property of the gametype and can't be calculated from your ROI (but the amount that you are likely going to win or lose can be calculated from ROI and variance figures).

I think what primedope is doing is that it uses some variance number that it thinks a "normal" tournament on some site would have. But it can't differentiate between Hypers, Turbos and deepstack tournaments in terms of variance (and these gametypes do have different amounts of variance).

I haven't updated for a while, and I have not been playing as much recently, but I have played a bit. I had a really bad session earlier where I probably played between 20 and 30 SnG and I could probably have literally counted the amount of all ins that I won with the fingers of my hands (I lost about 130€ that session). Then I went down in stakes to 4€ heads up SnG and lost some more. Ironically I felt like I had been playing better than ever during those sessions but the variance was just not on my side.  After a low of about 250€, I managed to turn it around a bit and my bankroll is now at 326,32€.

I have strated to realize that I have been underestimating the variance in Unibets HU SnG games. If I compare the "typical swings" of HU SnG to the "typical swings" in cash games, I would guess the variance of 10€ HU SnG games are about equivalent to NL 25€ 6-max cash games (or maybe even slightly more?). Maybe some other players who have been tracking their results more closely (or Unibet itself) would have more accurate data on this. If you think about it intuitively, HU SnG is not super far from HU Banzai, except in HU SnG the starting stacks are slightly deeper and the losses are capped to the buyin amount.

From some of the articles that you see floating around online, a bankroll of 20 buyins for cash games is considered "very aggressive", 30 buyins is "aggressive", 50 buy ins is "medium" (this is the standard in someones fantasy world at least) and a "cautious" bankroll management involves 100 buy ins. Converting these recommendations to HU SnG on Unibet assuming the variance of 10€ SnG equals the variance of NL 25:

- 50 buy ins would be a "very aggressive" BR

- 75 buy ins would be "aggressive"

- 125 buy ins would be "medium"

- 250 buy ins would be "cautious"

I remember seeing a Youtube video where some cash reg (maybe Ryan Fee) claimed that the biggest downswing he had experienced in his career was about 50 buy ins. So maybe people who grind HU SnG on Unibet for many years can expect a 100+ buy in downswing eventually (and a upswing of similar caliber). Of course my estimates could be way off too. I probably don't have the patience to wait until 75 buy ins to move up in stakes, but maybe 60 buy ins with a stop loss of 50 buy ins will be my go to from now on.

I might also play more of the other game types. As much as I like the volume that you can get playing HU SnG and the large role of exploitative play that HU SnG involves, the hands played there just aren't as intertesting (hence the lack of discussing hands in recent posts) because of the short stacks. HU SnG are also "more solved" than many other formats like cash games, which makes it easier for the regs to be quite good (preflop ranges play such a big role, even though they are not the only important thing of course).

What are you using as a simulator? I have started believeing more and more that the variance in Unibets HU SnG are severely underestimated by some of these online tournament variance calculators, because they do not take into account strating stacks and blind levels, which make a huge difference if you think about it.

Still happening regularly everyday. Lemme know what extra info, if any, is needed. Pretty annyoing bug to face as I almost time out my hands (sometimes I do time out) and the tables change places which is disorientating.

So I have played a bit and results have been quite breakevenish. Nothing really that interesting, though I feel like I have been running a slightly bad in all ins at least so considering that I'm pretty happy with the results. BR is now 427.64€. These are probably my stats:

You played 82 games and won 43 of them. You paid 24.6 in rake and your ROI is 1.73%

I recently tried to learn better preflop strats, since preflop is undeniably such an important aspect in HU SnG. I have come to realize over time that my simplified start at around 15 to 25 bb effective stacks of limping colse to 100% was not very good. Sure, it made prices for 3bets worse for my opponents, but it also let my opponent in the pots for free. All the better regs had some minraises (and reshoves) at those stack sizes and intuitively it made sense too. I looked at the charts here to get an intuition for what sorts of hands I want to put in different parts of my ranges in general https://www.pokersites.io/heads-up-hyper-pre-flop-charts/. I was playing only 2 tables for a short while today so that I could learn to apply the new preflop ranges while playing at a relaxing pace (but then that got boring so I started a third table).

