Message Board : Interesting bits from comparing winning sessions vs losing sessions

I spent some time today analyzing the data from the sessions I’ve played over the last 2 months.  (~5k hands, all NL25 full ring).  These are the first 2 months I’ve finished ahead money wise, but not by much.  It seemed like a good place to start, since my goal wasn’t just to figure out how to suck less.

I found some interesting things that I think will help me quite a bit going forward.  I’ve decided to post them, so maybe you’ll learn something, too.  Or, maybe you’ll have an interesting alternate explanation.

Historically, I’ve played too many hands from early position.  My first thought was that I’d see big differences in my VPIP/PFR by position. I may still be playing too much from early position, but the VPIP and PFR numbers by position were very similar for my winning and losing sessions.  What I did learn is that over this stretch, looking at either all sessions, or even just the winning ones, the only position where I didn’t make money was 3 off the button.  I haven’t dug into this yet, because it wasn’t my focus, but this may be interesting in its own right.

So, it’s not preflop hand selection by position.  Bummer, that would be easy to fix.  In fact, in my winning sessions, I was (very slightly) looser. Probably not statistically significant.

In fact, My statistics look similar until you get to Went to showdown, and won $ at showdown.

Win: WSD: 20.44%    W$SD: 48.66%
Lose: WSD: 26.59%    W$SD: 29.85%

Well, that sums up losing, right?  Go to showdown more, and win less often.  The obvious answer here would be that I’m calling too much postflop.  But I don’t think that’s the case.

A couple of other stats  (Bear with me, as I think these go together):
On my steal stats, W$SD(win): 77%.  W$SD(lose): 0. (!?)
First action on flop after PFR.  Check(win):11, Check(lose): 6.  Call also higher for winning sessions.  Raise + C/R were higher for losing sessions.

Perhaps most interesting were the aggression factor numbers.  Flop/turn/river for wins: 2.84/2.57/2.97
For losses: 3.17/3.54/1.92

Losing sessions also featured more check raising on the flop.

I think what this boils down to is one simple thing, and a few more complex.

Put simply, in my winning sessions, I’m letting the game come to me a bit.  In my losing sessions, I’m trying to outplay people after the flop, but giving up after firing 2 barrels.  (Normally, I’d say that’s the sane thing to do, but see the river discussion later.) 

Straightforward aggression doesn’t seem to have the desired effect.  I saw more than a few instances of C/R on a dry flop only to be called by A/K high.  A lot of those hands had the line of C/R flop, lead turn, check river (and the river went check/check, and A/K high won the pot, or similar.) This is good info, though.  This would be a great way to play a flopped set on a dry board.  The pots end up reasonably large because of the C/R and turn bet.

So, C/R a dry flop (and optional turn lead) out of position doesn’t have the desired effect.  What does?  I found that my best steal hands were in position on boards with obvious draws (flush draws work well).  Float the continuation bet, and you can steal if the flush card hits.  The benefit here is that with position, you often get 2 shots at the flush card for the price of 1 call.  Then they can fold when you hit your "obvious" flush.  Score yet another one for the power of position.

One other thing that I think is happening was pointed out by Todd in an earlier thread.  He pointed out how powerful the bet/bet/bet line is.  And my data bears that out.  I think that about the only thing that explains the big difference in WSD% is that in the winning sessions I’m betting the river more often, and getting folds on the river more often.  I’ve noticed that I’m getting folds on the river in reasonably big pots for 1/3 pot bets.  I’ve been using that bet size as a blocking bet, and seeing *way* more folds than I ever thought possible.  It’s like a lot of people are choosing call/fold before they even see the bet size.  Talk about exploitable.

So, if my data is bradly indicitive of small + micro stakes games, the advice I come away with is this:

- Don’t try to steal on the flop, unless you’re willing to fire 3 barrels.  Talk about high risk.  You better be able to assign preflop ranges pretty well.
-  If unsure about the proper river action, and you have a mediocre hand with some showdown value, bet 1/3 of the pot.  It’s amazing how many people fold what turn out to be better mediocre hands.  And I haven’t seen anyone catching on to raising that bet.
- Float more (this requires position, and allows you to play based on the board more than your cards).  I think if you were good at deducing when a player is likely c-betting with no pair, it would be profitable to play any 2 cards from the button by stealing on the turn.

Anyone interested in more info?  Different interpretations? 

One thing I found from this exercise is that it’s definitely tough to try to take this data and turn it into 2-3 things to work on to help your game.  But definitely worth the time invested.

Nice post weasel.  

Thanks.  Haven’t seen much like this, so I wasn’t sure how it was going to go over. 

5K hands isn’t very many at all.  It’s hard to draw any conclusions at all about win rate from a sample that small.  I read a really interesting post about win rate and variance over at 2+2. 

