River call effeciency question

[random]

Hello, we replied to your email 2 days ago, you have not sent us all the handhistories. However it seems like you figured out what is causing the difference in expected results.

As far as the formula goes, it has worked this way since HEM1 days, it has not changed.

It uses river pot size and opponents bet as the amount that can be won. So if river pot size is $3, and opponent bets $1, if you call $1 you will win $4 which will be your river efficiency $. In this case your river efficiency would be 4.

If it worked like you suggest, and it only looked at opponents bet size on river and ignored previous pot, then at most you could have river efficiency of 1 / $1. (Since at most you can only win $1 for opponents each $1 that he bets).

Here is one of the discussion on 2+2 if you would like to read more about it and how to interpret it:
River Call Efficiency - Micro Stakes Full Ring Games - Micro Stakes Poker Strategy Forum

[stevejpa]

Hello, we replied to your email 2 days ago, you have not sent us all the handhistories. However it seems like you figured out what is causing the difference in expected results.

Hm, I did reply to your email apologizing and attaching all hands, it's in my sent folder. (edit: lol I'm dumb I somehow sent them to myself instead of you :/ ) I won't bother resending because as you point out it's not really necessary anymore.

As far as the formula goes, it has worked this way since HEM1 days, it has not changed.

It uses river pot size and opponents bet as the amount that can be won. So if river pot size is $3, and opponent bets $1, if you call $1 you will win $4 which will be your river efficiency $. In this case your river efficiency would be 4.

Yes, and this is incorrect. The correct # is 5. Imagine a hand where there is $0 in the pot and your opponent bets $1. You call and win. Using your formula, your RCE would be 1/1=1. But that's silly, we profited one dollar on the river call. RCE of 1 implies that we broke even on our river calls. The correct way of calculating it is (1+1)/1=2.

To use your example imagine we call in the above situation 5 times, and win one. We win $4 the one time we win, and lose $1 each of the 4 times we lose. EV=$4*1+(-$1)*(4)=0. So we've broken even on our river calls. What is our RCE? 4 in the one hand we win, 0 in the other 4, for a RCE of 4/5 = 0.8. This is not correct. The correct RCE is (4+1)/1=5.

If it worked like you suggest, and it only looked at opponents bet size on river and ignored previous pot, then at most you could have river efficiency of 1 / $1. (Since at most you can only win $1 for opponents each $1 that he bets).

Not what I was suggesting though it turns out I was slightly confused about how HM2 calculates RCE. It is still wrong though as I've explained above.

[random]

Yes, and this is incorrect. The correct # is 5. Imagine a hand where there is $0 in the pot and your opponent bets $1. You call and win. Using your formula, your RCE would be 1/1=1. But that's silly, we profited one dollar on the river call. RCE of 1 implies that we broke even on our river calls. The correct way of calculating it is (1+1)/1=2.

Hello,

I understand point you are trying to make, however this is the way stat has worked and users are used to the values it provides.

We are not opposed to adding different river efficiency call stat if there was enough interest shown.

We are working on setting up Stat wiki (No ETA yet) where you will be able to comment on each stat, and that will be a good place to bring up your point and see how other users feel about it. Until then we will keep monitoring this thread and if there will be more users showing interest in it, we will see what can be done.

[stevejpa]

The values it provides are misleading at the very best! The stat is supposed to equal 1 when your river calls breakeven compared to folding. As currently calculated the breakeven point is impossible to know. If you always face 1/3 pot bets it's 0.8. If you always face 2x pot bets it's 0.6. For real players it's somewhere around 0.7 most likely. That's a completely ridiculous difference. Keeping it wrong because it's always been wrong does not make any sense.

In the sample I provided, that stat would tell me I'd have been better off had I folded every river I called when I won over 25 buyins on my river calls! I really don't know what else to say.

[random]

Hello stevejpa,

HM2 includes facing betsize filter that you can use, when analyzing your game on river.

That way, RCE stat will have a constant breakeven point that you are looking for.

One of the ways I like to do this is using Handmarkings, which basically gives you a way to create custom report. (you need to re-mark new hands when you analyze it in the future).

When combining facing river betsize filter together with river call or checkcall filters and using handmarkings, you can create something like this:
facing-riverbet.PNG

This way you can see how your river efficiency changes for different betsizes that you are facing on river, and you do not get mixed results.

Like I said, I will keep an eye on this thread and any similar thready about River Call Efficiency, and if there will be more users saying they would like some changes made to this stat, we will increase the priority of the issue.

[stevejpa]

cool I didn't know you could do that, thanks.

I do think there are two things you should do even if you're not going to change how the stat is calculated:
1. It doesn't work when you call and chop (RCE=0 for hands that you call on the river and chop (hh for this is pasted at the end)
2. Somewhere in your documentation make it clear that RCE=1 does NOT imply that you are breaking even on your river calls, because that is how everyone thinks it works.

