Colin252's avatar

Colin252

114 points

I did some (quite a bit actually) MDA (over large data samples).
I looked at the output, condensed it (probably over simplified, but integration is more important than accuracy for me right now) and applied it over a tiny sample of play (10k).
Given how differently I'm playing than I was previously, I'm honestly a bit lost as to how and when to attempt to determine which of these changes I should keep and which I should not.

Right now I'm so far out of my comfort zone it's not even true, on top of which I'm struggling to interpret the feedback I'm getting.

Thoughts on methodology for comparing the efficiency of changes (and samples required)?

April 27, 2019 | 7:36 p.m.

Comment | Colin252 commented on Population Exploits

EYV (first repsonse) has added a lot and shown how clever they are by laughing at you. I'm sure that, like me, you are in awe of how incredibly smart they must be.
Or not.

This is a very good question Vegas777. You would actually be silly not to ask.
It's such a good question, that nobody who knows the answer will tell you (for free).

What you need to do, is figure out a way to do some MDA for yourself.
Whilst this may not be something you're used to, breaking through barriers like that is a requirement for growth, which is why you are here :-)
Go for the low hanging fruit.
Say we open btn, BB calls, we're HU to the flop IP. I chose this as it happens relatively often compared with other situatuions, so even a smallish sample will yield insight.
Whats your cb success?
Whats your delayed cb success?

Now that we have these two numbers...

What can you infer from these two numbers?

April 27, 2019 | 7:13 p.m.

Build it in notecaddy. The stat exists in Hm2 but only for the bettor. i.e. dealyed CB success %.

In truth, doing some player pool analysis in reports is probably more telling/helpful than any individuals stats. You'll see what different player types are doing in this spot, who defends their checking ranges etc.
There are levels, like say BB CC OOP v BTN open:
1. Population OOP CC X-X/F freq
2. Player type OOP CC X-X/F freq
3. Player type OOP CC X-X/F freq by flop type
With each step you will require an exponentially bigger DB in order to get a sample capable of providing genuine insight.

April 27, 2019 | 6:53 p.m.

Comment | Colin252 commented on NL10 AKvsBTN

River raises are the nuts. At low stakes this silly simple rule will save you a ton of $$$.
Re turn bet a good question to ask is how does he play this very hand if you bet turn?
Given he has no SDV and still checks his slim but nut draw rather than bluff when you check to him, I'd say he'd call rather than bluff raise when you bet at him.
So if he calls a with GS no SDV, what do you think he bluff raises? Not much I'd say.
So then we can say: if it's true (as we deduce here) that when he raises it's almost never a bluff, we can pretty easilly make a fold and crush him by doing so knowing he has AJ/4/5 or sets an overwhelming majority of the time.

April 22, 2018 | 8:32 p.m.

My solver tells me to x/r 30% of my flop range here. Totally unexpected... Not at all what I would have suggested had I not looked at the GTO first.

March 2, 2018 | 5:33 p.m.

NL 5 I don't mind a bigger pre/flop sizes.
Tough to fold KK, as Kalupso says above, micro stakes guys do some non sensical stuff.
I do think the clickback won't be a set too often, as in only a fraction of his sets might play this way - people love to "slowplay". Also, people rarely play draws aggressively - which makes me think if we weren't beat otf then we aren't beat otr. These two observations combined I guess means I'm just gonna call off.

March 2, 2018 | 11:51 a.m.

Yeah you know if you check he will shove turn @ high freq. You know you can't call that bet.
You don't want to fold having put in so much money and not realise your equity, so you find a way to not fold.
You pay way too high of a price to realise your now relatively poor equity.
Not wanting to fold is not that same as making a plus ev play.
If you look at flop bet/3b ranges in these stakes you'll never use this line again.

Dec. 2, 2017 | 12:50 a.m.

One thing to learn from this hand is how much each card can change things.
Here you get probably the literal worst card on the deck for your hand. It completed both class of hand that beat the nut str8.
Its a good habit to just accept that no matter how good your hand is on the flop there will always be one or even a few cards that are just so awful that you need to slow up and consider folding v continued aggression.
It sucks when that terribad card comes and you have to give up what was the mortal nuts but it rarely happens. Most of the time, you will get a turn much better for you than Jd and you can fist pump get the money in.

Dec. 2, 2017 | 12:41 a.m.

Comment | Colin252 commented on NL10 river decision

Didn't read all the comments first....
Yes, I'm less tactful than you fo sho.
It frustrates the hell out of me that struggling micro players {i been there too} make all these - mostly emotional - attachments to things they think are a big deal or they want them to be a big deal when in truth these are the exact same things holding them back from doing well and if they focused on less "sexy" things they'd do so much better.
Like sorry OP I'm not trying to bag on you and if you want to PM me I'll for sure help you out with anything you are really struggling on by way of an applogy for this rant -
To be so confident you have a "good read" on a player when in reality you don't know too much from 2 unrelated Sd hands, but in the same breath you don't know what the pool does on average either multiway or otr - when you have literally thousands more examples of that than you do of this guy showing down is kind of ridiculous. It illustrates both paying too much attention to the wrong things - as one of these is a very reliable indicator of how to make informed strategy choices and one of them just isn't - and the logical blindspots that these false knowledge ideologies can create.
ie must adjust/outplay this guy by trying to win the hand
v
the pool under bluffs i'll just fold EZ game

One of these sounds like you're a badass who won't back down, but the other one will win you a lot of money at micro stakes.

