stt variance

nicnicnic

http://img409.imageshack.us/my.php?image=rainkhanxh5wk7.png



http://img340.imageshack.us/my.php?image=displaygraphtypeuserhisbe0.png



thought i was running bad until i came across these guys

SpermManJelly

Jesus thats sick if i've read those figures correctly.


Could you just clarify what the figures on the left and the bottom stand for?

Sperm Man Jelly

RoundTower

are these someone's real figures or a simulation? This isn't really interesting until you say who they are.

SpermManJelly

RoundTower wrote:

are these someone's real figures or a simulation? This isn't really interesting until you say who they are.

I think its pretty interesting,for me especially since I saw sit n go's as the most easily beatable from of poker going,and always thought had the least variance by an unbelieveable amount.Never thought someone could experince those sort of swings in them.

Sperm Man Jelly

Jimmy Hoffa

the last guys is elky supposed pokerstars superstar down 110k from stts madness.dunno about the rest but i would hound them like a doomed fox if i knew who they were

Hitman Actual

nicnicnic wrote:

http://img340.imageshack.us/my.php?image=displaygraphtypeuserhisha1.png


http://img409.imageshack.us/my.php?image=rainkhanxh5wk7.png



http://img340.imageshack.us/my.php?image=displaygraphtypeuserhisbe0.png



thought i was running bad until i came across these guys

Original link, Nic? I'm struggling to believe the swings here.

ArmaniJeanss

lenny_leonard wrote:

Original link, Nic? I'm struggling to believe the swings here.

If you are playing $2K Stts (they have these on Stars ???) then $120K downturn is probably achievable* fairly easily, especially if multitabling.

* achievable is probably too positive a word

Hitman Actual

ArmaniJeanss wrote:

If you are playing $2K Stts (they have these on Stars ???) then $120K downturn is probably achievable* fairly easily, especially if multitabling.

* achievable is probably too positive a word

Ah, 2K STTs. The figures make more sense now.

phantom_lord

I would go insane if I played anywhere near 17,000 stts.

sikes

SpermManJelly wrote:

I think its pretty interesting,for me especially since I saw sit n go's as the most easily beatable from of poker going,and always thought had the least variance by an unbelievable amount. Never thought someone could experince those sort of swings in them.

Sperm Man Jelly

each hand that you play has a different $EV, depending on the stage of the Sit n Go, so variance is certainly larger in them.

I have no idea what the long run for sng's is, but it would be well into the thousands, its so much easier to be fooled by randomness in sngs, I used to think I knew the game very well, turns out I hadn't a bog of the dynamics of them, especially the bubble play which has the most effect on your ROI.

nicnicnic

the graphs are taken from the febuary's low content tread on twoplustwo's STTS forum they are taken from different posts and the tread is a couple of hundred pages long so i aint going back looking for them.

I thought the results were so interesting that i saved them to my pictures and was going to use them when someone said 200 stts was a decent sample size for stts.

Hitman Actual

nicnicnic wrote:

I thought the results were so interesting that i saved them to my pictures and was going to use them when someone said 200 stts was a decent sample size for stts.

I remember HJ doing a 2000 STT challenge way back, but even he ran break-even over one 250 game sample-set, iirc. There may still be posts in his original blog if you can find them; it made for interesting reading at the time.

TheRock

Forgive the essay:rolleyes:
I've played a fair amount of STTs and have have both good varience and bad variance. My game of choice mostly.

A lot of people say there is less variance in STTs (and I used to think so) but I think I now disagree. Yes SNG STTS are beatable, but the are extremely streaky. You undoubtedly encounter more bad beats, purely because by the nature of the game you are allin a lot more. And I think there is definitely more LUCK involved in STTs than cash games (yes you can counteract the luck factor with correct strategy).

The vast majority of STT come down to all the chips going in on a 60/40. Now if you can manage to be the 60% more than the 40% you are going to make money. However again because of the nature of the game you a lot of the time have to push allin knowing your the 40% and are happy to take it. A ROI of 10-15% is great in mid stakes STTs. And with Rakeback and bonus whoring, I dont think you need be much better than a break even player to make money multitabling SNG's (nicnicnic knows vastly more about this then me)

Now if your a good STT player by the time you get to the bubble or ITM, you would hope to have a decent chiplead, so that you are not actually the one being allin at showdown.

What I am starting to wonder though is, because of the greater amount of luck, showdowns and Allins in STT's, is it more likely that bad variance will get you in the end?? And is it easier to control the variance in cash games, where you never have to be Allin unless you want to??(unlikely I know, but you get my point I hope)

Anyone have any thought on this?

