Hey Guys,

When you enter a hand range in a PokerStove type of application what behavior do you expect?

T8o+

[ ] - T8o, J9o, QTo, KJo, AQo

[ ] - T8o, T9o, J9o, JTo, QTo, QJo, KJo, KQo, AQo, AKo

[ ] - T8o, T8s, T9o, T9s

[ ] - T8o, T9o

[ ] - Other

For the record, PokerStove interprets T8o+ as {T8o, T9o}

My software (Poker Crunch) uses this one for T8o+.

I look forward to hearing what other people find intuitive.

In Poker Crunch, you can get that one by writing "T8o-AQo".

I'm between these two options:
T8o, J9o, QTo, KJo, AQo
T8o, T9o

And I can see reasons for both of them. The first one is looking at the hole cards table by the diagonals (number of gaps between cards). And the second one second one is looking at the hole cards table by rows in the case of the suited cards, and by columns in the case of the unsuited cards.

Holdem Ranger also uses the same interpretation that PokerStove.

But, what would be good to have is a "hole cards specification language" that is consistent across different softwares, and the it should allow us users to define the cards range with a minimum effort and at the same time providing a comprehensive language to define ranges.

I'm definitively against including here the suited cards in any case. Like if you are taking the PokerStove approach that is not by gaps, T8o, T8s, T9o, T9s will have to be acomplished by T8+. In this way the language provides you a way to define the three different ranges T8+, T8s+ and T8o+, otherwise you lose the only offsuit cards range.

And there is some small problem also, that the defacto standard is PokerStove (that seems will not have a 2.0 version), and it has a very limited hand specification language. Holdem Ranger did a good job there.

How about defining a public "common hand specification language"? May be even with a parser reference. That would be cool :-)

I'd probably chose to start from Holdem Ranger because of a few reasons:
1) It is compatible with PokerStove that is the de-facto standard.
2) It did a good job in expanding the hand specification language.
3) It has public source code.

I write some scripts for myself, and I picked J as the cutoff for this type of thing. So J8o+ == J8o,J9o,JTo, while 98o == 98o,T9o,JTo,QJo,KQo,AKo. In practice, that seems like a fairly useful way to do things, but I can understand why you guys do things differently.

T8o+ should definitely be interpreted as [T8o, J9o, QTo, KJo, AQo] and here's why.

32o+, nearly everybody will agree, means every offsuit connector up to AKo:

[32o, 43o, 54o, ..., KQo, AKo]

The behavior of "+" is established when it's a connector: take the initial hand and move each card up by one.

Why should the "+" behave differently when it's a one-gapper or N-gapper? It should still increment both cards by 1.

Now, on the other hand, ATo+ should mean:

[ATo, AJo, AQo, AKo]

because you can't increment the Ace. Therefore it's assumed that the bottom card gets incremented.

What we're really saying with hand range notation like T8o is:

I've got a such-and-such rank two-gapper OR A BETTER TWO-GAPPER. The + should not change the fundamental relationship between the two cards except when the top card is an Ace.

I realize this is not how every piece of software does it, but this seems the most consistent to me.

In Poker Stove, the + sign with unpaired hands always means to increment only the second card up until connected with the first card, and all are either suited or not as specified. It never increments the first card. And with pairs, the + means all higher pairs.

I personally don't prefer that method, but that's how it does it.

When you want to increment types of hands you should specify more information, such as:
"suited 1-gaps 53s+" which would mean ( 53s 64s 75s 86s ...)

but if you only say 53s+ poker stove interprets that as (53s 54s). Since other meanings are ambiguous, you just have to say if you mean to increment the hand type, as in my first example.

It would be good to have a standardized notation. One suggestion would be:
53s+ would mean only increment the second card, i.e. (53s 54s)
(53s)+ would mean increment both cards together, i.e. all 1-gaps.

ops I voted wrongly in the pole (perhaps) - I voted "behaviour I'd like" rather than "expect" (which was all offsuit one-gaps >= T8o)

The dash notation is much easier and more explicit. T8o-AQo is pretty clear, as is T7s-T9s. Even T7s-AKs is easy to understand [T7,T8,T9,J8,J9,JT,...,AQ,AK all suited] the only trouble comes when some bozo asks for 98o-ATs or some silliness like that.