Taking Frequency Notes

Poker Discussion
1

It’s hard to know how other people take notes, and I haven’t seen too much on this in poker literature that I like very much. I used to play a lot of on-line poker with a HUD, and started to wonder when playing on this site how best to get my notes to replace the incredible frequency data on your opponents that a HUD provides, and so thought I’d share 3 types of notes that I frequently take.

  • Pre flop jam notes: Here I just list the hands I’ve seen an opponent go allin preflop with. This is going to be most useful at the lower levels, where this happens more, but it helps me decide on the range I will call these bets with, and if I might make the same move, expecting someone else at the table to call with an overly wide range. Here are those notes from an actual player (a fish that plays on the high stakes tables frequently): 98, 54, KT, T5s, 44, KT, J9 allin
  • Pre flop frequencies by position: first number is seat, second is raises, then limps, then folds, so the player below (a top 100 player) is moderately tight, and limps more often than he raises, and is much tighter about raising from early position than from late. A number after a star is a cold call of a prior raise. A number in front followed by a star would be a 3 bet (this player has none out of well over 100 hands recorded, and so if he 3 bets you’d typically want a very strong range to 4 bet)
    B 0/1/0
    S 3/8/4 *1
    0 3/14/9
    1 3/6/14 *2
    2 4/6/13 *2
    3 1/12/26 *3
    4 2/15/14
    5 1/11/14
    6 0/5/10
  • Bluff/Value notes: I like to keep track of bets that have been shown down by street, and classify them as value, thin value, or bluffs. I also group the bets by size, to try and look for sizing patterns. Those notes for another strong player:
    R o1/0/3 p1/0/0 m5/1/0
    T o2/1/2 s1/1/2 m4/1/0
    F o1/0/0 r0/2/3 m1/0/1

“o” stands for overbet, “r” for raise over prior bet, “p” for pot sized bet, “n” for normal sized bet, “h” for half pot bet, “s” for small bet, and “m” for min bet. So on the river, I’ve seen this player make a number of overbet bluffs, while his min bets on the river and turn have usually turned out to be strong value (a bit of a rare pattern, especially among strong players)

1 Like
2

LOL, a fun game here might be: guess who the player is by their preflop frequencies and pot size by street frequencies. It is interesting how for many players this is almost like a finger print.

2 Likes
3

Most players on this site have high preflop frequencies, but it is difficult to assign them a range, because this consists mostly of limping in. Generally, these limp ranges are drawing hands and small pairs, but I have seen plenty of instances of limping with strong hands such AK and 10+.

For example, I recently played a hand where a player limped in with AA when I had Q3 off in the BB, and the flop came off K33. A preflop raise would have gotten me to fold the hand (no big blind defense for me), but instead I ended up getting my opponent’s stack. And people around here complain (including on this particular hand) when their aces get cracked – it’s because you gave your opponent a free opportunity to beat you.

Of course, I have gotten myself into trouble into these situations as well, not seeing the hidden monster and losing more than a few chips in the process. But I can tolerate these kinds of losses since I think that it is generally +EV to treat limping in as a weak play. It’s called “limping” for a reason.

I will never understand the high frequency of limping in, but it does create some difficulty. For example, say your are in middle position with a decent drawing hand (e.g., suited connectors 56-9T or small pairs 22-55), and 3-4 players have limped in and another 4-5 have yet to act. What do you do? In part, the answer to this question depends on your opponents’ preflop and call frequencies, but when facing broad preflop ranges (i.e., uncertainty), the answer is not always straightforward.

I know that a lot of players like to overlimp in this situation, but for me that tends to be a losing proposition. So I am faced with two choices: fold or put in a bet large enough (usually needs to be at least 6-7x bb) to thin the field to a manageable level.

On one hand, consistently pulling these hands from my range will decrease my frequencies to a point where I am an easy mark to read. On the other hand, consistently betting these hands tends to be -EV in multi-handed pots. There are a lot of players who will limp-call these bets, and strong players will often do so especially after other less strong players have already called, based on their pot/implied odds. This obviously makes post-flop play decidedly hard to play, and the aforementioned non-premium hands rarely hold up.

Most observant players won’t have too much trouble guessing what kind of player I am based on my frequencies, but I don’t worry too much about playing a transparent game, which is specifically adapted to the majority of the field on Replay.

3 Likes
4

Later in your post you talk about “broad ranges”, and I think that is the better way to describe what you are faced with when players limp with a high frequency. It is not that it is hard to assign a range, but rather that it is hard to assign a small range (since they don’t have a small range).

If you are always thinking in terms of ranges, the size of the range changes nothing: you are still evaluating how many combinations you are ahead of versus how many combinations you are behind to. Now, I’ll acknowledge that different players with the same limp percentage may have different cards in their range, but that is a problem true of both small and large ranges, and you have to watch the cards your opponent shows down to try and improve the fit as best you can (it will never be perfect… often the players themselves don’t know what their ranges are, lol).

