Guide: Analyzing your own hands [long]

As per this post I'm going to try to make a little (hah!) instructional text on what you should be doing when you analyze your own hands. The same can be used for analyzing other people's hands of course, but being able to break down your own decisions away from the table is a really, really useful tool. Here goes.

Preflop and basic information:

1. Find out what the relevant stack sizes are.
2. Guesstimate a preflop range for villain. Use the available reads/stats you have.
3. Consider your preflop options. Any merit to raising? Should you be calling? Should you fold? Take into account his style of play on the flop when you flat-call; some hands do better or worse versus certain players. Do not skip this step! You may think you have an "easy" decision, but leaks are - per definition - things you don't know you're doing wrong. Make sure this isn't one of them.

For each of the three postflop streets, do the following:

1. With the current board, what is your opponent likely to do with the range that you gave him preflop? Don't let what he actually did cloud your judgment, play against his range. Is he likely to c-bet this flop? With what? Is he likely to call a donkbet? Raise a donkbet? Call a check-raise?
2. Consider ALL possible lines. Donking, check/calling, check/raising, check/folding based on whatever you thought was likely in step 4. If you think he's likely to raise a donk-bet, is this a good outcome for you? With the range you think he'll c-bet, are you certain you want to raise? What will you accomplish with the different lines you are considering? Do not forget to take into account stack sizes and what you should do if you get re-raised or shoved on. Will stack sizes be awkward for raising? Or are they ideal?
3. Experiment with different bet-sizes and raise-sizes. What will betting $X instead of $Y do in terms of his continuation range, your fold equity versus the range you think he has, and committing him/leaving yourself fold equity for the river?

If you're in an observant mood today, you will have noticed that I haven't given the decision that you thought was "tricky" to begin with any particular place here. This is because

a) You'll get to it eventually anyway, and
b) There's some chance that the reason you had a difficult decision somewhere in the hand originally is because you made mistakes on earlier streets. Perhaps the decision you thought was difficult isn't actually hard at all once you look at the hand as a whole.

This may look like a lot to take in and like it would take a really long time to figure a lot of this stuff out. It is, and it will - in the beginning. After doing it a couple of hundred times (yes, really) you'll be able to do some of the steps quickly, be able to get to the core of the problem fast and know what to look for. With accumulated experience you can rule out a lot of lines very quickly and that will save time in doing future analyses. But taking that short-cut when you're new at it is not a good idea. You may potentially find a whole heap of leaks just in your preflop game if you actually take the time to look at it. Yes, really.

Another thing that is so very very important is that you're playing the role of a scientist. That's to say, your job is this:

Consider all available data and reach the best conclusion.

Not this:

Have a conclusion and look for data that supports it.

If that sounds obvious to you, good. But it's a lot easier said than done. For everyone. Including me.

------

Now, to soften things up a little (hah!), let's practise. For ease, I'll use one of the hands I marked from today's session as an example, chosen because I have decisions on all four streets.

This is a wall of text, and I know that. But this really is what I do when I analyze my own hands, and the only comfort I can give you is that it takes a LOT more time to write all this down than it does to just perform the steps yourself.

View hand 545352
The Official DeucesCracked.com Hand History Converter

UTG: $200.00
CO: $210.00
BTN: $200.00
SB: $39.00
Hero (BB): $421.40

Pre Flop: ($3.00) Hero is BB with T 9
UTG raises to $6, 3 folds, Hero calls $4

Flop: ($13.00) T 7 9 (2 players)
Hero checks, UTG bets $11, Hero raises to $28, UTG calls $17

Turn: ($69.00) 5 (2 players)
Hero bets $50, UTG calls $50

River: ($169.00) 8 (2 players)
Hero...

1. Effective stacks are 100bb. Nice and easy.

2. Standard 20/18 reg opens UTG in a 5-handed game. HEM says his average opening range is 21% in that position. The table as a whole is pretty nitty, but there's an aggressive shortstacker in the small blind. I can't really say how these things will affect his range, if at all, so I'll just take the 21% at face value and guess that his opening range is (using PokerStove for this):

22+,A2s+,KTs+,Q9s+,JTs,T9s,98s,87s,76s,A9o+,KJo+,Q Jo

3. Folding is definitely an option. Suited connectors don't play well out of position.

He folds 63% to 3-bets, too, so raising is probably profitable, but he also 4-bets an insane amount (4.3%) and that would suck if it happened.

