I'm heading to Vegas to play the $1500 2-7 TD in 2 weeks, so I thought I would try to study the game as much as I could before then.

I want to work out optimal IP an OOP ranges in a 1:1 spot on the river. As I'm not aware of any 2-7 solvers, my friend had an idea of using Piosolver - by setting the board as 22234 and assigning both players ranges as 55-AA, which effectively cancels out the board and leads to the best holecards winning. To us, this seemed like a fairly accurate but hugely simplified way of representing the linear ranges of 2-7 hands.


After getting the results from Pio, I then compiled a list of all the hands both players can have. I assigned both players the exact same range, which consists of most 4card 8's (minus OESDs and the roughest gutshots), and the smoothest 4card 9's. I weighted each hand according to how many ways it can be made from 4card hands i.e 23457 will occur 5 times as much as 2345J, as the wheel can be made from 2345, 2347, 2357, 2457 and 3457. Paired hands like 23477 are worth 0.75 of hands like 2345J, 3457K etc. I didn't go as far as to weight the 4card hands from their respective 3card hands or consider blocking effects to keep it simple. Will these greatly affect my weighted ranges or are they negligible after the going through 2 draws?

By using Pio results with my compiled river ranges, this is part of my suggested strategy in a 5BB pot OOP:

Top 10% (86532+): Mostly bet with mixing in X/R about 1/3 of the time.
10-20% (87643-86532): Mostly bet, mixing in X/R about 1/5 of the time.
20-50% (Q7542-87643) - 60% Bet + 40% X/C
90-100% (87654-66532): Bluff 50%, which for me works out to almost exactly all straights and every 88xxx hand.

Seems reasonable so far, but here's where I'm doubting my calculations a bit.

50-90% (66532-Q7542) - X/C 92.5%, X/F the rest. This means folding everything up to 44xxx and then X/C everything 33xxx - Q7542. This seems quite loose!

IP's strategy when checked to is fairly simple, it bets the top 52% of hands (Q8532+), then X/B everything weaker up to 90% (55xxx), then with the bottom 10% of its range it wants to bet 75% (all straights to 66xxx) and X/B the rest

My main question then - have I made some bad assumptions or is this actually what optimal play might look like?

Wow, without dissecting it too much, a lot of those assumptions seem reasonable - which wouldn't have been my prediction from the thread title alone. Kudos to the creativity and work you put in using Pio for an "off-label" solution.

Some of the assumptions might be off. Betting 52% in position seems too wide. I'm not sure how to account for the frequency that you would be x/r, as I'm not very familiar w Pio.

Bluffcatching 33+ or 22+ seems about standard against optimal playing opponents, though. I don't think it's too loose (though not all opponents will be playing optimal).

I don't see any x/r bluff or bluff raise candidates listed. Those opportunities won't occur often in a 3 day MTT, but might be worth considering, in general.

Thanks. Pio does actually suggest X/R bluff for OOP about 0.5% with hands in the 60-90% portion of our range, and then bluff raises IP about 2.3% with hands in the 50-90% portion. I didn't include these as they are v infrequent and in my limited experience its hard to imagine people in live games bet/folding at all unless I happen to catch them on a pure bluff. I suppose its infrequent use could be an argument for including it in my range as well!

About the high betting % IP, it is actually 59% sorry. I figured its just a function of how often villain is forced to call due to pot odds. I tested it out with different pot sizes starting at 1BB which suggested betting 38%, 3BB which suggested 55%. At 10BB it wants IP to bet 63%, and then OOP has to X/C almost everything in the 50-90% range which includes everything between Q7542-66xxx. A funny one is at 5000BB, IP bets 77% and OOP is forced to call pretty much everything including 88xxx.

This of course assumes I have weighted hands correctly which I'm sure I haven't. I just need to find a way to better approximate how much of each 5card hand I'll have on the river, but I'm scratching my head trying to figure out the best way.

Very cool. Very educational what percentage to bet/check-fold/check-raise.

Your bluffing range could be improved by considering blockers. We see that OOP is calling 33+ and folding 44. Instead of bluffing with the bottom of your range (straights and 66+), IP its better to bluff the same number of 44+ combos that have seen a lot of wheel cards. Likewise, for OOP, instead of bluffing {straights + 88}, you should bluff same # of bottom 10% hands that have seen many wheel blockers.

Yeah I agree. However, assuming we start the hand with 3 wheel cards, draw 2, then draw 1, pair at least once along the way and then pair one of our top 2 cards on the final draw, this will occur about 6% of the time if my maths is correct. This works for OOP who wants to bluff about 5% of the time, but IP wants to bluff about 10% (in a 5BB pot).

Adding all straights and 88xxx the times we dont see any dead wheel cards conveniently brings us up to 10% though which is nice.

Sent from my SM-G935F using Tapatalk

Very cool thanks for posting 👍🏼

Yes, though yesterday I saw someone check/call 33 vs someone bluffing an ace high. I find that amusing...when I see it happen to someone else of course.

In a $1500 tourney might have a lot of people making some big mistakes, so I'd be careful against strictly living in piosolver land. Though this thread is interesting.

I played the $300 event at PH yesterday and had a guy bluff raise with 22xxx in a large pot on the river in 1:1 after I lead a T (I folded and was v shocked to see it). Im playing the 2-7 event today and I dont intend on getting too fancy with x/r mixing and loose bluffcatching (depending on who Im playing against ofc).

When I'm back from Vegas next week, I want to go a bit more indepth with this analysis though, refine my ranges etc. I may also flick Pio over to no limit mode and see what it comes out with for 2-7 NLSD.

Sent from my SM-G935F using Tapatalk

Yeah I play with same people a lot so I observe and note play on all 1/1 with hopes of gathering enough puzzle pieces to play each specific person optimally

But GL today man!

Saw some stuff yesterday I didnt expect to see really, people checking back decent 9s in 1:1 etc. And stuff I was able to make adjusmtents for - nobody ever went for x/r on the river so their checking ranges were almost exclusively bad 9s - mid pairs. And thanks, bagged up about avg for day 2 in a few hours. Not the best table draw though!

Sent from my SM-G935F using Tapatalk

Aaaaand I've made the FT. Got to play with Death Donkey + Negreanu/Chris Ferguson which was cool.

Sent from my SM-G935F using Tapatalk

Awesome! Good luck!!

Oh man wish I woulda put this together last night so we could have talked about this thread (I don’t want to post about it), good luck today. I feel better about folding 22 in that 1:1 spot since you probably bet a reasonable range there. That tournament is so weird in river 1:1, nobody value bets a ten and they bluff all their pairs, so you can exploit by check call / raise everything.

Thanks mate. I was going to mention this at some point actually, but you ended up all-in pretty quick good fold anyway, I had a 97. OOP I'm not betting as wide, not too sure what the bottom of my betting range would be though. I know IP its somewhere round a J, maybe a Q5xxx if the pot is huge (probs too spewy though).

Im still unsure how to adjust to these players, Ive never really played a live 2-7 tournament. If they dont value bet thin enough, it makes me want to lead more OOP to prevent check/backs but then it also makes me want to check more because their range is quite defined and I can save/make extra bets where necessary.

Sent from my SM-G935F using Tapatalk