More Tables = More Variance

Had a bit of an epiphany just now, however obvious it may seem...

I've recently been running real good, my graph's shot up and I feel like I'm crushing 50NL even without a load of luck. Anyway, with it being Friday, I've played a decent amount today, and at first I won a couple of buy-ins in an hour, and then I got distracted on MSN and stuff for a while, so when it got late, I thought to myself "Right, I better put an hour or so in now!" and I pulled up the lobby and saw the vast amount of fishy looking tables...

So I must have clicked on about 20 different tables and joined the waiting lists, just because there were that many juicy looking tables and I couldn't help myself, and by the time I knew it, I was 16-tabling, which I haven't done in a long time (since I realised that I wasn't ready for it in terms of applying solid exploitative strategy)... My plan had been to just get on the tables, and leave some, and make sure I had nice seats and such on the ones I stayed with, but I thought to myself that I needed to try to build up some rakeback over the weekend so that I get a nice lump on Friday...

The session went pretty bad and I ended up losing 4 buy-ins over the space of about 20 minutes and quit, feeling quite tilted.
In my head I was like, "****! Why do I do this ****?! I know that if I'm going to mass-table properly I need to build up my brain muscle over time like Jared says! God, I'm such a *****! I've just lost all my winnings for the day!"
Obviously this is a very results-oriented form of tilt. The first stack I lost in the session was against some total monkey, I had AK vs. A8s AI pre-flop blind vs. blind, and he sucked out. I didn't mind that, but after I lost a couple more (I won some stacks too, so I actually lost more than 4 stacks in this session), I was really feeling quite agitated, and regretting having returned to the felt for the evening.

I quit after about 20 minutes, and upon reviewing the session I realised that I hadn't made any errors in stacking off. This confused me, because I usually only get annoyed when I make a bad play, and even then, I don't really get as pissed off these days. I then decided that the reason I was pissed off at myself was because I hadn't disciplined myself to play less tables and pick out of the best of the tables that were available to me.

A few hours after the session, about 5 minutes before I'm now writing this, I realised what it was. It's the fact that when I'm playing that many tables, I'm allowing more of my money and bankroll into poker, and the variance of poker will do what it will with what I put on the table. So by playing double the amount I usually restrict myself to, it's almost as though I might as well have moved up a limit and played my usual amount of tables.

I realised that the reason I'm not ready to play this many tables is not due to my ability to make betting decisions properly overly reduced when I mass-table, but due to the fact that I haven't become accustomed to the multiplied power of variance and the increased speed at which it acts when I put 16 stacks down. This has come as a bit of an epiphany to me, so I thought I'd share my thoughts for those that are trying to add more and more tables to their sessions but then struggle, as I have done, to understand what's going wrong and why sessions seem to be so much more intense.

I'm not saying that our betting decisions can be as good as when we're playing less tables, but I think that more tables going wrong is often diagnosed as a struggle to play as good, whereas one's capacity to handle the eccentuated variance is also a big issue here, so if you're in this boat and you're thinking, "What's going wrong here? Am I really playing that bad when I play more tables? God, this is frustrating!" Then maybe consider this. Or maybe it's just plain obvious. Lol.

If you can play 16 tables the same as you can play 1, then variance is reduced, not increased.

Yeah, more tables = less variance.

Yeah I know more hands = less variance, obviously... My point here is that when you play a really high volume of hands in a really short amount of time, the natural swings of poker come at you at [new amount of tables divided by normal amount of tables]x the speed that they usually do, and this can make sessions far more intense...

I didn't mean there was like increased variance in actuality, but the volume of variance that you have to deal with comes at a much higher rate...

No, variance has nothing to do with it. You felt more rushed because you were playing more tables than you are use to. With more of your bankroll out than you are use to. That has nothing to do with variance.

^^^ This.