I’m pretty sure I’ve read the same post.  But I’m not so concerned about my exact win rate at this point.  I grouped my development into 3 winrates: defnitely a loser (goal: lose less money), about break even (goals: win more when I win, lose less when I lose), and definitely a winner.  I’m pretty sure that with a couple of positive months, I’m now about break even.  As long as I only play 1 table for a couple of hours in the evening, I doubt I’ll get to where I have much confidence in my exact winrate, but I don’t really care that much. (I don’t see myself ever quitting my day job).  (My original goal with poker was to be able to go to vegas, or a casino, and not lose too much $$ too fast, and ideally be roughly break even.  I’ve done that in 1/3 NL games in vegas last summer, so I’ve met that goal.  Next goal is to be able to pay for my trip at the tables. )

Many live game players will play 100K hands in a lifetime.

That’s why you see so many 20-somethings (sonething low) winning big tournaments.  They have as much experience at 20 as live pros have at 40 or 50. 

That said, your vpip, pfr, aggression factors, steal numbers and a couple of others converge very quickly.  Those numbers really do tell a strong story about your game.

Sure.  One thing that I also neglected to mention was that since this is a relatively small sample (and I know that), I actually looked at the hands that generated a given stat (as much as possible) individually.  At least the outliers–the big wins + big losses.

How does your delta from EP to BTN look in terms of vpip and pfr? 

Well, on the button, I’m playing close to any 2 if no one has shown much strength, so it varies from 17/12 UTG, to 29/17 (hijack), 36/22 (CO), 56/21 (B).  I know you play higher, and I think if I were playing higher against people that could actually use position, I’d have to play tighter.  Or if there were more preflop aggression…

I play something like 9/7 UTG and UTG+1.  I’m working towards something more like 7/7.  All pairs, AK, AQs, KQs is ~8.  I’m trying to play pairs and AK, but I sometimes lack the discipline.

If there are aggressive players in the game, I’ll cut down to about this range for EP.  They don’t show up often, so my EP range is mostly what I can play postflop. 

1 or 2 big hands will push a postion into the red over 5k hands.

Sure.  But, like you said, I think it’s worth a deeper look.  It’s nice to be able to filter a sample, sort by $result, and then replay the hand.

For example, it wouldn’t surprise me if those weren’t multiway pots where I had mediocre hands with someone yet to act after me.  We’ll see.  A study for another day, or night.

losing overall with with KKs, and overall with AKs. 

That could also be overplaying 1-pair hands, especially if other players have you tagged as such.  (I’m not saying that you are, but I know if I were losing with those hands, I’d probably want to know why) 

Did you take a look at your win vs equity with pokerev? 

Last time I looked into that, it wasn’t ready for prime time.  Sounds like it’s worth another look. 

For example, I win a ton of small pots.  I don’t always make great decisions in the bigger pots.  PokerEV was a good tool to take a look at some of those things.

That’s great to know, since it’s pretty much my story, too.  For a while, it seemed like I couldn’t win if the pot was > 50BB. 

So really, you’re only looking at a difference of 5 or 6 hands going to showdown vs winning and losing. 

VPIP about double, and lots of checked big blinds, but you make a good point.  Even if it’s 20, that’s still pretty small. 

One of the things that is probably worth doing is looking at all the hands that got to showdown and how they got there. 

I’ve done that to some extent with the losing hands, and I found too many C/R dry flop, bet dry turn, check river, and have them turn over 2nd or 3rd pair or something equally silly.  I defnitely have to save the C/R  /  bet line for when I know I’m ahead.  Because it’s not getting as many folds as I would have thought. 

Did you just run into a couple of coolers?

I wish I could blame it on that.  I didn’t really see any huge pots I lost where I was thinking it was a cooler. 

Yeah, those players are worth a note.  Value bet only.  Any players that will call a flop c/r and turn bet cold with A high should be your friends. 

I would look more closely at those hands where you c/r and lost.  Take a look at the board texture, the positions, hand strenght, pre-flop action, etc.  It may well be that you were picking bad times to bluff. 

I’m not convinced that’s the case.  I actually brought the move over from my tournament game.  In tournaments, I pretty rarely see someone call a flop C/R with less than top pair, but in NL25, I’m seeing it all the time.  I won’t do it as a bluff if there are any cards above a jack, or any straight or flush draws.  And these hands usually get to showdown, as I haven’t seen anyone thinking deeply enough to think it’s a pure bluff and to play back.

But, given your opposition, you may want to take bluff c/r’s out fo your game entirely and just c/r for value.

I think at a minimum, I need to show down a value C/R before trying one. 

That’s the thing about loose/passive oppostion.  You feel like you can run over them, but you just can’t.  It’s infuriating.  "You called with that?!?"  Value, value, value.  Repeat that to yourself over and over.  I find when I’m playing my worst, I’m trying to run players over and not waiting for values too bet solidly.  I also find that I get impatient and start making thin calls that are nothing but tilt inducing.  You see some guy drag a monster pot when he hits a gutshot on the turn when he should never have seen the flop… Drives me crazy.

I’m *so* glad I’m not alone here.  This is a lessone I have to keep learning again and again.  I think it’s because the home games I play in don’t have anyone who’s L/P.  So, every time I play one, it messes me up again.  

