I posted this to the hold'em manager thread, but they said they can't answer strategy. It is my best understanding (from looking at my stats) that if I folded every river (even when I had the nuts) I would still be adding 1 bb to my winrate. I just wondered what a winrate would be for a good player when they flat on the river. Here is my original post:

" I have a question about a leak I have in my game. Basically I have been going through every preset filter and I consistently find that I am crushing in every category (3 bet, cbet, flat flop, etc.) but the only area where I seem to be getting crushed is my river flat, which is almost always losing and by a pretty wide margin. I guess this just means that I need to start flatting rivers tighter? Assuming I can just fold for 0, then it seems I would add 1 bb to my winrate by just folding every river.

What is a normal river flat win rate for a good player? What other stats might suffer if my fold river is higher? (turn flat and flop flat are both big winners, don't know if these will be affected or not)"

The two major mistakes I seen on the river:

players that will bluff draws on the flop and turn, yet lots will just give up with hands that have no chance of winning showdown.

players that will bluff draws on the flop and turn, yet lots will incorrectly construct a bluffing range with the wrong hands from the wrong part of their available bluffing range.

the result is that I consistently face a river bet that is value heavy.

Well you cant start folding 100% of rivers, obviously.

For starters, position matters. Let’s assume oop for discussion. Your oop river check range is too weak, for locating your hand within this range and deciding to call. Likely, your river betting range is too strong as a consequence. Your opponents are folding when you bet, and value-towning you when you check. Move some value from bet range to check range and this will balance the extra bluffs/thin bets that opponents fire at you when you check. Your check range has the potential to be powerful and profitable, often having the nuts at least some of the time. Don’t be too greedy by always betting your best hands oop, you have to let villains bluff or value-own themselves often enough so they respect your checks.

When playing the turn, you should have an idea of where your hand will be located in range on various rivers. Blank rivers, rivers that complete draws, rivers that pair the board, etc.

If you are losing money with calls in position, then this is just stationing and overcalling. Filter for these and figure out what is worth a call and what is not.

-Rob

There's an error in reasoning here: you have more than one alternative to flatting

Also it's possible you can fix it prior to the river - it's possible you're getting to the river too often with hands that aren't good enough to raise but that you don't want to fold

I don't really use HEM - is the stat saying your win rate *overall* is impacted -1bb/100 by your river flatting? Or just that you lose 1bb per 100 river flats? If you only flat the river in 1/100 hands then your actual net loss from river flatting is -.01bb/100 if you see what I mean.

If we assume that no one noticed and adjusted, it shouldn't affect other stats. But, people will notice, and adjust. I would expect people to bet/fold the river against you a lot more.

How do you come to that conclusion? Every time you fold, you lose the hand, along with any money you put into the pot.

The EV of an individual folding decision is zero (you neither add or subtract money from your stack if you fold), but when HEM collects your data it just looks at how much money you had at the start of the hand, and how much you had at the end. You can't end up with a bigger stack than you started with if you fold.

If you want further advice on river play, a useful HEM stat is "river call efficiency". Add that to your HEM report for position. You want it to be >1 if possible, since that would indicate you're breaking even with river calls (based on pot odds), even if you actually have the worst hand the majority of the time. e.g. If villain pots the river, you only need to be "right" once in every three times to get a breakeven RCE of 1.0.
If your RCE is <1, you need to fold (or bluff-raise) more often, because you're apparently a pay-off wizard.

Thanks for the good replies

This is a good point.

I was just trying to indicate that my strategy is so far from optimal that folding 100% of rivers would net me an additional 1bb/100

I agree with the point you made and it's something I will chew on and consider. But combing through the DB it seems like the vast majority of these calls are IP. Mostly bc I try to avoid playing oop as often as possible.

Didn't read the last part until I replied above but this is pretty much spot on.

Good point. Noted.

I think this another good point, but with the way I have constructed my ranges. It seems like this is almost exclusively a river problem.

My overall winrate is impacted by about 1-1.5 BB. I am -352 bb/100 on river calling.

The stat I am looking at only takes into account river decisions. It's just saying that the sum total of times I call on the river is negative while folding every hand would just be 0. The money I lose from a river fold is already accounted for with the other HEM stats, to the best of my knowledge.

Thanks, looking into this now as we speak.

I added RCE and it is 1.12. So does this mean that my hypothesis about folding every river is wrong?

Yes, your hypothesis was wrong. An RCE of 1.12 indicates that for every $1 you put in the pot on the river as a call, you got $1.12 back on average.
Most winning players at lower stakes will have an RCE much higher than that (because they fold more often because #TheyAlwaysHaveIt, so their RCE will be 1.3 or 1.4 or something like that), while fish will have an RCE of less than 1. (They don't win often enough to break even on their river calls).

I'm a bit confused that you said you had a loss rate of -352bb/100 on river calls, but I'd have to look at the filters you were using and check some of the hand histories to look at the size of some of the pots. If the filter was simply "River call = true", then it means that you're losing 3.52bb per hand when HEM looks at all the hands in which you called the river. It sounds bad, but some of the pots will 150bb or more, and you'll have put 75bb in the middle. Losing 3.5bb per hand is better than losing 75bb, obviously. That is to say, calling the river will often mean you end up with fewer chips than you started the hand with, but calling the river with the best hand often enough, is still better than folding every single time. [EDIT: It's probably possible to set a filter in HEM to "fold river = true". It would have very little strategic value, but I imagine the lossrate for that filter is frightening.]
I mean, just imagine you were playing against a fishbot whose strategy was "Call pre, call flop, call turn, fold 100% of rivers". He would be busto after about 10 hands. You have to call some rivers, because sometimes you win, and the amount you win on the river is greater than the amount you lost by calling flop and turn. i.e. calling and being breakeven (ish) is better than folding every single time. Like, suppose you'd put 15bb in the pot on average when the river card appears. If you folded every river, you'd be losing 15bb per hand. If you call sometimes and win, your losses might go to down to 3 or 4bb per hand.
As long as your RCE is >1, you don't have a major problem. In fact, the low RCE of 1.12 would make you a difficult opponent to play against, as you're apparently calling/bluffcatching at near optimal frequencies, unlike the regnits that you can run over, because they fold too many rivers.