Are there reasons to flat? Well, he c-bets 95% of flops, which is good for a hand that can flop something big (because of implied odds) but he doesn't fold versus raises often (40%), which is a little bit of a problem for T9s since I probably want to check-raise draws and I'd love a bit more fold-equity. Not an ideal opponent to play T9s versus (as opposed to all pocket pairs because high cbet% + low fold vs. raise is awesome when you hit a set). I think flatting is +EV, so I approve of my original decision.

For the flop

He does c-bet an awful lot, but this is also the text-book example of a board that "should be checked back for pot-control." I think he's less likely than average to c-bet here, but probably still c-bets very wide. All flush and straight draws, all sets, all two-pair combos, all overpairs, for sure. I'm not sure what he'd do with a hand like AT, but I'm going to go ahead and assume that he c-bets that too (wrong play versus me specifically, but he probably doesn't know that).

Is there merit to check/calling? What reasons would there to be to check/call with top-two in general? I can think of the following:

1. My opponent is unlikely to call a raise with worse.
2. My opponent is unlikely to outdraw me.
3. The turn card is unlikely to kill my action.
4. The turn card is decently likely to improve his hand to second-best and give me a chance to win more money.

None of these apply, with the weak exception of #4. AK/AQ are such a small part of his total range (especially if we remove AsKs and AsQs) that calling in the hopes of him spiking TPTK is a bit of a long shot, and hoping that he'll somehow stack off when that happens is even more so.

And check/raising?

Now I have to guess a little at his c-betting range and then guess a little about his continuation range when I raise. My guess is that versus a raise to $28, he'll not fold with the following hands:

77+,AsKs,AsQs,AsJs,ATs-A7s,As6s,As5s,As4s,As3s,As2s,KTs+,Q9s+,JTs,T9s,98s ,87s,76s,AKo,ATo-A9o,KJo+,QJo

That's a whole crapload of hands. And my equity versus that range (says PokerStove) is a whopping 70%. Fistpump!

Alright, we can clearly skip check/folding.

Donking? The argument for it could be found if he's the type who raises a lot postflop rather than just flat, because he c-bets so often that we are pretty sure we'll get to check-raise. This particular player has never raised a flop in the 500 hands I've played against him, so I'm going to rule this out pretty quickly as well, but it's good to know what to look for.

Is there merit to check-raising to a different amount? Smaller? I think that would be a mistake because raising smaller prices him in way too much with his draws (assuming he doesn't re-raise those; hard to say). What about bigger? There are three things to consider for a bigger bet:

1. Will that affect his continuation range? Probably, if we make it upwards of $36. With the $28 he might feel he can continue with 76 and A7, say, but might be a little more tempted to fold those if I go bigger. Then again, those hands have a decent chunk of equity versus me anyway, so it's not a huge loss.
2. Will the difference in SPR change things if he re-raises? In this case, no. I'm committed as it is and this isn't really the kind of board that he's likely to 3-bet/fold, so sizing my raise small enough to give him rope to pull that stunt is probably not worth anything to me.
3. Will a bigger bet make it easier for me to get stacks in by the river? Only marginally so, I think. If he calls the flop raise, I don't have to bet anywhere close to pot on the turn and the river to stack him when he has an overpair even with the smallish raise. This is probably a non-issue here.

Conclusion: I should probably have raised bigger, because I think his calling range remains unchanged up to at least $32. 2bb may not seem like much, but this is where you go from a decent win-rate to an awesome one, in getting these decisions right. Here, I made a mistake. Lesson learned.

For the turn
The turn card is a great one for me. The only hand in his range that was helped by that was 86s, if it was even in there to begin with (it wasn't in my original range, but it well may have been) and that would be a very small part of his already wide range. I'm not worried.

Any reason to check-raise?

No, I think check-raising here is a huge mistake. I will often end up giving free cards to hands that have outs versus me, and the worse hands that will bet/bluff if I check to them are typically hands that will call (or even bluffraise) if I just bet myself, with the exception of gutshots like KJ that may fold if I bet but might plausibly try to bluff if I check, but I think he's unlikely to bluff me in general because he can't expect me to fold often. The important point here is that his range contains exactly no hands that are drawing completely dead against me unless he somehow called the flop raise with AK/AQ. Inducing bluffs is a trick best used when villain can have air.

No other lines make any sense whatsoever, so I have to bet. What about sizing?

Here is where things get a little dicey, because I'm trying to balance two separate parts of his range on the edge of a knife:

1. His draws, that will call a turn bet but not a river bet unimproved, and
2. His made hands (e.g. AT, overpairs) that I want to keep in but at the same time bet an amount sufficient to make him committed on almost any river.