Jumping straight into 16 tables when you haven't adjusted to it is almost always a bad idea. You said you'd done it before but it sounds like it's been awhile, so I'd argue that it's not like riding a bike. Massively multi-tabling requires a completely different strategy and -- for me, at least -- required a gradual adjustment that came from adding tables slowly. And even then, I still tried to play clever poker way too much. Even though I could handle the volume, I could never truly get myself to play nitty enough for that many tables, and I wound up overplaying stuff. I too took a break from mass tabling and now play no more than 4 at a time, which allows me to actually think about what I'm doing on each one. I miss the rakeback and easy IronMan levels and bonuses, but what's the point when you start losing more than you're gaining?

So yes your winrate can drop substantially and you can lose a bunch of buy-ins in a hurry, but it's not due to variance.

Yeah, hate to burst your bubble here but if you INCREASE the number of tables you are playing, your overall variance actually REDUCES, not increases. As long as you are playing sufficiently with your bankroll and employing a +ROI, your variance is going to go down. IT sounds more like you need to keep your table # down until you can handle that kind of decision-making.

You guys who think more tables mean reduced variance fail to see the whole picture.
More tables mean increased variance.

So you are saying running it twice doesn't reduce variance?

Lol, you guys are right in that obviously increased volume = less variance over results, but... Consider this...

You play 1000 hands of poker... Variance is a big factor, right? 1k isn't a huge sample, right?

Let's say you play these 1000 hands over 1 day, playing an hour at a time over 4 tables... Pretty comfortable poker, right? You have time to make your decisions, accept your results, and have breaks in between...

Now let's say you play these 1000 hands in 20 minutes or however long it may take over 16+ tables... Exactly the same betting decisions are made, hypothetically... It's going to be much more tough on you, right?

What I'm getting to here is that (this may just seem really obvious), building one's self to mass-table involves not just making our betting decisions a lot more automatic, but also building up our psychological and emotional durability steadily, which also takes time and work (adding tables slowly)...

Yes, but that doesn't mean variance has been increased...

Ok, so bad subject title, lol... Unfortunately I can't edit it... My point still stands though. You can't play 100k hands in 1 session. Pretty much any amount of hands you play in a session, your session is really vulnerable and subject to the variance of poker, and playing huge volumes in short amounts of time requires one to be more stone-cold in handling the increased power that variance has on you psychologically...

Fair

I agree completely.

More Tables = Less Variance.


It's like the pain doesn't last as long.

To clarify:-

More tables = same variance
More volume = less variance

What I don't understand is how you think playing more hands per hour is more tough or less comfortable. Most people feel the exact opposite because the variance goes by so fast it makes sucksouts so easy to deal with.

You can never play an amount of hands in a day such that variance doesn't play a big part in your results. Having more time to deal with results psychologically with more time between hands and breaks considerably increases your durability. I think in order to mass table well, one really needs to be considerably more 'numb' in this respect.

What? Are you on crack sir?

... Lol? :P I'm not sure that's a valid counter-argument?

I've been thinking about this for a while now and it is interesting to me.

We might need a mathematician to weigh in. I am NOT one, as you will soon see.

But I think most of us are defining variance as runbad/good or down/upswings. These things are related to variance, but they do not define it.

Now, I am not sure if we should consider winrate when we talk about variance. But if we should, then we are missing an important part of the calculation ITT. At first I wrote a post here supporting the idea that leaving out the decreased winrate induced by MT play we are forgetting the effect it would be having on our swings. But the more I am thinking about it the more I become convinced that winrate is another thing entirely.

But swings might have more to do with playstyle than skill. For example many losing players are steady consistent losers over the long haul.

I think what many are talking about ITT is not decreasing VARIANCE but decreasing variance/expected value. Or perhaps what we want is to make the dispersion of our variance to be more manageable over the long term. If I flip coins... and start flipping 10x more at a time my "runs of X over N" are not going to change one bit.

One problem with high volume play is the benefit of smoothing out the valleys more quickly, is offset by the speed at which you approach the next swing.

And I can't let the idea of winrate completely off the hook. Even if our variance remains the same, where our winrate falls in the distribution matters quite a bit whether or not we are seeing lots of swings.

Finally this thread has proven that high volume play in X time can have the effect of either reducing or increasing a players tilt, and this depends on the player.

Lots of people convinced in this thread that they are right.

I am not quite convinced we are all even talking about the same thing.

