I play mostly low stakes mtt and some medium stakes mtt. I also use holdem manager 2. I read in threads that bb/100 is how you can see if you are a winning player or losing player. Can someone tell me specifically what am i suppose to look at for this? I read that if your bb/100 is 10 or more, its crushing. If its like 1-3bb, small winner. A bit above that winner. And a bit higher than that solid winner. And of course anything that is under 0 is a losing player.



In holdem manager 2, in tournaments, i have a tab that says stack size.


It shows bb/100 at different bb stacks such as


101-150bb
81-100bb
61-80bb
41-60bb
26-40bb
16-25bb
11-15bb
0-10bb


It shows the bb/100 for each level. Then at the bottom, like an overall one. Are you specifically suppose to look at what your bb/100 is specifically at stacks of 40bb and below since that is where most tournament stack sizes are later on? Obviously the bigger the stack means your deep stack play etc.



There is also option of filter in tournament stack size and it shows your bb/100?


For that it shows


Greater than 200bb
very large 61-200bb
large 36-60bb
medium 26-35bb
small 16-25bb
very small 11-15bb
short 6-10bb
micro 1-5bb


So basically both of these are the same right? However, i do not see anything that says bb/100 adjusted that i read in threads.

Some people mention look at pre ante vs antes. All tournaments have antes nowadays so isn't that useless?



I heard ppl mention filter it to bb greater than 200... which i assume they mean bb at 100/200 or higher. But how does one do this on hem2? But is this is another way to see if someone is a winning or losing player in mtt? I know that in sngs, the holdem manager EV graph would show how much you should be up or down in sngs. As there is that green line and red line. I know that they have that in cash games as well. But for mtt, how does one do this?



So if someone was to post their hem2 stats with the bb/100, someone can pretty much look at the bb/100 or adjusted bb/100 and can make a good guess if that player is a winning player or not?

You need to win chips in order to win tournaments, so yes if you have a positive bb/100 (enough to cover rake) then you are more likely to be a winning player. In a vacuum that doesn't mean you necessarily are since you might be throwing away money in high value spots etc.

Generally if you don't spew at low blind levels you could filter for <40bb or <30bb to get a better look at how you play the middle stages of a tournament. 150bb all ins in the early levels kind of skew those stats quickly.

Hi yes i know you need to win chips.


Yea i get 150bb all in early level skews everything. Thus it skews the overall bb/100 right? But do you look at that bb/100 or adjusted bb/100. I cant find the adjusted stat on hem2.


So do i look at bb/100 or adjusted bb/100 at specific bb stacks? Like 1-10bb... 11bb-15bb, 16bb-25bb, 26bb-40bb etc? See what is my specific bb/100 at each one? Is that how you check if you are a winning or losing player on hem2?


People say filter it to BB greater than 200. Others say make it ante and not preante, but all tournaments now start with ante so isn't this incorrect?


I want to know how you specifically filter these stats in hem2.

It's all arbitrary anyways. There is no set of filters that tell you "hey you are a winning player". If you select 11-15bb then you get your results for those stack sizes. Big deal.

You can use these filters to check what stack sizes you're doing well at and which you don't. If you see you're leaking chips <10bb then you can work on that. There isn't a magical formula that gives an absolute answer so don't go look for it.

Adjusted is always better since it cancels out the variance from all ins.

Well how about checking the bb/100 with each level such as


101-150bb
81-100bb
61-80bb
41-60bb
26-40bb
16-25bb
11-15bb
0-10bb


So if you have a positive number in each one, would you say that is pretty good at looking at a player's skill level? What is the difference between adjusted vs just the regular bb/100. Adjusted is similar to like the red line in holdem manager 1 where it says how much you should be up or down?


Is it possible to have at least a 6bb/100 in every bb stack 40bb below and be losing? Should the bb/100 and justed bb/100 be a huge difference?

When you're looking at a report (e.g. "Results by stackize" or "position") in HEM, click the STATS button on the top left and then type "Equity" in the search box. Find the "All in EV equity big blinds per 100 hands" stat and add it. You can move the column next to the usual bb/100 one. The EVbb/100 (or "all in adjusted") stat is more useful than bb/100, since it factors in some of the luck element.