Betting big extracts the most from category 1, betting smaller (~$40) extracts more from his bluffcatchers. Remember, if he has AT, he can easily put me on a draw. I don't want to bully him out, although I don't really know if that's even possible. To break this down for me, I use Pokerazor, and find out that of the hands that I'm ahead of, the distribution is like this:

1. Overpairs + top pairs: 30%
2. Second pair: 12%
3. Third pair: 10%
4. Gutshots: 20%
5. OESD: 32%
6. Flushdraws: 10%

(To those of you who notice that it sums up to more than 100%: It's because it sums up made hands and draws separately. Some made hands, e.g. JT, is simultaneously top pair and a gutshot. It's good enough for our purposes, though)

What I'm learning from this is that his range is weighted somewhat more towards draws than towards made hands. For that reason, I may want to bet slightly bigger on the turn in order to squeeze out what I can before the river kills my action. What's a good amount? I'd say $50. It gives him the illusion that he may have some fold equity for those rare times when he decides to bluff-shove (an awesome outcome for me) but it's also not so small that I'm offering him direct odds to outdraw me. This is really thin ice when it comes to guessing, and perhaps $52 would be better, or even $55. At this point, the SPR is meaningless since the pot will be a lot bigger than the remaining effective stack on the river anyway and I'm just looking to not fold out his bluff-catchers.

On the river
Here is the raison d'etre for this hand being marked. Should I check/call or shove?

Pokerazor tells me that at this point, a whole 35% of his range is now a straight (and 9.2% a set). Overpairs and top pairs constitute only 27%, 11.5% are second pair, and another 30% are just overcards. I'm committed, so I'm obviously not check/folding. Or am I? And he's never folding a better hand if I bet. Is he likely to check back his overcards or will he desperately try to bluff them? That's the key here, because I think it's unlikely that he'll bet any overpair/top pair if I check to him. I can only assume that he's not going to fold overpairs getting $285:$116 if I bet.

So my problem is that when I bet, his calling range will have me beat more often than not unless he makes a hero call with second pair. I think this is unlikely. But here's the kicker: When I bet and he bluffcatches with top-pair/overpairs and beats me with straights, I lose less than I do when I check/call if he only bets his straights when checked to. I invest the same amount of money in both cases, but my equity is better when I bet.

Can I check/fold? If I think I'm losing money on average when I bet, and I think it's unlikely that he'll bet a weaker hand when checked to, is check/folding unreasonable? It's a 0EV play, so in order for check/calling to be better, he'd have to bet a weaker hand some of the time. How often?

I'm getting ~2.4:1 (rounded down because I want to be profitable, not slightly in the red) and that is equal to needing 30% equity when he bets. Differently put, he'd need to bet his air (which is 30% of his range) about half the time in order to make up for the times that he bets his straights.

I still have no idea what the best play is. I think it's close between check/calling and check/folding, but the good news is this: Before analyzing this hand (that is, before writing this post) I thought I had screwed up the river when I checked instead of betting, because I was convinced that betting was actually the right play. So whether or not I get the final decision right, I'm at least glad to have learned what the right second-to-last decision was.

Wall of text ITT.
But since its FPs its awsome so i actualy read it, A+++ as usual.
It amazing how deep your thought process is at times.

Great post FP and this sure looks like a lot of work (guess that's why you're playing the stakes you are and I'm not).

Do have a question on the analysis though:

You say, "When I bet and he bluffcatches with top-pair/overpairs and beats me with straights, I lose less than I do when I check/call if he only bets his straights when checked to. I invest the same amount of money in both cases, but my equity is better when I bet."

So why do you think either ch/c or ch/f is better than betting the river? I'm sure I'm missing something but after reading through it a few times I still don't get it.

The caveat lies in "only bets his straights." IF that's true (big if) and if he WILL call with some bluffcatchers when I bet (small if), then it's necessarily true that I do better versus

straights + bluffcatchers

than I do versus

straights

The decision to finally check instead of betting after all hinges on the fact that betting flat out isn't profitable. I should never take a -EV line when folding is an option. That said, it's not necessarily so that check/folding is the best option, because I could well be wrong about the fact that he will never bluff the river (or for that matter value-bet with worse). He needs to do that 30% of the time for me to show a profit from check/calling, and whether or not he would is anyone's guess.

(But if I'm hell-bent on showdown and I think he WILL only bet his straights when checked to, I would lose a little less money by shoving than I would by check/calling, yes - I can see why that might be confusing)

Oh god, I almost forgot to say this:

Checking doesn't necessarily mean check/calling or check/folding, either. It will pretty often mean just seeing a showdown. And that is obviously hugely +EV for us.

