Hmmm... I'm not very enthusiastic.

First, while you can view AA as a group, I think it would also be useful to create some subcategories there--depends a little on how this is supposed to fit into the program. But if one is trying actually to evaluate hands over very large numbers, I think that would be really useful.

There's some overlap between the following subdivisions, but I think these would be pretty good:

1) AAds
2) AAss
3) AAsss or AAr
4) AA with 2 sidecards T or better (allows straighting)
5) AA with other connector or 1-gapper (AA76 or AA75)
6) AA with another pair

One could also do the same thing on JJ-KK, but probably too tedious.

However, I really wouldn't lump JJ in with TT. In my book, JJ is a pair that can stand alone, TT isn't. If you're striving for some degree of universality, however, it might not be bad to make separate categories KK, QQ, JJ, TT and 99. I know some people view only KK and QQ as stand-alone pairs, looser players may view 99 as such. Anyhow, that seems to me the only way to get around variations in the way different players view the quality of these hands.

I'd completely scrap the category 22-88. I don't think these hands are normally worth anything. If you have a category for that, then you should also have a category for anything at all double-suited.

And, by the way, I think a HUGE priority on improvements of pokertracker should be getting suit structure into the categories. On some hands, it doesn't matter a lot, but it's always relevant and on some hands it makes a huge difference (A678r is really pretty junky whereas A678ds is a monster). I think Mike promised something in that regard months ago, and there are filtering possibilities with suit structure, but it's all very cumbersome, and it would be REALLY useful to incorporate it as a column in categories (as well as a column for "includes a suited A").

I don't like your "high wraps" category. AKQJ is a strong hand mainly because of high card value, and it's a very different kind of hand than 89TJ, which has value because of it's straight possibilities with the bonus of some high card value.

Instead, I'd make a category "Big cards," including any 4 where all cards are 9 or better but actually excluding the wrap QJT9, which has more value as wrap than as big card hand (I guess you could include that both in the wrap category and in big cards, but I think it's high card value is noticeably lower).

On the wraps, I don't think there's really much difference between 4567 and 89TJ, and I find it insulting to the former hand (which I view as huge) to put in the same category as 3456, which I'll often fold.

On the gap wraps, I don't think the rank is nearly as important as the location of the hole. In my own categories, I have the premium gap-wraps as Low- and mid-gappers (3567, etc. and 4578, etc.--noting that both of these are the bottom end of having full extension). High-gappers imo are noticeably worse (4568, etc.).

On the pair wraps, I also find it insulting to a strong hand like 7655 to put it in a group with complete junk like 4322.

I'd really just put the pair wraps where the pair is smaller than JJ all into one group, and the only good ones are imo 4456 through TTJQ, 4556-9TTJ and 4566-89TT.

On the pair gap wraps, I think it's critical to have to closure, although I'm open to being proved wrong on that. Anyhow, I think 4467 is still a premium hand but 4457 isn't.

Ace wraps: Well, here's where the suited limitation plays a huge role. Also, if one wants to avoid duplication with the big card hands, you can go A456-A89T, but there's an enormous difference depending on whether or not they have a suited A or not. Also, imo sss (3-suited) makes them noticeably worse and ds, as usual, a little better.

On the A-wraps with a hole, I have the same quibble with location of hole. I still really like A467ss(sA) but am at best unenthusiastic about A457ss(sA).

On the double pairs, I think the connection value is very important: 9922 is maybe in the decent group (which I'm not sure 4422 even makes) but 9988 and 9977 are pretty strong. I'd really go with 4455-99TT (since TTJJ is already in the JJ category, just a particularly nice one), then 4466-88TT as separate categories (or a single separate category) and then other double pairs TT or lower as a different one.

Another nice category might be what one could call suited pairs, like 66AX where the AX is suited and not of the suit of either 6. I think 66AX-TTAX of this type are all moderately strong hands, 22AX-55AX for my taste (pending any substantiating data) less so.

Oh, one other group (whether one group or more) should include the following:

1) 2-hole wraps 2467,8TQK
2) Big-hole-on-bottom wraps 3678-8JQK
3) Big-hole-in-the-middle wraps 4589-89QK.

I'm struggling with a few ideas at the moment, as I don't play Omaha Hi yet so I'm not sure what you want. I'm trying to balance ease of use with level of control. This is what I'm currently using for suits:



s1 is single suited to the highest card in the hand, s2 is single suited to the second highest card, etc. ss2 is double suited to the second highest card. For example the hand selected is s1 because the highest card is suited. So to filter for suits, you can either select an option from the suiting box, or type in "ss" for all double suited hand, "1" for all hands where the highest card is suited, etc. Combined with the hand filters you get a decent amount of control.

Of course, this doesn't allow you to select all suits of a particular rank, for example, all suited tens, nines and eights, regardless of where they occur in the hand. So another possible method is to have a scheme such as "ssA9" to mark a double suited hand to the A and 9. For example, the selected hand above would be "sK". Then if you wanted to only look at all your single suited hands where the suits were 8,9 and T high, you could type "s8,s9,sT -ss", in combination with any hand or other filters you might want to apply.

I'm trying to keep the interface simple, and this will confuse the hell out the idiots, but since only you guys will be using it I guess it won't matter anyway. Monk will just have to learn to cope.

The other issue is, do you want to be able to group multiple categories together? For example, do you want to lump high double suited wraps together with AAxxr? It will significantly complicate the interface, but it won't take long to do.

Anyway, let me know what you want.

Aisthesis - thanks, amazing post. I'm implementing everything on the list.