Nov. 29, 2017 | 8:17 p.m.

Comment | Colin252 commented on NL10 river decision

Yes I'd likely fold the turn.

You saw the guy show down 2 hands, depending on the circustances of those hands, they might in no way relate to this situation.
So you're basically making BIG adjustments which woud be really bad v most players. You'd need a very good reason to do that, your "read" here isn't near reliable/applicable enough to want to make major strategic adjustments which would be very VERY bad v the pool.

You so want the read to be correct you're being guided by that instead of objectively assessing how useful the information you have is, if at all.
Basically you're trying really hard to "outplay" the guy when in truth he's probably so bad that by just value betting appropriately and folding well you'll crush him without all the fancy "read based" nonsense. It's not a sexy answer and doubtless the opposite of what you wanted to hear, but that doesn't make it any less true I'm afraid.

Nov. 29, 2017 | 7:15 p.m.

Comment | Colin252 commented on NL10 river decision

If you filter where you called 2 streets turn and river with 3rd pair in a 3 way pot, I'd wager you're losing a lot of bb. In small stakes where people don't bluff enough generally, especially multiway and especially otr I can't imagine why on earth you'd showdown here. Other than is that TT looked good preflop. I seriously doubt you'd defend A7s otr, but it's the same hand - pure bluffcatcher.

Nov. 27, 2017 | 11:01 p.m.

River raises are the nuts generally.
Later street raises on paired boards are trips + way more often than they should be

Nov. 22, 2017 | 8:30 p.m.

Comment | Colin252 commented on KJs vs x/r flop

Big database, hours of study/filters etc.

Nov. 20, 2017 | 7:51 p.m.

Comment | Colin252 commented on KJs vs x/r flop

It's not totally implausible that he has AJ, but I think it does get played this way rarely enough in my experience/DB that it probably shouldn't be the difference maker in your decision here. Probably this is where a few stats on villain would be very helpful, the tighter he is the more I'd fold.

Nov. 20, 2017 | 7:51 p.m.

Comment | Colin252 commented on KJs vs x/r flop

I think otr you'll get shown A3s at a similar freq to 45s. IME at these stakes when villain is bluffing otf they don't follow through ott often unless their eq improves, and they very rarely bluff river by this line.
Of the four most common player profile, when they take this line they have TPTK+ 86% of the time on average. Pretty easy easy fold v unknown based on population data.
Often in low stakes we're faced with a "difficult" spot like this and curse our luck, say something like "I'm too high up in my range to fold" then donate the rest of a buy in to the nut peddler we're playing against.
Truth is people aren't aggressive enough, thats never more true than otr. Yes it sucks emotionally folding, but it will be both the more +ev line in the long run v this pool - AND build all important discipline for the future. It isn't sexy and so doesn't get much coverage but being able to fold "good" hands is a pretty significant part of your edge at these stakes.

Nov. 19, 2017 | 9:05 p.m.

Yes sample is an issue, I'm only seeing 35% folds otr v my bluffs but much greater folds in general. I'm massively underbluffing river as I have 10x the sample of bbb river value bets to bbb river bluff.
Over what sample/stakes are you seeing these fold numbers?
Re turn folds, your tiny flop size causing wider call range is likely the cause of the high turn fold.
Fwiw when calculating turn bluff value remember every hand has a higher than zero ev when checked back so with say 2/3 turn sizing 44% folds could well mean your worst bluffs would be a higher ev check, with the plan of bxb when you get the chance.

Nov. 12, 2017 | 10:24 p.m.

Mental block I run into in this spot is that I think I'm gonna value cut myself because I only think about which made hands will call and how they do v AJ. In reality, on wetter boards a TON of our value will come from charging draws.

Nov. 12, 2017 | 2:58 p.m.

On wetter board it might be reasonable to get 3 streets via B-B-X/C when draws don't come in.

Nov. 12, 2017 | 11:21 a.m.

Of the three criteria you mentioned I think opponent type is by far the most important at micro stakes. How capped he is or which strong hands you block are totally irrelevant by comparison if villain is happy to call 2nd pair good kicker.
Also paying more attention to this will help us value bet more effectively - which somewhat unglamorously is likely worth more EV at these stakes than bluffing well I'd guess.

Nov. 11, 2017 | 11:30 p.m.

1.Done alittle crev, didn't think I was cbetting too much on middling co-ordinated flops. Wrong about that - 68% cb - 38% cb success, and facing a x/r way more often than I am in general. Great point.