Thanks, that makes sense. I wasn't thinking about the times when he checks behind and we see a SD and I couldn't see him betting with worse more than 30% of the time.

Jeez, I'm being really unclear about what I mean here. Let me try again:

Betting can never be a losing play in the sense that it will be -EV. I shouldn't have said "it's not profitable" because I can't actually lose money on this river. That was clumsily phrased. What I meant was that betting can't be better than checking because the range that calls us has us beat on average. When that's the case, betting is never better than checking.

Often (half the time?) when we check, he will check back and we will win the pot very very often. Nice big pot. Whether or not to call or fold once he bets, however, is the tricky part. So checking will always be profitable (and always better than betting), but check/calling might be worse than betting and check/folding might be worse than betting. The profit in checking comes from the cheap showdown, not necessarily from what we do once he bets.

Hope that clears it up. Sorry about the confusing way I phrased it.

Just for shits and giggles, I'll mention that he shoved KTs when I checked to him. That's just so bad that I don't really know how to explain just how bad that is.

But yes, I would never have expected him to do that, and I won't particularly let this one super-bad move make me inclined to change my analysis.

Yeah, sounds like something I'd do when tilting. Take a hand with some SD value and turn it into a horrible bluff.

Sorry I've kind of derailed the point of the thread (how to analyze a hand) with questions on the actual analysis. It really was nice to see how thorough you are with your reviews.

Do you prefer to do your reviews after a session or before you start your next session?

No particular preference. Somewhere in between sessions; usually after (because when I have time to play I like to get into it right away) but not for any reason other than how I allocate my time. I also tend to talk hands through with Taylor to check my thinking, which is of course also a nearly invaluable asset to have (poker-buddy in general, Taylor in particular).

By the way, and I didn't mention it in the original thread because this isn't really a HA thread, but you're (knowingly or not) making a very important point here that I think a lot of people tend to overlook when they analyze hands: People spazz. Not often, but it happens. And so when we check, we shouldn't overestimate how often he will do something really stupid, but we most certainly shouldn't assume it's never because that's even worse. The spazz/tilt/bad play factor is in fact so important that leaving it out of your analysis completely will often lead you to the wrong decision. In this particular hand, I didn't get a chance to worry about him spazzing because I was willing to get it in at any point before the river, but assigning, say, a minimum 5% chance of your opponent doing something extremely stupid in any hand at any point is probably not a bad idea.

i just was wondering, if we're willing to call a shove in this situation how come you didn't bet? Also, what was it exactly that made you feel that calling the shove was right(for example did he have a low WTSD or what?). Thanks!

I think this was answered in an earlier post

He never said if he thought calling was right, just that he did.

So if you were to list the options by precedence of profit it would be:
1: checking- him checking behind
2: betting(shoving)-him either folding or calling
3: checking then calling his shove
4: checking and folding to his shove
Thats a question I think.

Outside of the thread derailing illustration- I really like what you have to say about self hand analysis. I've just (yes literally just) been using that to look at some hands of mine from a recent session to figure out what could have gone better/identified a couple leaks- turns out a few "standard" things I do preflop might not need to be so standard.

Its interesting to see this post because it couples very well with Professional No-Limit Hold'em I have been reading just very recently- or at lease it does in my brain.

Thanks!

I am going to have this tattooed on my forearm. An invaluable post.

Wow, read it 3x already, a lot to absorb, but I am sponging. I play - at the moment - only STT, but am starting to do some serious hand analysis. I did bad in almost all aspects of my playing, so, on advice from a lot of players better than myself (not such a big mountain to climb, but hey:-) I am concentrating on only one game for a while, then I will see where I go next. I am assuming the principles you outlined above can benefit my STT HA as well. And on another note, I have read a lot of your threads, if my playing is improving at al, you are one of the people responsible, so ty 4 that!!

Any recommendations on how many hands to gather data on before doing any serious analysis? I have about 4200 hands gathered from .02/.05 and .05/.10, full ring...was going to try to get to about 10k hands total before i went and checked things out. But it seems the analysis is fairly in depth...I imagined a more streamlined sort of data mining procedure

I'm both sad and happy there isn't I suppose

wow great post, will be reading over again later.

thank you very much, i know i need to pick apart my game more often

much appreciated

TY for this post. It widened my view on how to analyze.

Brilliant post. I shall be attempting to do this myself. Thank you FP.