Hitman Actual

TheRock wrote:

Forgive the essay:rolleyes:
I've played a fair amount of STTs and have have both good varience and bad variance. My game of choice mostly.

A lot of people say there is less variance in STTs (and I used to think so) but I think I now disagree. Yes SNG STTS are beatable, but the are extremely streaky. You undoubtedly encounter more bad beats, purely because by the nature of the game you are allin a lot more. And I think there is definitely more LUCK involved in STTs than cash games (yes you can counteract the luck factor with correct strategy).

The vast majority of STT come down to all the chips going in on a 60/40. Now if you can manage to be the 60% more than the 40% you are going to make money. However again because of the nature of the game you a lot of the time have to push allin knowing your the 40% and are happy to take it. A ROI of 10-15% is great in mid stakes STTs. And with Rakeback and bonus whoring, I dont think you need be much better than a break even player to make money multitabling SNG's (nicnicnic knows vastly more about this then me)

Now if your a good STT player by the time you get to the bubble or ITM, you would hope to have a decent chiplead, so that you are not actually the one being allin at showdown.

What I am starting to wonder though is, because of the greater amount of luck, showdowns and Allins in STT's, is it more likely that bad variance will get you in the end?? And is it easier to control the variance in cash games, where you never have to be Allin unless you want to??(unlikely I know, but you get my point I hope)

Anyone have any thought on this?

Are you playing the 6-seaters the same as nicnicnic? And what's your approach on the bubble? Do you ever use ICM as a guide?

TheRock

Yes, was mostly 6 seaters, but playing about with 9/10 seaters now again.
And yes I have used ICM.

My question is not how to play these, but rather do people think you can control the impact of variance more in cash than STT.

dvdfan

Well its alot easier not to blow your bankroll if you play badly or tilt for example your playing cash and youve your stack built your stack from the $50 buyin to $200 in a couple of hours and in the space of 20 mins you have a run of bad luck and loose 2 buyins with AA twice cracked. You then tilt and lose the other $50. This is assuming your playing 25c/50c max $50 buyin.

Id say a comparitive level bankroll wise with that on sngs would be $20 sng's. If you build your bankroll up +$150 after a couple of hours playing sngs and hit bad variance, it would take alot of tilting to lose that $150. So you would have to lose 7 sngs in a row to lose this. Giving that some sngs can take up to an hour (obviously less if your tilting and getting called with 46o all in) then you should recognise your tilting before you ruin your profit. The problem i find with cash games is its too easy to donk off a large portion of your bankroll in a short period of time.

SpermManJelly

If you are multitabling 2000 dollar sit and go's even slightly lower you'd presume that elky would not have enough time to actually take into account the person he is playing so I presume he uses theory.
At that level I don't see theory working as well being that the standard of players should be much higher for such a large buyin.
Therefore such variance isnt as shocking.

nicnicnic

the consensus of the bigger players on 2+2 is that the margins are getting smaller and expected roi of greater the 10% in the >100 entry bracket are just not achievable anymore

As for sample size i used to think 2000 but now im more to think that maybe 5000 games are needed to forecast your long-term roi with any confidence


Ive had a 500 game stretches where my roi was 38% my last 1000 games have barely 3% roi(6.5 if u include rakeback) this bad stretch is down to bubbling to many before the streach my win to bubble ratio was about 5:4 now its about 4.5:5 at the small margins stt players operate this is catastrophic to the roi. It probably only works out at two or three bad decisions per 100 games but that's the effect with the margins so small.

I know the problem is the amount of tables im playing at the moment. I'm playing up to seven 50/1 omaha and three stts at the same time (chasing a wsop main event seat in a promo) but my roi is suffering badly its a trade off I'm beating the omaha

SpermManJelly

From what i've seen of all this variance in sit and go's I don't see the ultimate value in multitabling them.
It's a know thing most sit and go players are the most graced players so if you 2 table say you can fully exploit these weakness's whereas when you play extra tables your chances to keep an eye out for the weaker players/ultra tight players decreases.

This results in you playing maybe tighter than you should,mis playing hands and choosing theory over feel even.

Sperm Man Jelly

nicnicnic

SpermManJelly wrote:

From what i've seen of all this variance in sit and go's I don't see the ultimate value in multitabling them.
It's a know thing most sit and go players are the most graced players so if you 2 table say you can fully exploit these weakness's whereas when you play extra tables your chances to keep an eye out for the weaker players/ultra tight players decreases.