In the long run, all good poker players are profitable, right? They may lose money in the short term (next 100 hands, next 1000 hands, next 100000 hands) but eventually they will show a profit.

Therefore, the more hands they can play, the 'faster' the long run is to reach

i multitable sngs and put in a good amount of volume most sessions.

more tables does not = more variance nor does it = less variance.

imo variance is still the same no matter how many tables you play.

for example if you ussually play 4 tables and then decide to play 8
tables then you will see twice as much variance as you normally would,
but its still the same amount of variance cause you are playing twice as many tables.

playing more tables and putting in more volume does not create less
variance. but it does help you overcome variance much better.

so in the long run, if you play more tables/volume you will go through
some big downswings/upswings cause of variance, but it will go much faster and easier then playing fewer tables/volume.

so op is kinda right, playing more tables doesnt = more variance, but you will see a lot more variance though.

but keep putting that volume in and you will go through variance much smoother.

thats my opinion.

IMO losing a few more SnGs than expected when it goes rough is never as bad as when you get hit by a downswing and lose multiple stacks in quick succession in cash games. There is no form of poker as emotionally testing as NLHE 6-max Cash, I don't think. Furthermore, to mass-table this game requires real emotional strength, I think. That's my point really...

My belief is you can make up for Bad varience at more tables, For instance
If you only played 2 tables for 2 hours and you get pocket AA and get your 100 bb in and get sucked out on. Well with just playing 2 tables you might not even get the 100 bb back that night with playing 2 tables, but if your multi tabling 6 tables or more you might get trip 5s or AA again and have them hold to make up for the 100bb you just lost..

Obviously this is the argument which undermines my bad title choice, which stands (more tables = less variance, in actuality, lol). But my point is, if you're playing 16+ tables, you're likely to frequently go through bad times where you lose 3 or 4 BIs in a row, same as you might win that many, and sometimes you'll go through real bad patches where you lose more, same as you'll go through real good patches where you win more... Granted, a lot of the time it'll be a more steady climb. My point is though, even if you have it in you to make good betting decisions and click and think quickly enough to play that many tables at a time, learning to comfortably play this much poker at a time requires more than just knowing your betting decisions inside out. One must also be completely detached from their bankrolls and results emotionally and psychologically.

I think everyone is basically saying the same thing, it is simply that if you're playing more tables, you will simply move through these swings at a faster rate. If you play one SNG a day and only cash once a month, it is different than if you play 30 a day and cash once. The amount of time between hot and cold is the determining factor, which can be bent by our own mental mindset.

Uhm guys, the topic title is correct, more tables = more hands/hr = more variance. If you're playing 200 hands/hr your potential swing in that hour is significantly greater than if you're playing 50 hands/hr.

Of course more tables also means that your results will converge on expectancy more quickly, but that's a separate issue from variance.

Please elaborate for me as it seems that more tables = more volume...great example of RUSH POKER. Wally

Exactly what I was thinking....and statistically over time there will be a convergence towards the mean.

Aha! I knew I was right! Woop... i r wisdom xD

Thank you Dorkus <3

Is it? I beg to differ.

More hands/hr = 80/20's running closer to 80/20 and 60/40's running closer to 60/40 = less variance

SO, if I play more tables I will get sucked out on:

Less often - or more often?
More times per hour - or less times per hour?
Less times in a row - or more times in a row?


I have a feeling that we are arguing with more than one conception of what variance actually is.

This is precisely what I've been thinking.

When I first read this thread, my knee-jerk reaction was to ask how anyone could think more tables == more variance, when everything I've learned up to this point said the opposite and I thought it was commonly held as fact.

But then I started thinking it through.

If you define variance by swings, and accept that you're going to have X number of swings over Y number of hands (grossly simplified, I know, but bear with me), then if you play Y*16 number of hands within the same period of time, you're going to have 16x more swings within that time, i.e. more variance.

OTOH, as TPC points out, the higher frequency of variance gets you to the "long term" faster, where we all agree the odds converge towards expectation, therefore you get through the negative effects of variance in a shorter time. In a sense, the perceived effect of variance is reduced or shortened.

So I can see it argued from both directions, I guess. Or am I totally off base here, lol.