With small pairs or smallish suited connectors and with limpers in front of me, on a passive table I generally prefer to limp in, but sprinkle in some raises. On an aggressive table with lots of raises and 3 bets, I prefer to usually fold or raise, though I’ll still limp here some frequency of the time. I just find I make money in those situations, since evaluating where you stand post flop tends to be relatively easy, and those hands perform quite well in multi-way pots. That said, I don’t think you are losing too much equity just folding those hands, especially as you play higher and higher stakes, or if stacks are not deep. You’re not really narrowing your range much though, in that your raising range is not really impacted by your limping frequency. Two players that have the same raising percentage, but one has no limps and the other limps a ton… well, the raising range for both is still likely quite similar.

By the way, I don’t think your game is quite as transparent as you suggest, LOL. I’ve been surprised by hands you’ve shown down on quite a few occasions. :astonished:

1 Like
5

Your note-taking strategy is very interesting, marking players’ frequencies by position and type of play. Maybe if I were more cognizant at the tables, I would use that approach for regular opponents.

I agree with @AKFolds points in general about wide ranges and limping. It applies to note taking because I don’t need notes on most players when my base assumption for 99% of them is that they are passive/limping/calling stations. I only take notes on players who I see frequently, have very high rank, or who make noticeably different plays. I typically note something like “Called OOP with K4s” or “called flop with just a gutshot” or “limped KK”. Plays like these inform range-based decision-making by showing how wide or narrow their range is in a particular situation without having to track a frequency.

2 Likes
6

Just following up on this in a way to hopefully illustrate its power, here are my notes on two players in the top 100. Note too, that if you have players like this in the top 100, just imagine how much more common it will be at lower ranks to find gigantic leaks if you just take the time to note frequencies.

Just to quickly summarize again how the notation works:

  • there are sections for bets on the river (R), turn (T) and flop (F)
  • the first letter in each section denotes bet size: over-bet (o), pot (p), normal (n), half pot (h) and small (s) (and some others I might not need to use, like raise, min-raise, and min)
  • the numbers are strong value / thing value or marginal / bluffs

Player 1:
Most Exploitable Leak: pot and half pot bets on the first two streets are tilted severely toward bluffs. Small bets are strong.
R o1/0/3 r1/2/0 n1 0 0 p7/4/12 h24/6/16 s10/0/2 m2/3/0
T o1/1/0 r0/1/0 p3/2/16 n2/0/1 h8/8/21 s13/4/3 m3/3/1
F o0/2/2 r1/2/3 mr1/3/0 n1/1/0 p2/1/17 h5/4/16 s5/3/3 m2/3/3

Player 2:
Most Exploitable Leak: pot sized river bets include no bluffs
R p12/1/0 h11/10/4 m3/7/3
T r4/1/0 mr0/0/1 p5/3/1 h10/7/2 m2/1/6
F o0/2/0 r2/3/0 mr0/0/2 p1/3/1 h2/1/0 m1/2/2

Hopefully any of you could suggest effective strategies for beating these players, once you know that they have frequency leaks this large.

1 Like
7

One more thought related to this: I find it very valuable to reverse this practice occasionally, taking notes on my own play. It’s surprisingly easy to pick up sizing tells, and taking notes on your own betting, at least occasionally, can help you notice if you are developing frequency leaks of your own.

Obviously there are times you are happy to have “leaks” like this. If you are playing at low stakes against wildly aggressive, out of control players and call stations that almost never fold, you’ll probably want to drastically reduce your bluffing rates. Conversely, if you find yourself against very tight players, you’ll often find you can over-bluff in quite a few spots. That said, there will be some players that are very observant, and against this group it is nice to try and aim for intentional targeted ratios of value to bluffs for different bet sizes, and on different streets.

1 Like
8

Here’s another player in the top 100. What are some of the most profitable exploitative responses?

  • R r1/0/2 n2/0/0 p9/0/8 h10/1/7 s1/0/1 m2/0/0
  • T mr0/0/2 p3/0/3 n2/2/0 h10/9/8 s0/2/1 m1/1/3
  • F r1/1/1 mr0/5/1 p1/0/2 n1/1/0 h3/4/4 m1/0/0
9

Here are some of my thoughts on potential exploits:

  • Bluffing frequencies generally seem likely to be too high on the river (keep in mind that our sample is somewhat small). Ranges are very polarized, and so in general I’m just expanding my calling range, although you can also add some bluff raises with hands that lack showdown value.
  • Min raises on the flop in particular seem to be very marginal value. Given that it seems consistently weak, you can expand your bluffing combinations quite profitably. You’ll also need to call quite wide when you don’t have good bluffing candidates, as you’re being offered fantastic pot odds, and so you don’t actually want to fold all that many combinations.

I find the second point to be a real area where you can see a lot of profit. A long time ago, you could hear a lot of players advocating for this kind of raise, to “find out where they stood”. The thought was that you could reduce your losses with marginal hands through a small “bet for information”. There are a number of problems with this approach, though.

  • the information you get against stronger players is of questionable value
  • combinations of hands that call and beat you typically outnumber combinations of hands that call and are behind - so it doesn’t function well as a value bet
  • combinations of hands that fold and are behind generally outnumber hands that call and are ahead, so it doesn’t work well as a bluff
  • while the worse hands that fold generally surrendered a sizeable amount of equity, this typically isn’t as big a factor as the prior bullets
  • if you make this move with only a narrow class of hands, you’re surrendering far more information about your own holding than you were hoping to derive with your “bet for information”