Open Side Menu Go to the Top
Register
**** Official Stats and Graphs Analysis thread **** **** Official Stats and Graphs Analysis thread ****

06-24-2011 , 09:55 PM
Oh there's some spew for sure and going through this thread (particularly Mpethy's posts) I've already identified some issue in the blinds but the BTN WR is puzzling me. I purged my DB earlier this year so I can't know for sure but I'm almost certain long-term my WR from BTN is ~30/100 so it's a huge concern to me it's so low in this sample.

The only thing I can point to that is different in this sample is for a lot of it I have been opening min from the BTN and previously I opened 2.5x as standard. I can't imagine that's the reason though. I think I posted this similar breakdown a year or so ago over a a larger sample 500k+ where my WR was more than double I'll see if I can find it and compare.
06-24-2011 , 11:54 PM
I couldn't find my old positional breakdown but I had a suspicion maybe I was defending BB vs SB to frequently (overestimating my ability to outplay post with position) so added these filters:

- My Position = BB
- First raiser = SB
- Cold call = True
- no. players = >2 (excluded HU)



I imagine this is a situation where I should be winning right? Can a decent winner please run these same filters and confirm how terrible I am? TY

Last edited by ChoiceAsBro; 06-24-2011 at 11:59 PM.
06-25-2011 , 02:54 PM
i use PT3 not HEM so im no pro but your filter is for "could cold call" = true which would include all of your folds in the graph right?
06-25-2011 , 08:19 PM
Ahh yes, I meant DID cold call = true



I'm still losing however though the sample is now only 350 hands making it not particularly relevant.
06-25-2011 , 11:58 PM
Looking at your winrate from the BTN over 22k hands is pretty much irrelevant
06-26-2011 , 01:05 AM
Its not 22k hands its 22k BTNs in an almost 200k sample
06-26-2011 , 03:38 AM
Quote:
Originally Posted by ChoiceAsBro
Its not 22k hands its 22k BTNs in an almost 200k sample
Yeah 22k hands, are we not talking about the BTN in isolation here?
06-26-2011 , 04:24 AM
Hmm maybe you're right but I still think it deserves some review. I consider having a BTN wr lower than MP wr even over only 22k instances as an indicator of some leaks certainly.
06-26-2011 , 10:28 PM
gone back and checked that ATS filter mpethy gave about 5pgs and 3 months back.

my wr from CO/BTN when i steal are 62 and 56bb/100 respectively. however, when i see a flop they both drop to around -40bb/100. so i gotta be making some fairly blatant and bad cbetting leaks. any ideas as to what filters i should run to narrow in on that?
06-27-2011 , 01:10 AM
Quote:
To determine whether it is a stealing leak, go to filters/main filters/preflop action facing player/deselect all except unopened. Then go to more filters and look at PFR = true. save and close. Your win rates in the CO, BTN and SB should be slightly positive. If it is red, it indicates that your stealing range is unprofitable, and this is a giant leak. You'll want to compare this to your long term results to see whether it is close, or represents variance. If it is close to your long term results, you are having problems for sure.

Before you do any analysis there, though, you want to go back to more filters and add saw flop = true. Now we are looking at your steals that got called (or 3 bet, and you called the 3 bet). If your preflop WR was green, your WR should go UP by about 20%. If it goes down, it means that your post flop stealing game is a leak.
Is this what you're talking about? I'm not sure I understand Matt 100% here... CO, BTN, SB wr's should be slightly positive when stealing? Mine and yours are massively positive it seems, I'm 88 and 55bb/100. When I add in saw flop = true I get 118 and 7bb/100 so I'm doing even better from the CO but much much worse from BTN. It seems I have a similar leak (at least from the BTN). I'm going experiment with some different filters later, I'll let you know if I find anything interesting.
06-27-2011 , 01:35 AM
Quote:
Originally Posted by mpethybridge
Here's how I examine a player's button stealing range:

In HEM, on the main filters page, deselect all "actions facing player" except unopened. (PT3 is actions tab/click raise first in box)

In more filters, add PFR = True, save and close.

On the position report, you'll be looking at your win rate for your steals.

Win rate in this spot tends to correlate with your stealing %; in general, I see people between 80-120bb, obviously inversely correlated with ATS between, say, 25 and 40.

If you're in this range, you probably don't have a serious problem

Step two is to go back to more filters and add "saw flop = true." (PT3 is misc. tab/situations tab/click saw flop box).

I look for your win rate to go up here. If it doesn't I can normally find a systematic leak in your c-betting game. There are too many variables and potential leaks to list here, but the general rule is that if your win rate goes down when you see the flop, you have at least one of the several potential leaks.
thats what i was referring to
06-27-2011 , 12:12 PM
one thing I just thought of is that you should also select "Faced preflop 3bet = false"

any thoughts?
06-28-2011 , 03:13 AM
Quote:
Originally Posted by Jah Onion
one thing I just thought of is that you should also select "Faced preflop 3bet = false"

any thoughts?
that wouldn't change anything because we've already deselected all actions to us except unopened. so that filter is only when action is folded around to us and we decide to open.
06-28-2011 , 03:38 AM
I think Jah is meaning to filter out the times you ATS and then face a 3b. Could be interesting to look at both this on its own and then filtered out to determine whether calling 3bets from the blinds might be the leak rather than c-betting?
06-28-2011 , 07:54 AM
hmm yea ok that makes more sense. that's definitely an option. but i'm still thinking it would be a cbet leak for the both of us, because cbetting from steal positions happens way more often than playing 3bet pots in those spots.
06-28-2011 , 08:20 AM
OK so this got me thinking and I've run some more filters. I purged a couple months back so total sample size is only 200k but regardless...