I also had a personal project to make a calculator for Sökö in Python. Sökö is a variant of 5 card stud. But it is much better than 5 card stud because it contains 4 card straights and 4 card flushes in the hand rankings (so it's actually feasible to chase draws, and what is a good poker variant without the existance of decent draws). Also unlike most stud games it is pot limit and not fixed limit (fixed limit is super boring :D). I pretty much completed the project today, and it has been interesting to learn about staring hand strategy (also now I have a calculator that I can adapt to other game types pretty quickly too if I want).

I think I am quite good at figuring out strategy games intuitively, but starting hands are the hardest part to figure out intuitively because the effects of starting hand selection are the the most indirect to the final results of the hands you play (there are still many betting streets to go and many brances of the game tree to take). Sökö is a game that can be played on only one Finnish poker site as far as I'm aware and there are little to no resources on the game. It's a fun challenge to figure out games that have not been "quite so solved" and to try games that are really different to NLHE.

I probably will be spending less time playing poker for the next few days. I want to try and get other stuff sorted (including my terrible sleep schedule).

I have also played a lot against LieutenantHU and NR23 (especially NR23) at the 10€ level. I know their games better than almost anyone elses that I play against :D.

I find sleep, excersize, some nice relaxing social events in good company, and nature, in no particular order seem to help most my mental state reset. Enough excersize seems to be crucial for my mental health in general (and how well I think and by extension play poker), so finding some form of excersize that you genuinely enjoy would probably be a good thing. I have also noticed that it is easy to play too much poker, burn yourself out a bit and fukk up your sleep rhythm (and I'm not even a pro so there is no pressure for me). I'm the type of person that tends to focus way too much on the things that interest me and I can forget other things in life, so it's important for me to be aware of that weakness (I haven't figured out how to fix that weakness yet though). Idk if you are similar.

Thanks @Dennis_ ! If there was one good thing about the session, it didn't leave any doubt in my mind that today was up to much else than bad variance. The way I was running, the money was going to go in any case :D. And when playing against decent/good aggro players and they are running great, it is very hard to counter/adjust to them. It's easy to play too trappy/passive, or call down too light vs them and bluffing becomes scary if it feels like they always have it. Let's hope that we can all turn these downswings around soon :)

Quite a few times while playing on Firefox today my session timed out ("istunto aikakatkaistu") while I was playing and I had to log in again to continue playing.

A recent freeze a few days ago. Edit: Time was Sat 29 Feb 2020 11:41:46 PM EET

Well, I ran horribly today at 10€ HU SnG. When I bluffed they called, when I value bet they folded. There was one particularly annoying aggressive opponent who managed to somehow constantly have it even though they were almost constantly betting and constantly betting big, very hard to play against, and they won most of the all ins too. Not that I was running well against other players either. I started tilting so I had to stop. One of those days I guess...my BR is at 413,44 € after a recent high of about 550€

Since the last update: "You played 113 games and won 55 of them. You paid 33.9 in rake and your ROI is -5.58%"

Thanks for the advice. 50 buyins sounds like a good plan

I was playing as "ILikeBallzXD" 

So now my bankroll is 476.44 € after playing just a few 10€ SnG games:

"You played 23 games and won 17 of them. You paid 6.9 in rake and your ROI is 43.39%"

Wtf lol. I noticed it really helped to play slower though, to force myself to think of each decision longer when playing 3 tables. Anyway, I should really go to sleep now :P

Some of that ROI was probably from playing me today :D . Did you have a chess themed name (a mutant of two chess legends)? @Dennis_ 

So today was a pretty terrible day pokerwise. I decided to take shots at NL 25€ HU SnG. It did not go well at all. I probably lost 75% of the all ins that I played, especially towards the end of the session. The last game was pretty indicative of the session. I was up in chips and I lost probably 6 or 7 all ins in a row. My bankroll is now 376.64€. Not the end of the world, I can still grind it back up, but pretty tilting.

But I realized a few things about my HU game also when playing against the good regs at that limit:

-I need to study limp/push/fold ranges more. I thought I knew the correct Nash ranges, but I could be playing slightly too loose ones (or maybe the bad session is giving me the wrong impression? Idk)

-I probably bluff a bit too much in general. Estimating the best bluff frequencies against different opponents for different spots while 3 tabling HU is not that easy, but I need to get better at it

-I might have some timing tells, or maybe not. I think my timings tend to depend on if I have big decisions on other tables. So sometimes I act almost instantly, sometimes slower. Maybe I should just slow down in general.