This results in you playing maybe tighter than you should,mis playing hands and choosing theory over feel even.

Sperm Man Jelly

I disagree as Sikes says the dynamics of a stt is that every situation has an ev so if your making the right ev decisions it doesnt matter what your opponent is doing. Now I'm not saying that your roi wont be less but its simple maths if you can make a 15% roi two tabling and 10% six tabling; well i know which one I'm choosing

SpermManJelly

I havent played nearly as many sit and go's as you have and with the roi you suggested I would also go for the 6 tabling but with all these figures and players of elky's ability showing such sick variance in the long run I would cut down the amount of tables being played basically due to what seems some of the cons of overly multi tabling

Obviously Sikes is a great stt player but as you've said your playing ev situations which is theory but maybe in the long run playing this theory eventually catchs you as shown from the figures above.Thats why I said cutting down the amount of tables you play and literally watching maybe even typing notes on opponents would be better in the long run.This is merely a suggestion and there is no proof to what Im saying.:rolleyes:

Sperm Man Jelly

sikes

SpermManJelly wrote:

I havent played nearly as many sit and go's as you have and with the roi you suggested I would also go for the 6 tabling but with all these figures and players of elky's ability showing such sick variance in the long run I would cut down the amount of tables being played basically due to what seems some of the cons of overly multi tabling

Variance is inherent to poker and is worse in the short run, not long run. What we can do is get more samples into the time we are analysing, in this case our immediate hourly rate, to reduce its effect on that figure.

For example.
You have the choice between 1 $300 table with a 10% ROI
and 6 $50 table with a 10% ROI, we should go for the 6 table because variance will have less impact on our immediate hourly rate.

SpermManJelly wrote:

as you've said your playing ev situations which is theory but maybe in the long run playing this theory eventually catchs you as shown from the figures above.

Playing the correct strategy is not going to cause this variance. The reason SNGs are higher variance is that every hand played does not represent the same value in $ terms, unlike cash games. If you get AA early on and you win a 16BB pot, this represents very little in the way of $ to you, however if you are 4 handed and you pick up AA obviously winning that 16BB greatly increase our $EV because we are on the bubble and we have picked up blinds which represent a larger amount of chips in play. So the hands have a different $ expectation depending on what the situation is.

Also we have no idea about these graphs, did they move up too fast and not move back etc.

SpermManJelly wrote:

Thats why I said cutting down the amount of tables you play and literally watching maybe even typing notes on opponents would be better in the long run.This is merely a suggestion and there is no proof to what Im saying.:rolleyes:

Sperm Man Jelly

If your decision making improves drastically when you play one table then you should stick to that one table until you become more comfortable and start recognising situations. Players who multitable may take a slight knock to their ROI but their hourly rate will go up by multitabling becuase they get more games in each hour, and they move through their swings faster than someone single tabling.

also im not a great stt player

nicnicnic

SpermManJelly wrote:

I havent played nearly as many sit and go's as you have and with the roi you suggested I would also go for the 6 tabling but with all these figures and players of elky's ability showing such sick variance in the long run I would cut down the amount of tables being played basically due to what seems some of the cons of overly multi tabling

Obviously Sikes is a great stt player but as you've said your playing ev situations which is theory but maybe in the long run playing this theory eventually catchs you as shown from the figures above.Thats why I said cutting down the amount of tables you play and literally watching maybe even typing notes on opponents would be better in the long run.This is merely a suggestion and there is no proof to what Im saying.:rolleyes:

Sperm Man Jelly

I kinda see where your coming from Mr jelly and there must be a point where adding tables becomes -ev alot of the bigger players in these play up to 12 or more tables. now there has to come a point where an extra table reduces the hourly rate. But I'd imagine that with practice a decent player will increase there hourly rate up to eight tables at least.

also when you speak of getting a feel for the other players. this is going to sound like a contradiction but your talking about playing good poker. Now this is only my opinion but i really don't think you need to play good poker to be a winning stt player. What you need is a strategy to beat the game consistently. Because of the dynamics of stts there is basically an optimum play for every instance now the optimum play may not be pretty. For instance say a stt is five handed blinds 150/300 you've got 2600 and k6os the optimum play could be a push ( I'm not saying it is ) now this will not be considered good poker in fact people will probably have you down as a donkey if you walk into a hand but is the optimum play if you get what I mean.

RoundTower

well yeah but only because people are stupid.

nicnicnic

RoundTower wrote:

well yeah but only because people are stupid.

I like your thinking on this:)