I'm a recreational player but wanted some feedback on some numbers that I've been looking at over the last ~13k hands re AKs and AQs.

So I've been dealt those hands a total of 89 times (AKs-48, AQs-41) and, if removing two donk plays, these hands are only marginally profitable (at best). Looking through the hand history, I also see that I only completed a flush exactly one time, and I've hit the flop hard (two matching suited cards)only twice with one going to the river without completing the flush and the other hand I took down on the turn.

How aggressively should these hands be played? Based on my experience, it doesn't seem profitable to play these hands aggressively.

As you probably know, AKs and AQs are generally rated in the top 3%-4% by most rating systems. The probability of a flopping a flush draw or flush with those hands is 9.6% so out of 89 hands you should have gotten 8 or 9 draws or flushes. The chance of getting 2 or less is about 0.007. Either you have been extremely unlucky or the data is unreliable.

I’ll let other discuss strategy but I will say this. If you have a flush draw on the turn with those hands, an EV implied odds analysis I did showed you generally need implied odds of less than 10 to 1 to call a turn bet that is -EV at that point. The model I used includes reverse implied odds and the possiblility villain doesn’t call a river bet if you hit.

Say, the turn pot is 10 and villain makes a half pot bet. You need equity of 25% to make the call and the draw only gives you about 20%. If you hit the river flush and if villain will call your bet, say 50% on the time, and if your flush wins 80% of the time, you need implied odds of 6.5 to 1 meaning you should call the turn if you can make a bet of least 18.

Details on the model applied to flush draws can be found here:

https://holdemmathology.tumblr.com/p...sh-draw-part-3

dont discount the times it is checked down from there an you win with ace high. and times you make a pair and win. or are able to bluff out the winning hand on the river.

Valid point. The model I used assumes that if you don’t hit, you check/fold.

Including the case of a check down is relatively easy. Including the case of hero betting if he doesn’t hit (either bluffing or having some value) is quite a bit more complicated adding four more variables, bet probability, bet amount, win probability and villain call probability.

~13K hands is a LOL sample size for this kind of analysis ("AQs is bad"). Sorry but online is like that. Data converges better the more hands you play. Try 200K hands.

Many thanks, statmanhal and Ray Zee! Appreciate the feedback.

To expand a bit, AQs is +1.47bb on my database but if I look back to some previous months' play with this hand, it looks like this:

Hands bb/100
328 +25.98
255 -23.92
156 -20.26
203 +290.64
124 +116.13

So you can see that 40 hands with AQs means nothing, when 200 or 300 hand samples are up and down tens or even hundreds of bbs. Overall +1.5bb EV is a massive hand but it takes a huge sample size to converge.

The hands that just crush, giving you a win rate that you can observe really quickly, are exactly AA and KK. Other than that, other hands that still are very valuable, like JJ and AKo, can show substantial losses over a surprisingly large sample of hands.

+1 to what WereBeer said. I can't be bothered to open my database right now, but while AKs has been my fourth most profitable hand over the course of several years (slightly better than jacks, slightly worse than QQ, but way behind AA and KK which are the only real premiums), I've had 50k samples where it was breakeven or even losing. AQs is significantly worse, (and AQo is worse still) but all of the big aces are still +EV hands in the long run. (If you play them correctly and don't get ridiculously unlucky forever).

P.S. To answer the other question in the OP, you should usually/almost always play AKs/AQs aggressively pre-flop, but that doesn't mean you have to keep piling money in post. On some flops that are bad for AKs/AQs, you'll be putting money in the middle with 9-high instead, if 9-high has more equity, for example.

most times those that have bad runs with the best looking hands are playing them poorly in some way post flop.
they may play them by what they think is correct but that type of play may be grossly wrong for particular opponents.

i played a hand last night with this lady i have played with many times and i had opened for 3 times the bb and most went out and she raised to 5 times the bb and i folded my 2 kings. and of course she showed her aces to the table.

^

Too shallow to set-mine?

Is this correct? With AK/AQ you have two (or at the very least one) over with a very good kicker that will likely give you a winner if it hits.
So I'd put the equity in the 25-30% range. (Let's say 25% discounting for opponent already having two pair or better)

yep she was short but i was making a point. that a normal way to play a hand can be totally wrong against different players, and most players i play with, play about the same against all players. although will be more careful against very tight ones but will still get beaten when they shouldnt have been in the pot in the first place.

Say, the turn pot is 10 and villain makes a half pot bet. You need equity of 25% to make the call and the draw only gives you about 20%.

The 20% figure was in reference to hitting the flush draw. As Ray pointed out and I agreed, having AK,AQ can also win other ways but I did not include that possibility in doing the implied odds analysis.

Biggest leak with these hands is mindlessly C-betting with them, especially out of position.