Thanks, for detailed post.
We can see that the red line is plummeting and holmeboy already mentioned that
I was never paying attention to this, because was hearing everywhere - its micros, red line at micros is very often not good. So was focusing on some other things.
Since you play looser than your average player you enter more pots and it seems that you abandon them more frequently. One of your biggest "silent killer" is the small pot. That's generally the pot where you raise, make a cbet and then give up without giving it much thought after. These pots are small, they evade the session review and you rarely give them a second look.
And they are not even silent it looks. When I am playing - I often feel like, if I dont win some big pot, my session is at beast breakeven, at worst - negative. Just did not kwow what to do about it. Was trying to think - oh, its just variance. Like people say 10K hands is not sample, 20K hands is not sample, 30K hands is not sample. So keep playing same way hoping variance will even out ♦
But somehow variance should not always be negative. It should have good wining strikes also.
I remember I was adviced at 2012 december 09 database review that my flop agression is big and turn agression is low, so I was working on that, but looks like it was unsuccesful.
Before 2012 dec 09
afq f/t/r: 55/32/19, flop cbet 79, turn cbet 47
after 2012 dec 09
afq f/t/r: 51/35/23, flop cbet 69, turn cbet 47
So kind of fixed a bit, but did not know if enough and then probably decided to wait for some sample.
So the first thing I would do is reduce the range of hands we play based on what hands are not profitable, especially in EP->BTN.
I guess that one will be tough, because the ranges are very dynamic especially in BTN which is probably warying 40-100% depending on players in the blinds. Probably will need to take individual hands where I open loose and post in forum.
You bet more than half the pots you see a flop on. Then on the turn you bet only less than one third and by the river you are down to almost 1 in 7. So clearly there's money being put into the pot when it shouldn't be or left there when we could attack it (also most likely missing valuebets on the river).
I guess maybe I should reduce flop aggression, because on river I think I was betting for thin value and with maheepsangari we discussed who plays same limits and wins - he says I bet too thinly. And I think it makes sence, because maybe because of this my wtsd is low - people don't call my bets, and W$SD is low - when they call - they have better hand. Or maybe I should use smaller sizings when value bet to let them call more light.
It also makes sence - when I bet flop with trash, I have weak range on turn and river, so I cannot get value and my bluffs also dont work.
If the player is going to fold, I probably don't necessarily need to bet flop, because he can fold his weak hand on turn also. That is problem when I am OOP - then if I don't cbet, he might bet into me and if I have trash, I have to fold.
Which means - I have to check top pairs to have range to call. And I try to do this and it feels that I catch lot of bluffs with this when I check.
I'd look at BTN only and see hand groupings, and then move to each position. You may be surprised.
VPIP = TRUE, all hands
BTN:
Spoiler

Suited 2 gappers check:
Biggest pot won J8s - 68bb when I hit 2 pair
Biggest loss: T7s - 167bb flopped flush vs flopped nut flush ♦
CO:
Spoiler

Suited 1 gappers:
biggest win: QTs 57 bb nut flush vs 7 high flush 4tone board
biggest loss: 79s 112bb trips vs FH
MP:
Spoiler

UTG:
Spoiler

SB:
Spoiler

BB:
Spoiler

So the tendency is - suited hands loose often which is bad.
I'd also be interested to see your fold to 3-bet stats
By position all hands
SB: 72.7
BB: 58.3
EP: 57.5
MP: 68.0
CO: 72.9
BTN: 79.1
and after match 19
SB: 71
BB: 45.5
EP: 56
MP: 67.6
CO: 70.5
BTN: 73.2
Called preflop 3bet, all hands (terrible):
Spoiler

Speed, how you get Bogdan doing this? grin I want it too
I guess the first thing to do is to create a thread in Begginers forum section
And btw I think my mistake is to wait too long for variance to even out. Maybe its enough like 10K hands for agression stats to check.
FOr example this month filter 11K hands, afq f/t/r 48, 34, 25.4 which is very similar but flop reduced bit.
And I really reduced flop cbet which is over this sample - 60.5, turn cbet 39. ANd I was trying to reduce flop agression to make turn agression bigger. But turn agression still is low