In those 200k hands I've been in either the CO/BTN with a chance to steal 20k times

I did steal 7274 times 36%
I was called 2465 times 34% WR = 48bb/100
I was 3 bet 1385 times 19% WR = -100bb/100
1 4bet 169 times 12% WR = 230bb/100
I called 360 times 26% WR = 59bb/100

I'm not totally sure what to make of this yet as the samples are quite small and I don't really have a baseline of what "good" numbers look like but I do notice one thing though straight away:

- when looking at your play in steal spots how you re-act to 3bets is going to be massively important. If you ignore the times my steal went through the blinds successfully 37% of hands are vs a 3bet. The pots are much larger in this situation too so mistakes are magnified.
06-28-2011 , 12:24 PM
Quote:
Originally Posted by bleffo19
that wouldn't change anything because we've already deselected all actions to us except unopened. so that filter is only when action is folded around to us and we decide to open.
faced 3b vs the blinds dude ldo wtf bbq
07-01-2011 , 02:44 PM
Hey, guys I need opinion. First are these stats allright and second could someone give me advice, how can I minimize the blinds losses. This is my last month stats, I play 25nl 6max at mergenetwork. Thank you all!
07-02-2011 , 12:41 AM
steal small blind more, cold call more in position, resteal from blinds more, win some showdowns
07-02-2011 , 08:18 PM
i'm trying to switch from MTTs to cash games and failing pretty miserably so far. this is my first few K hands at 50nl FR on stars. what am i doing wrong? is it too small a sample size too soon to matter? am i playing too many hands? any advice much appreciated.



07-03-2011 , 01:09 AM
Quote:
Originally Posted by zedveron
i'm trying to switch from MTTs to cash games and failing pretty miserably so far. this is my first few K hands at 50nl FR on stars. what am i doing wrong? is it too small a sample size too soon to matter? am i playing too many hands? any advice much appreciated.



sample size is really quite meaningless, but there are already some things that stick out.

you are going to SD far too often and also losing too often at SD. this, in essence, means you are being too much of a station and paying people off far too light (which is also why your river call efficiency is fairly shocking too). you are also not winning enough when you see a flop - be more aggressive to try and take people off of the weaker portion of their ranges (this goes hand-in-hand with your low-ish (but not drastic) AF).

cbet a tad less and fold to people's cbets a touch less too - so many people cbet once and give up, and you can print $$$ by floating them even just once.

hope this helps
07-21-2011 , 06:01 PM
getting crushed at 50NL...kinda small sample. just this month, moved up from 25NL at FT to 50NL stars. most of my loses come at the end of sessions, because of tilt calling and flipping. I know my blind play needs a lot of work. I am just lost there. with the high amounts of stealing that goes on at that stake i feel a need to agressivly defend my sb/bb. Im sure this is a good approach, but im doing exactly the wrong thing at exactly the right time at the blinds. (if that makes sense, anyway) other than that, i think my play is OK from all other positions. ive read that COTW thread on red line and stuff and that helped at 25NL can show graph to confirm if need about (125Khands). as for 50NL (22K sample) my graph is awful, Red line are all my loses, and blue line is even (rare in my case, usually its a christmas tree, so even though im running above EV$ i am getting coolered alot, as well TILT play is half of the loses)

I would appritiate any analysis or input on these stats, if there are any spesific filtered graphs you need/want to see, just ask.

here are my stats filtered by position
(hope the picture works)

thanks in advance
07-21-2011 , 06:23 PM
a follow up by stake... its all at 50NL just had to make some room for the whole line to fit in the screan shot...still missing ("VS 3BET FOLD%" which = 61.0)

07-21-2011 , 06:38 PM
1. the difference between your 3bet and sqz is weird
2. your wtsd is too high, so youre either not value betting thin enough on rivers, or youre being too much of a station...but youre w$sd is high so maybe youre just running good? didn't expect to see that
3. your bn/co/blind rates are all not good but its too small of a sample size to really break it down like that
4. your aggression numbers are kinda low
5. your steal is much too low, particularly on the CO
07-21-2011 , 07:26 PM
Quote:
Originally Posted by SA16
1. the difference between your 3bet and sqz is weird
2. your wtsd is too high, so youre either not value betting thin enough on rivers, or youre being too much of a station...but youre w$sd is high so maybe youre just running good? didn't expect to see that
3. your bn/co/blind rates are all not good but its too small of a sample size to really break it down like that
4. your aggression numbers are kinda low
5. your steal is much too low, particularly on the CO
1. What should it be at? like what is acceptable, or what did you expect/what hsould i expect?
2. ive always had a high showdown% and won a lot of money their. i have pretty good reads IMO (this is where im right, or either make a complete fool of myself) but my showdown statistics i feel have never been a problem. however, this may be why i am losing, so i am open to all insights.
3. what do you mean by btn/co/blind rates as not good? too loose? to tight? spewy? (also im assuming youre talking about CutoffUO PFR/btn/SB)
4. I found that too, at 25NL my agression was much higher, maybe it was because of the "jump" and the scared money kinda feeling? but i know that i am seeing high pockets JJ QQ KK and getting ace high flops a lot or 3 flush on the flop. (that should not be a just reason why, but may explain why they are somewhat lower)
5. i find that when i steal from CO iam frequently getting 3bet from btn

that being said...

1. what should i aim for 3bet and sqz%
2. not 100% sure what you mean by this (the thin value river part ish)
3. how do i improve my co/btn/blinds rates?
4. how much should i increase this? cbet more? barrel more? check raise more, raisecbet more?
5. CO play improvement tips? ie..facing 3bet from btn/blinds

      
m