Over a small sample size (less than 100,000 hands), it's very possible - due to variance - that you'll have a loss rate at some stack depths. I can't be bothered to open up my HEM, but I remember it saying things like I was winning at 25bb and 10bb, but losing at 15bb. It's just variance. :/

Hi. I just did that... clicked on stats on top left and added the EV bb/100. Is there a reason there are 2 different ones? One is EV bb/100 which i added, other is EV BB/100 which is big bets. So the first one right?



So do i look at this EV bb/100... in tabs like stack size or tournament stack size and not overall correct? So this shows the EV bb/100 for every stack size so look at your EV bb/100 especially at 40bb and below since most tournament stacks are this size?



So when people say their bb/100 is 6 in tournaments, that is their EV bb/100 and not bb/100 on hem2... is that correct? And that is the overall EV bb/100?


But does the overall bb/100 at the bottom matter? That is the average EV bb/100 for all stacks right? But should that be looked at or not?


I just added the EV BB/100 and notice its exactly half the EV bb/100. So ignore the EV BB/100 correct?

Since you play in big blinds, not Big Bets (which are used in fixed limit), the stats you want are bb/100 and EVbb/100, not BB.
I don't think the stacksize report is particularly useful until you have a massive sample size. The report I look at most is POSITION, and I want to make sure that my EVbb/100 is reasonable in all positions (usually losing quite badly in the blinds) and that OVERALL EVbb/100 and bb/100 are positive.

After all, you make money by having a positive winrate in bb/100 overall. How well you do with a 15bb stack in the cutoff position specifically isn't very important.

In the snippet of the POSITION report I've posted below, the number circled in green is the biggest indicator of whether the player is a theoretical winner or loser.

Hi there. Yes i heard about filtering it to position.


But in the image you posted, does this completely disregard stack sizes? Such as under 10bb, 11-15bb, 16-25bb and onwards? Or is that for a specific stack size?


If so, then why do people talk about their bb/100 at certain stack sizes like under 10bb or 11-15bb etc?


So in this example, you are running just a tiny bit worst right because the bb/100 is just a bit lower than the EV bb/100? So what you posted, is that what everyone means when they say their bb/100 is either 5 or 10 or whatever number it is? So they filter it to position? Or some do it but others don't?


So its possible to have something ridiculous like your bb/100 at say -3bb/100 and your EV bb/100 to be something like 12 EV bb/100? Thus if someone is running way way below expectation? Or is that way too big of a range. So if someone is playing solid but run very bad, i guess something normal would be like your bb/100 is a bit over 2 , but their EV bb/100 could be much more than 10?

Sample size dictates how likely it is that bb/100 and EV bb/100 diverge significantly.

Keep in mind that EV bb/100 is not the whole story of luck.

If every time you get KK someone at the table has AA then you're very unlucky. That won't reflect in EV bb/100. Neither will not getting AA as often as expected. Etc.

Well if you get KK and they get AA, you should win about 18% of the time. But say you only win 12% of the time in a huge sample. Let say you have years worth of hands on your computer. Also say in overpair situations, you only win about 73% of the time vs 82%. And say in flips, you only win 43% of the time in a huge sample. So based on that, your EV bb/100 is going to be much higher than your bb/100 because of how much you run below EV right?


Well i don't consider running KK into AA very unlucky. That sucks. But if it only wins like 12% of the time vs when it should be 18% and your overpairs only win 73% over a huge sample, then that would be all reflected in EV bb/100 right?

You don't consider running KK into AA being unlucky?

Why?

In my image, it's just my overall numbers by position. If I used one of the "Stacksize" reports, I'd get different numbers for each stackdepth, but the sample size wouldn't be big enough to draw any firm conclusions. Even over 100,000 hands, there's a lot of variance.
Maybe they are looking for potential leaks (if they have a significant sample size). More commonly, they just want to brag about "crushing" with a short-stack, or they want to moan about how badly they run in flips.
Running more than 2bb/100 below EV in 125k hands is kind of significant, since it meant I was full 27 buy-ins (100bb) below EV. Many players only have an EV of 2bb/100, so if they run more than 2bb/100 below EV they actually lose money.
The numbers in the orange section at the bottom are what people use for their overall winrate. In the sample I played, I was basically a (thereotical) 5bb/100 winner. Some people will filter for position. Some will pore over the stacksize report, or run filters to find out how they do in 3-bet pots or whatever. Some just look at their graphs. As I previously said, the overall EVbb/100 is the main indicator of whether you're a winning player.
In a very small sample, it's possible to have a very large numerical difference between bb/100 and EVbb/100. Over time, the numbers tend to get closer (due to the law of large numbers) although the difference in number of buy-ins can keep diverging forever.
e.g. Your EV might be 10bb/100 and your actual winrate might be 2bb/100 after 10k hands (so you're running at 20% of EV due to runbad and you'll be a few buyins below EV), but after 100k hands, you might 'stabilize' at an EV of 7bb/100 and an actual winrate of 4bb/100 (running at 57% of EV, but tens of buyins below expectation, just because the runbad has lasted for an extended period).