Hi Arty, thanks for your advice in this thread. I was able to improve my RCE from 1.12 to 1.52, which has made a big difference in my winrate. I appreciate your posts and hope I continue to learn from and hopefully provide value for you as well.

This doesn't necessary improved your river play.
If you fold everything except the nuts on the river you would have a higher RCE but you would lose a lot in your non showdown winning.
A 1.52 RCE is probably way to high and indicates that you can call more OTR.

The majority of calls and check/calls on the river is bluffcatcher (unless calling is your only option because your opponent is all in). This means that you should be close to breaking even when you call and your RCE would then be 1.00. You want your worst bluffcatcher to be 1.00 if that makes any sense to you.
I would aim to have what you had before (1.12). If you have to high RCE you should probably start checking how you build your range that gets to the river.

Just to make it clear, when I say "A 1.52 RCE is probably way to high and indicates that you can call more OTR" I don't mean that you should just call more hands for the sake of having a more "correct" RCE. Your worst bluffcatcher should still not be under 1.00 and if you can't find any more hands that have over 1.00 then you need to re-build your range that you get to the river with.

This is very true, imo. Don’t go crazy, OP, but there is a reason you are only arriving at river with good value calls or obvious explo folds.

Enjoy the winning, but remember perfect calling is a goal of 0EV. You are defending, not attacking.

Some days I feel like the biggest fishy call station, and those days suck. Some days I feel like a master bluff catcher, picking off anything that does not smell right.

In the end, I am just a difficult player to bluff, and this makes my opponents less profitable on previous streets as well as the river.

Thank you both for the input and I agree with both of what you are saying in general. However, I play lower stakes and the ranges I am running into on the river are not well-balanced at all and my increase in RCE appears to match an increase in winrate as well, so I think I am making better folds in spots where I was previously calling too loose. That being said I will keep this information in my mind as I attempt to move up in stakes, as I think it will be more applicable the higher I get.

My river call eff is ~1.20

Very nice number, Broken, thanks for sharing.

What stakes do you play?

200nl - 1knl

the river can be thought of as completely decoupled mathematically from the rest of the hand, it's where the poker "skill" comes into play the most

you are losing money on the river because you are not as good at poker as your opponents

on the river one player is 100%, the other player is 0%, so if you are calling when ahead more often than folding when behind you win, if not, you lose. the turn is the end of poker theory

in all honesty i think you might benefit from reading sherlock holmes. he considers all possibilities and discards the least likely to find the culprit

start eliminating hands from your opponents range, the one that is left is the one he has

This is a terrible post.

I specialize in terrible posts!

thread shows how you can beat the game using GTO: proper frequencies, pot odds, etc etc

the OP is actually right tho, you are supposed to fold the river 100%, theoretically, if every river was proper GTO

notice there is no "skill" anywhere, if we just play everything GTO from pre flop to river then you should never flat or raise the river, you should only fold when being bet into

that's fine if you want to break even, if you don't, you need "skill", i.e. figuring out who has the best hand, the fact that the OP is losing money flatting is a critical leak in his poker skills

this is exactly correct

a GTO player has EXACTLY 1 bluff combo and 1 value combo in his river leading range, giving the other player a yes or no choice that is completely indifferent as both players make zero money regardless of player B's actions

so what is the GTO response to indifference?

well, calling and folding have zero ev, so the only difference is VARIANCE, and GTO will seek the lowest variance because it assumes a FINITE bankroll, therefore, the OP is correct, GTO folds 100% of rivers

Nope. A villain can play optimally and some of your range will be precisely breakeven (the "bluffcatchers"), some of it loses if you call, and some of it will be profitable, because you'll have some hands that beat the weakest value hands as well as the bluffs.
e.g. Villain can bet a set for value as part of the GTO strategy, but you might have a few combos of straights. Those hands are clearly profitable as calls, since they are better than pure bluffcatchers. Folding 100% of your range would be completely ludicrous, let alone massively exploitable. (If the correct strategy was to always fold, it would mean that your opponent should always bet with ATC.)
It was established earlier in the thread that OP is not losing money with his river calls. His river call efficiency was slightly more than 1.00.

Well the OP is obviously a very intuitive mathematical player, because his thread's premise is correct

Player A can have a mix of bluffs to value ratio but if he is a GTO player he will give the correct pot odds, i.e., tailor his bet sizing to fit the window, in fact he can always give player B an indifferent decision, he can have the exact nuts and min bet into a huge pot for example, maybe his bluffing hands are down to 2% of the time and he's value betting 98% of the time. This is an extreme example but Player A's sizing can make all of his hands indifferent to a call

Player B is indifferent and cannot make a profit no matter what he does, the only difference is variance. Should player B have an infinite bankroll he should always call. If he has exactly a kelly bankroll he can call 50% of the time. anything other than a kelly bankroll he should fold 100%