  1. Yes turn is merged as I'm seeing a better ev B-B-X than B-X-C with all the likely candidates. hour glass idea fits well with my overall btn cb success stats at 45/38/58 but as I said above, I have a tiny sample of 3 barrel bluffs but they really aren't going well. Hard to say if I'm just picking terrible spots - wholly possible - or just sample is insignificant. But it's pretty hard to ante up the 3rd barrel when it's costing me money as of yet.

Nov. 11, 2017 | 4:31 p.m.

Yeah my flop cb is too hgih no doubt - it's lower over my last 100k hands but still 65%. I use 1/3 - 1/2 pot sizing mostly so 45% folds is a pretty good profit with my worst hands so bluffing v typical villains seems a decent exploit as I capture around 20% eq with my lowest value hands.
Re board textures, yes not a strength of mine at all. I get the concept in the general sense but struggle to apply it well given the unpredictability of my opponents call ranges.

Nov. 11, 2017 | 4:21 p.m.

Noted, for regs sb range is very different/stronger than BB range. For fish it's almost identical, so yet another reason to pay close attention to opponent type.
I am very exploitable otr, but as discussed in this thread:

http://www.runitonce.com/nlhe/on-exploitability-ip-hu-srp-via-b-x-c-li/

All the hands I might use for B-X-C have a higher EV as B-B-X, so exploiting trumps trying to be inexploitable.
Re weak draws, cb success stats BTN HU v BB are 45/38/58.
Looks like you want to TB all day long, but in practice when I look at when I bluff that cb success freq is way down at 35% and I'm better off checking back river when I can't value bet.
On the other hand, when I B-X-B bluffs I get 68% folds and a huge profit with 2/3 sizing so that seems much more appealing given the low fold equity ott followed by poor river bluff success.
River bluff sample is small though and overall TB success is high, so maybe build a bigger sample on that. Seems like I rarely do it as I have 500,000 hands in this DataBase and I only TB bluff 30 times.

Nov. 11, 2017 | 4:17 p.m.

Good fold. out of 81 chances to VPIP he has done so 9 times, only once with aggression.
What can we say with relative confidence?
He is highly selective of hands to play, even then he is rarely aggressive. Then he 3bets you - you think it's 85s?
It could be, but his stats suggest he'll show you AQ+, 99+ ten times for every one time you see a sc.
It's counter intuitive to outplay someone by folding, but in micro stakes recognising when to fold is a MAJOR part of your edge.
A great example of totally owning this guy would be open 92o from btn v his bb then fold to 3b. You will crush him by doing this.

Nov. 10, 2017 | 7:21 p.m.

a) not my handle
b) try not to take feedback personally
c) rather than attempt to change the subject, it would behove your development to accept people here don't think the villain was stupid. There is something you can take from that to improve yourself in the future - what a great opportunity - don't waste it.

Nov. 7, 2017 | 9:46 p.m.

it's not so much about strength NOW but vulnerability to future cards and chance of improving on those cards.
KQ is significantly better than TT in both of those areas. Mid PP are basically a horror show for multi street bluff catching.

Nov. 7, 2017 | 12:35 a.m.

^^this^^ with the caveat that whilst some pools over fold v 3b, EP fv3b is not going to be exploitable in the way btn fv3b may be.

Nov. 6, 2017 | 11:27 p.m.

His line is only stupid if people exploit him by actually folding gto calls.
If people know he never bluffs but always call anyway then his line is REALLY SMART.

Nov. 6, 2017 | 11:25 p.m.

I just hate the way this hand is played.
If you think he raises flop light then x/c rather than bet.
If you think he doesn't raise flop light then bet/fold is good.
But bet flop v aggro player then call down TT which almost never improves is just spew.
Would you bet/call down 89? It's all but the same hand in practice v his range but 89 improves way more often.

Nov. 6, 2017 | 11:20 p.m.

Nobody 19/13 is value betting Qx ott. They just aren't. Let alone bet the river. Let alone bet/3b the river.
Your hand ranging process is madness from the Planet Zorg.
Rather than thinking about what you actually get shown here by this type of player, you're thinking about what he'd need to have for your play to be ok. For me it comes down to the fact that you are looking at: absolute hand strength
ie "I have trip aces thats a great hand woop woop"
rather than what you should be thinking
ie " I have trip aces but the worst hand this tight and likely conservative player who exercises very controlled aggression with well selected high quality hands will bet for value is AJ which beats me.
His type is renowned for under bluffing river, I can exploit him by folding because my seemingly strong hand is actually very weak relative to his betting range in this situation."

Nov. 6, 2017 | 5:29 p.m.

Samples for stats vary from stat to stat because some things have the "opportunity" to happen very very often and some almost never.
VPIP for example gets the chance to happen 100% of hands, so 60 hands is a sample of 60 for that stat. It's not certain but you can infer he is tight with reasonable confidence - where as you point out this player has only 2 chances to 3b in the 60 hands we played so it'll take quite a few more hands before you can rely on his 3b stat. His x/r/f river in 3bp will take a crazy number of hands to get a sample by comparison, 5000 hands may not see him get the chance to do this even once.

Nov. 6, 2017 | 5:28 p.m.

Load more
Runitonce.com uses cookies to give you the best experience. Learn more about our Cookie Policy