When you're looking at a HEM report there are some buttons under the stats. One of them looks like a mini bar chart. Clicking that will show you the graph of whatever stats you're looking at. e.g. You can see your BTN graph, or your graph for how well you do with 15bb. On the graph, the orange line will show you your All In EV, so you can immediately tell if you're running above or below expectation in races.

Okay so i looked at the overall EV bb/100 and have few other questions.



1. Let say you run real bad in a session or many sessions in a row. Such as coolers with sets vs straights, higher set vs lower set, aa vs kk, but when you have overpair it does not hold to anything close to 80%.


Also let say you play lot of tables and volume. And say 2 and 3 outers are very standard when you are playing 12+ tables.


So if you keep on getting coolered... but say it keeps happening when all the cards are exposed such as you call all in or shove all in on the river and get called but all cards are exposed. Well your bb/100 is going to be negative since you are losing. But your EV bb/100 will be very similar since you call or shove all in and keep getting coolered right?


However, let say you get it in over and over as a favorite before turn or river and keep getting outered on river mostly but you got it in good as favorite most of the time and keep losing. So that would mean your bb/100 would be huge negative. But your EV bb/100 could be even positive or slight negative? Is that how it works?


I ask this because i had many days when my bb/100 is -5 but my EV bb/100 is like 10 or even more. I read if its 10 or more you are a crusher at the stakes. Is that true? I heard if your EV bb/100 is like 5-6, thats good player. 7-9 ev/bb 100 is like very good. 10 is more is crushing?




2. So basically if you keep getting in as a coinflip but keep losing coinflips, your BB/100 could be very high negative and your EV bb/100 would be close to 0 or so right? So let say you keep getting coolered over and over. You lose almost every single big pot and its coolers. Your bb/100 is obviously going to be a big negative. But the EV bb/100 will pretty much very similar since well you are suppose to lose right?



3. Can someone have a EV bb/100 of 10 or more and play many months or year or more in that sample and be losing? Thus huge sample of hands etc? The other thing is do you include progressive tournaments in this though? For example progressive tournaments, players are more aggressive in bounties. Or are you suppose to not include bounty tournaments when looking at ev bb/100?

How many hands do you have in your database?

100 000 will give you some idea of your win/loss rate.

It doesn't take as many as that to have a reasonable guess that you are a winning player if your win rate is high enough.

At the moment, i dont have 100000 hands.



But that isn't hard because i average at least 2500 hands each session i play. I use to have over 1 million hands or more because i never deleted anything. But recently i deleted all the hands etc. But you do include progressive tournaments in it right? But do you include say some mttsngs as well or you do not include those? Or do you only look at freezeout tournaments when it comes to the EV bb/100 and bb/100? I recall someone mentioned progressive tournaments make your EV bb/100 skewed. So you include that or not?



Well is it possible to have like 500000 hands for example and say at least a EV bb/100 of 5 and be down money? What about a EV bb/100 of 10 and be down money? But in these cases if it happens, will your bb/100 be around the same number or it would be much less?


Example can you have 5 EVbb/100 and say 3bb/100 and be down money over 500k hands? I assume yes.


What about a 10 ev/bb/100 and say a 7bb/100 and be down money over 500k hands? Or it has to look something like 10 ev/bb and something like 3bb/100? Is it possible to have both the EV bb/100 and bb/100 to be like around a 10 or more and be down over 500k hands or more?

It's possible to never win an all-in over millions of hands.