That is crazy to fold there for sure.  I would say that the bet/bet/bet line has a lot to do with board texture.  You need to be able to represent a strong hand.  I think this may be a bit tougher to pull off in your game where people are calling lots of bets without much. 

It seems like they drop on the river, though.  I stll haven’t figured out what sorts of hands justify call/call/fold on a board with no draws.  With draws on board, I understand, but… 

Yeah, it takes a lot of heart to 3 barrel a lot.  I generally only pull that out when the board is just right and my opponent is just right.

I’m starting to wonder if a small third barrel isn’t profitable in these games.  Maybe the small bet looks like a value bet?  And they’re being "clever" and folding to my "obvious" value bet? 

I think this is better if you have no showdown value.  Maybe out of position its worth it too block in this fashion.

Again, I’m seeing so many folds lately to my blocking bets that I’m starting to wonder.  Maybe it depends on having been the aggressor, but I uaually am.  

Floating can be pretty poweful. 

If there’s anything that demonstrates the power of position…

Thanks.  Haven’t seen much like this, so I wasn’t sure how it was going to go over. 

One of the things I really like about NPA is that you can make these long form begginner/intermediate posts and get thoughtful responses.  You don’t really get much of that over at 2+2.  It’s hard to understand concepts in difficult hands when you get an expert response that is something like "b/f flop, c/f turn".   

I’m pretty sure I’ve read the same post.  But I’m not so concerned about my exact win rate at this point.  I grouped my development into 3 winrates: defnitely a loser (goal: lose less money), about break even (goals: win more when I win, lose less when I lose), and definitely a winner.  I’m pretty sure that with a couple of positive months, I’m now about break even.  As long as I only play 1 table for a couple of hours in the evening, I doubt I’ll get to where I have much confidence in my exact winrate, but I don’t really care that much. (I don’t see myself ever quitting my day job).  (My original goal with poker was to be able to go to vegas, or a casino, and not lose too much $$ too fast, and ideally be roughly break even.  I’ve done that in 1/3 NL games in vegas last summer, so I’ve met that goal.  Next goal is to be able to pay for my trip at the tables. )

I’m in roughly the same boat you are + 1 table.  I have no intention to be a professional mid-limit grinder. 

How does your delta from EP to BTN look in terms of vpip and pfr? 

Well, on the button, I’m playing close to any 2 if no one has shown much strength, so it varies from 17/12 UTG, to 29/17 (hijack), 36/22 (CO), 56/21 (B).  I know you play higher, and I think if I were playing higher against people that could actually use position, I’d have to play tighter.  Or if there were more preflop aggression…

I’d really consider playing fewer hands.  Especially in EP.  I know you’re not getting punished for it particularly, but you are probably drip, drip, drip leaking out BBs.  You will almost certainly improve your win rate if you cut the number of hands you play in 1/2. 

Sure.  But, like you said, I think it’s worth a deeper look.  It’s nice to be able to filter a sample, sort by $result, and then replay the hand.

It is always worth looking at your big winners and big losers after every session.  It only takes a few minutes and is well worth it.  I find it a lot easier to see something as a mistake, even in pots I’ve won, when I’m reviewing it.

losing overall with with KKs, and overall with AKs. 

That could also be overplaying 1-pair hands, especially if other players have you tagged as such.  (I’m not saying that you are, but I know if I were losing with those hands, I’d probably want to know why) 

There were one or two badly played hands for sure, but it was mostly cold decks.  KK vs AA.  Overpair vs set.  AK vs PP’s all-in preflop.  That sort of thing.  It only takes a couple of hands to flip you over to the red. 

Last time I looked into that, it wasn’t ready for prime time.  Sounds like it’s worth another look. 

I really like pokerEV as a tool. 

I’m not convinced that’s the case.  I actually brought the move over from my tournament game.  In tournaments, I pretty rarely see someone call a flop C/R with less than top pair, but in NL25, I’m seeing it all the time.  I won’t do it as a bluff if there are any cards above a jack, or any straight or flush draws.  And these hands usually get to showdown, as I haven’t seen anyone thinking deeply enough to think it’s a pure bluff and to play back.

It sounds like you may need to just bluff a bit less in a game that loose. 

But, given your opposition, you may want to take bluff c/r’s out fo your game entirely and just c/r for value.

I’m *so* glad I’m not alone here.  This is a lessone I have to keep learning again and again.  I think it’s because the home games I play in don’t have anyone who’s L/P.  So, every time I play one, it messes me up again.  

I find this even tougher live than online.  Live, low limit games can be sooo much looser and slower.  It can be really hard to find the patience.

Yeah, it takes a lot of heart to 3 barrel a lot.  I generally only pull that out when the board is just right and my opponent is just right.

I’m starting to wonder if a small third barrel isn’t profitable in these games.  Maybe the small bet looks like a value bet?  And they’re being "clever" and folding to my "obvious" value bet? 

I think this is better if you have no showdown value.  Maybe out of position its worth it too block in this fashion.

Again, I’m seeing so many folds lately to my blocking bets that I’m starting to wonder.  Maybe it depends on having been the aggressor, but I uaually am.  

Aggression is also very powerful. 

---