This is something I usually don't do, but it is weighing down on me some so I'm opening myself up here a little bit to my readers. My 10 tabling hasn't been going so well and after being almost even on the month I had my worst losing session ever yesterday. I lost in the region of 15 buy-ins playing in fairly soft 2/4 and 3/6 games over 5000 hands. This is a lot playing ring games that I would not consider to be tough, ie obvious soft spots at the tables. I've been messing around with ev++'s variance simulator and assuming they haven't botched the programming I get very surprising results. Assuming over that 5k hand session that I were a -4bb/100 loser with a standard deviation of 100 the liklehood of a 15 buy-in downswing is in the region of 5-7%:

Here is the same graph for a break even player:

And for a 4bb winner:

The liklehood of this downswing occuring is slightly lower for the break even player, but what is interesting is that for a 4bb winner with the same standard deviation the liklehood is still around 2%.
First I would like to add that it is extremely unlikely that I was a 4bb loser during my session yesterday. I am not a heavy tilter. Sure I was frustrated by losing that much, but wasn't that aware of it at the time as I don't check my results whilst I play unless I'm taking a shot at something. I've been going over the big pots I played and I can't see spew anywhere. Just the obvious setups and one or two plays I made where I had a read and was wrong/ran into the top of villains range, but we're talking about 200bbs in said spots. What I'm trying to play at is that even given extremely bad play (-4bb/100) based on the variance simulator it is still pretty unlikely to even lose that much.
In contrast to this my empirical winrate at 2/4, 3/6 and 5/10 prior to my downswing was 5bb/100 with a standard deviation of 75bb/100 over 250k hands. Inputting that into the variance simulator makes it almost impossible to lose that much over 5k hands.
Interestingly enough Nassim Taleb (Author of the Black Swan and Fooled by Randomness) actually briefly spoke about NLHE having similar extreme variances as financial markets which makes one wonder if the tools we use to analyze variance underestimates extreme occurences.
So what now? Well my results for this month are very bad by my standards and I am willing to accept that my play may well have been subpar.
I'm not giving up on my hand goal for this month and I'll continue to put in volume. However to ensure higher quality of play I am capping the amount of tables I play to 8 and will make sure I take frequent breaks. I will also stop playing hu when multitabling as this has been source of my losses now and prior to my downswing. I don't mind playing anybody hu, but I think for it to be a worthwile experience I should be able to concentrate on my opponent better.

All the best,

Daniell