It's not impossible, but it's very unlikely. Indeed, (in cash games at lease) there's only about a 2% chance of losing money over 100,000 hands, if your EV is 5bb/100.
Tourneys are weird though, because all the money is up top. So you can be a frequent mincasher with a reasonably high EVbb/100 and bb/100, but still lose money if you don't go deep and get a big score sometimes.
As with all stats, bb/100 tells a more accurate story about your play as the sample size gets bigger. No one truly knows their winrate, because it's constantly changing. And tourney ROI is even more subject to change, due to the mindbending amounts of variance.
PKOs really complicate things. I'm not exactly sure how HEM/PT track them, but I presume that the bb/100 stats are not impacted by bounties whatsoever.
Naturally, you need to win hands and indeed blinds to make money in bounty games, but your actual monetary gains depend more on winning the "important" flips against players with big bounties on their heads. You'd need to run all kinds of filters to try work out from your stats if you're "good" at PKOs. The EVbb/100 and bb/100 numbers are somewhat useful for comparing yourself to other players (e.g. "I'm winning at 5bb/100, so I'm beating the opposition") but in PKOs, in particular, a positive bb/100 does not necessarily mean you're actually making money.

Hi i understand with cash games its unlikely. But i want to know for mtt.


Well how big of a sample of hands do you need for your EV bb/100 to be good sample? At least 100k hands for mtt? 500k hands? 1 million hands?


The thing is i mentioned last time i deleted my entire database. But the last week or so i been looking at my EV bb/100 and bb/100.


On the sites i play on, i had many days where my EV bb/100 is over 10. The bb/100 is much lower than that such as 3 bb/100. I had days where my EV bb/100 is over 10 but my bb/100 is -5. I seen way too many beats that overpair and dominated hands losing on flop or runner runner is very common.


On another site i play on, the last few days, my EV bb/100 and bb/100 is I'm not joking... its over 20. But im down on this site almost every day. I lost count of how many overpairs i lost and runner runner. So that means on this site if my EV bb/100 and bb/100 is over 20 but im losing these days, that means im running extremely bad right? But how is the bb/100 even 20 then? Shouldn't it be like 5 etc? Basically in these tournaments i build a big stack early on but cannot win flips or get coolered at the middle to late parts of tournaments.


But if the last week or so if i play on 2 sites and say ones EV bb/100 is over 10 almost everyday but the bb/200 is like 3 or even negative... that means im doing fine right? I am literally not joking when my overpairs this week and ak vs ax, im not winning more than 50% in this short sample. I also play lot of tables, thus to get more volume and this literally happens the entire time.


Yea i play progressives so i wanted to know if this affects the ev bb/100 or not. But most tournaments nowadays are this format.

I highly suggest giving up this crusade and just focus on playing well. The answers you want have little meaning in the first place.

Approximately one ****ton.

Don't keep looking at them every day. I've had sessions where my EV was over 150bb/100 but I've lost money. You need (very) LARGE samples to draw conclusions about winrate stats, as people keep saying.

I wouldn't focus that much on bb/100 on MTTs!

The reason is simple: MTTs are soft and so everyone who takes the game somewhat seriously will be a winning player. The effect of variance is huge, though (for large player-pools, I wouldn't draw any conclusions without a 5K-tourney sample)

So focus on learning relevant concepts and in hand analysis instead!

I’m surprised this thread has so much focus on win rates of bb/100 for tournaments when ROI is the more applicable measure of performance IMO.

Anyway, the issue of sample size was raised. That requires an estimate of profit variance. I uses the PrimeDope tournament variance calculator to derive an estimating equation for the standard deviation based on a player’s ROI for this particular tournament:

Number of players: 100
Number of payouts: 18
Buy-in 10 Fee 1

Using the Central Limit Theorem to justify assuming a normal distribution of profit, the required number of tournaments one has to play to be 90% confident he is a winning player ranged from 3400 with a sample ROI of 0% to 1565 with a 100% ROI. An ROI of 35% required a sample of 2300 tournaments. For large tournaments, the required sample size will increase and it will be lower for small tournaments. Similar up and downs will occur for changes in confidence level.

I agree with statman that roi is most important. To see why it’s true think about the difference in ev loss between an early tourney 10bb mistake and an end game 10bb mistake. Good chip ev winrate is then a function of the sum of all decisions which increase your prize pool ev. If a play doesn’t increase your share of the prize pool, then it’s a bad play regardless of whether or not it is +chip ev.