Please comment on this. I am working on a method to classify flops as either wet or dry. Everything that I have read on the subject is somewhat subjective. I am trying to find an accurate method that is not subjective.

1. Sort the cards in order, high to low. So 7cKs9c becomes Ks9c7c. The 19,600 possible flops are reduced to 10 buckets.

HHH HHM HHL HMM HML HLL MMM MML MLL LLL

2. For each of the 10 possible Flop buckets, look for the following:
Ace High - Any Ace.
Pair - Any 3 cards the same rank.
Gap 0 - Connected cards. 8c7d.
Gap 1 - One gap between cards. 7C9c.
Rainbow - No cards of the same suit.
Three flush.

For the gaps, I add points if it can result in an OESD. Also for combinations such as a gap 0 and a gap 1. I use a simple scoring algorithm.

Some of the initial results, with errors, are:

AhKcTh non 1 HHH A G0 4 0
Kh7c3h dry 2 HML 0 0
8d6c5s dry 3 MML R G0 6 0
Ac8c7c wet 4 HMM A F3 G0 4 0
QhQs3c dry 5 HHL R P 0 0
8h7h6h wet 6 MMM F3 G0 12 0
9s7s5c wet 7 MML G1 0 4
AhQc2h wet 8 HHL A G0 4 0
KdTs9c wet 9 HHM R G0 9 0
Qs9c7s non 10 HMM 0 5
Th5h2s non 11 HLL 0 0
Qs9c2d dry 12 HML R 0 0

How does this look as a starting point? Which are right and which are wrong?

If there is interest, I will post the algorithms when testing is complete.

865r is "dry", when any 7 has an OESD? (You have 975tt as wet).
Monotone A87ccc is "wet", when the only unmade hand with high equity must contain the king of clubs? (Similarly, you have 876m as 'wet', when 70% of most ranges hate that board because they completely whiffed it and could be drawing dead even if they have some kind of connection).

As I think I mentioned in the other thread, I think the search for flop categorizations is somewhat fruitless. Even if you could categorize flops numerically, they would still have completely different solutions for the ranges that see those flops. e.g. 765tt plays completely different in a BTNvBB pot, to how it plays UTGvMP, or in a 3b or 4b pot by anyone.

Thanks for the input. Pointing out why a hand is either wet or dry, and the logic behind it, helps everyone.

Board texture is valuable in bluff raising opponents who bet flops when they are unlikely to have a hand that they think is worth defending.

For me, Dry / Wet boards are a step on the way to Static / Dynamic, then to board analysis. The end goal is figuring out things like "Is this board friendly or unfriendly a given board is to barreling and bluffing?".

One aspect is the probability that your opponent will have Broadway cards in his hand, pairs, or suited connectors. A8 is far more likely than J8 , because J8 will generally be folded preflop.

I saw a post in Poker Theory claiming approximately 187 types of flops, followed by "might as well be Rainman" jokes. I think it's an interesting topic....if the system is beautiful enough it might not be too difficult to navigate. E.g. the Stove preflop chart has 169 squares but is no more difficult to visualize than a plotted number line.

There's a natural ordering to hole cards, not sure there's such an easy way with flops.

there are 22100 unique flops and 1755 strategically different flops. best to delete wet and dry from your poker vocabulary.

When I am comfortable with the wet / dry classification algorithm, my plan is to then match that classification to how often opponents hit the flop. The classification, taken alone, is not that useful. But a measure of how often opponents hit the flop would be very valuable.

I think you're still missing the point that each range hits the flop differently. Someone flatting in MP has a completely different range than someone defending the BB, so a good flop for MP could be a pretty terrible flop for BB, whether you describe it as "wet", "dry", or somewhere in between.

What I am initially trying is to relate wet / dry to made hands and draws. I will calculate the percentage of the time that hands like a Flush draw are made on the 3 types of boards to see if there is a strong correlation of how often an opponent hits the flop.

The step after that will be relatively simple, expand the correlation to position. I don't know about specific position vs. specific position, too many combinations.

I hope to post initial results by tomorrow.

I remember an old application that never became to popular. I liked messing around with it. It would be great to use today for organizing flops for pio runs. Anyway I finally found the link. I remember it using some crazy matrix to visualize the flops. It is a Range vs Range of flops app. I only messed with the trial, if it had a way to export in pio format with probabilities I would probably buy it, but I doubt the coder is even around anymore.

http://flopinspector.com/

Thanks!!!! I just downloaded the application and evaluate.

wet, dry, mono, flushdraw, rainbow

I find this discussion interesting because I've been experimenting with different ways to rank or examine different hands or situations. Even if you don't get exactly the results that you hope for, digging into things this deeply can be useful, both for what you learn and for the discipline of studying.

I was recently thinking about which hands are better (have an equity edge) against other hands. I started messing around with an equity calculator and I put hands into five groups, from all hands better than a random hand all the way up to all hands better than five random hands.

When I looked at what I came up with I saw right away that there were some very useful patterns in that chart, for example:

Any ace, king, or suited queen is better than a random hand.

Any suited connector 65s or better has an edge against as many as five random hands. The same is true for 76s against four random hands and 87s against three random hands.

At first I wasn't sure what to do with all of that information, then a couple days later, it hit me--final table play! If I'm the short stack at a final table with four opponents, I now know that hands like Q4s, K7o, and T9o, along with many suited connectors, might be reasonable hands to shove.

In many cases my equity edge will be quite small. I'm not worried about that because I'm a tournament specialist (I have never played a cash game.) Playing over a thousand tournaments a year, those small edges will add up.

Studying a situation as deeply as OP is can only lead to good things, even if his classification system doesn't work out the way he planned.

seems like it would be messy for a human to do. card rank is also something to think about but this is where i'd start.

I am having a lot of trouble trying to simplify classification. I have written code that does it pretty accurately, but have failed at simplifying it. I have not given up yet. Thanks for a helpful post.

What you guys forget, it does not make sense to distiguish flop in terms of how ranges would hit the board because then indeed it makes no sense to have terms of "wet" and "dry. I would spare out the differentation between heavy and light boards.

But if you solely account for the possibilities of texture shifts in general your approach makes sense. I would not make a difference between highcards or lowcards MMML or ALLL.
That problem is already solved by Tom Chamber regarding "Advanced Theory for PLO". The tables are very complex and not very helpful for simplified studies. An approach I had for plo though is here:

https://forumserver.twoplustwo.com/1.../#post54555779

AhKcTh non 1 HHH A G0 4 0
- 3+
Kh7c3h dry 2 HML 0 0
- 1+
8d6c5s dry 3 MML R G0 6 0
- 0
Ac8c7c wet 4 HMM A F3 G0 4 0
- 1+
QhQs3c dry 5 HHL R P 0 0
- 1
8h7h6h wet 6 MMM F3 G0 12 0
- 1-2+
9s7s5c wet 7 MML G1 0 4
- 0+
AhQc2h wet 8 HHL A G0 4 0
- 2+
KdTs9c wet 9 HHM R G0 9 0
- 3
Qs9c7s non 10 HMM 0 5
- 2+
Th5h2s non 11 HLL 0 0
- 1+
Qs9c2d dry 12 HML R 0 0
- 2

If your starting hand is usually suited, like in PLO, a suited flop hits more often and you can usually add +1. A board like 568s (1) and maybe 678ss (1-2) may not be very wet heads up but they have other dangers.