-I need to play better against new ultra agressive opponents. If I don't have reads on an opponent yet, it is quite difficult to avoid getting run over by such opponents, because it's hard to tell if they are just running really well or are really bluffy or make very thin and very aggressive value bets (e.g all in with bottom pair). So it's easy to make either too loose calls or too many fold against such players.

-Maybe I need a bigger bankroll for taking shots :D. This result is a pretty normal bad run probably.

I programmed a quick calculator in Python that can calculate different possibilities for how many HU SnG games were played and won, and other stats, based on the profit/loss of the session (and buyin and payouts for the SnG). I'm too lazy to count while playing :P (so this is the easiest solution if I don't want to constantly bother Unibet employees for the stats). Here is the output:

You played 74 games and won 30 of them. You paid 55.5 in rake and your ROI is -21.35%

In the unlikely event that someone actually wants to use the calculator, here is the code (just paste in a python interpreter, change appropriate values and run):

#HU SnG games amount and games won calculator with ROI and rake paid#Calculator only works if rake 0% < rake <50% of the buyinbefore=771.64 #bankroll before sessionafter=376.64 #bankroll after session. Exclude any profits/losses from rakeback and other gametypes and buy in sizes.buyin=25 #buyin for the heads up sit n go in your currencypayout=48.5 #payout if you win the heads up sit n go in your currencypossibilities=5 #checks the first n possibilites profit=after-beforedef calc(profit,buyin,payout,possibilities,winslist,buyinslist,rakelist,roilist):    profit_remainder=profit%buyin    rake= 2*buyin-payout    n=1    while len(winslist)<possibilities and n<10000:        amountW=(n*buyin-profit_remainder)*1.0/ rake        if len(str(amountW%1.0))==3 or str(amountW%1.0).count('9')>=9:            winslist.append(int(round(amountW)))        n+=1    for i in range(len(winslist)):        amountG=(payout*winslist[i]-profit)*1.0/buyin        buyinslist.append(int(round(amountG)))    for j in range(len(buyinslist)):        amountR=(rake*1.0/2)*buyinslist[j]        rakelist.append(float("{0:.2f}".format(amountR)))    for k in range(len(buyinslist)):        amountROI=profit*1.0/(buyinslist[k]*buyin)        roilist.append("{0:.2%}".format(amountROI))def printer(profit,buyin,payout,winslist,buyinslist,rakelist,roilist):    print "After making a profit of "+str(profit)+" in "+str(buyin)+" buyin heads up SnG games with a payout of "+str(payout)+" when won:"    print "Possibilites:"    for i in range(len(buyinslist)):        print("You played "+str(buyinslist[i])+" games and won "+str(winslist[i])+" of them. You paid "+str(rakelist[i])+" in rake and your ROI is "+str(roilist[i]))        winslist=[]buyinslist=[]rakelist=[]roilist=[]calc(profit,buyin,payout,possibilities,winslist,buyinslist,rakelist,roilist) printer(profit,buyin,payout,winslist,buyinslist,rakelist,roilist)    

Shame about the guinea pig :(. They definately are affectionate, and surprisingly intelligent as well. Awesome creatures

How reggy are the 25€ HU SnG games? I can imagine that most of the better pros and endbosses are lurking on the 25€ level and beyond, but they probably don't care too much for the €10 level.

Played some more €10 HU SnG today. Bankroll is now 770.44€. I was playing and especially running well at the start of the session. Towards the end of the session I started losing focus and tilting slightly and my results were breakeven, but probably would have been better if I was more focused (I slept about 4 or 5 hours last night so fatigue started catching up to me). Yet another reminder not to play too tired, emotional control is so much worse when tired and it's easy to play on impulses instead of logic :D. Anyway, can't be dissapointed with the results!

Nice ROI! How many tables do you play? @Dennis_  Yeah, HU is a good way to get a lot of volume, especially in the browser version where you can tile a maximum of 4 tables.

Thanks @Dennis_ ! Lets hope it lasts