Steve

File: HH20140601 Aenna #3 - $1-$2 - USD No Limit Hold'em.txt
PokerStars Zoom Hand #116993280697: Hold'em No Limit ($1/$2) - 2014/06/01 20:33:50 ET
Table 'Aenna' 6-max Seat #1 is the button
Seat 1: stevejpa ($262.57 in chips)
Seat 2: bdiddy1313 ($435.78 in chips)
Seat 3: thebestBarça ($123.27 in chips)
Seat 4: PkrJack ($200 in chips)
Seat 5: DawgKillah ($753.29 in chips)
Seat 6: FreeBeingM ($265.46 in chips)
bdiddy1313: posts small blind $1
thebestBarça: posts big blind $2
*** HOLE CARDS ***
Dealt to stevejpa [4h 4c]
PkrJack: folds
DawgKillah: folds
FreeBeingM: folds
stevejpa: raises $3 to $5
bdiddy1313: folds
thebestBarça: calls $3
*** FLOP *** [Jd 7c 2h]
thebestBarça: checks
stevejpa: checks
*** TURN *** [Jd 7c 2h] [7s]
thebestBarça: bets $8
stevejpa: calls $8
*** RIVER *** [Jd 7c 2h 7s] [5h]
thebestBarça: bets $15
stevejpa: calls $15
*** SHOW DOWN ***
thebestBarça: shows [4d 4s] (two pair, Sevens and Fours)
stevejpa: shows [4h 4c] (two pair, Sevens and Fours)
thebestBarça collected $27.22 from pot
stevejpa collected $27.22 from pot
*** SUMMARY ***
Total pot $57 | Rake $2.56
Board [Jd 7c 2h 7s 5h]
Seat 1: stevejpa (button) showed [4h 4c] and won ($27.22) with two pair, Sevens and Fours
Seat 2: bdiddy1313 (small blind) folded before Flop
Seat 3: thebestBarça (big blind) showed [4d 4s] and won ($27.22) with two pair, Sevens and Fours
Seat 4: PkrJack folded before Flop (didn't bet)
Seat 5: DawgKillah folded before Flop (didn't bet)
Seat 6: FreeBeingM folded before Flop (didn't bet)

[Witz17]

Hello random and stevejpa,

I just noticed the problem with the RCE calculation and also already made a thread here:
http://forums.holdemmanager.com/bug-...lculation.html

As I explained there, the biggest problem is, that the way it is calculated two different people may have the same RCE but one could make bad river calls while the other does make good river calls.

So it's kinda really useless. If people are used to this value, well they may be lulled into a false sense of security or they may call well but still get bad RCE values.

And the other smaller problem is with the RCE$ value, which is much more meaningful.
First the rake isn't subtracted there and second (as also here mentioned) it's corrected wrong for chopped pots (as if lost).

Would be really great, if you could fix that. If you need help, just tell me As computer scientist I would gladly take time to work myself through the code to fix it well just an offer

[fountouris]

Hi.
I have the same problem with river call efficiency. I still need to understand how this specific stat works so I need an answer to a specific question

In a specific stake (I examine ONLY the hands played at the SAME SPECIFIC stake and game), the current "River call efficiency" stat for me is 0.87

Hm2 says that River call efficiency is how much dollars I got for every single dollar I have put on the pot by calling the river.

This information is displayed in HM2 stats page, and it is also displayed on Leak Buster (I use Leak buster Pro).

I AGREE WITH THAT CURRENT DEFINITION OF THE STAT and it is EXACLTY the stat I need to have.


But if the definition is true, if the stat really calculates what you say, then I should have LOST money on the river when calling .

But my River call Eff$ stat in the same stake and game is POSITIVE..

So you have to explain me how it is possible to get 0.87$ for every 1$ investing for calling bets on the river, and at the SAME STAKE to have the river call eff$ stat with a POSITIVE value


Please give me an example of how that could happen to be able to understand.

I can't find any explanation to this by myself

And yes, I think that is true that everyone expects River call efficiency to be 1 when you are break even on your river CALLS... Yes people are used to see that stat now, but I don't think that they are many of them that understand what it really means. I don't. For example if I was break even on my river calls how much should be my river call efficiency with the way you calculate the stat now? I don't know, and I think that you also don't know and nobody knows!

LeakBuster Pro also says that I have a CRITICAL leak at the specific stake using that 0.87 river calling efficiency value . I can't understand how it is possible to have POSITIVE RIVER CALL EF$ and to have a critical leak. But if I really have a critical leak I need to know how you calculate the river calling efficiency to be able to fix the leak. Now I just see only a number that I don't understand it's mmeaning


The guy above says that he is a computer scientist. Before I start a poker career I had two bachelors, one in Maths and one with honors in Computer science and studies in applied and theoretical game theory in a top world class level, I had work also on scientific research and Artificial Intelligence . So the fact that someone with all those studies can't undestand the meaning of your stat values about river calling and find them unreasonable and can't interpret them, it should say something to you. And if I have a problem to interpret this, think about users that they are not math experts.

I don't agree with that "people are used from HM1 to see some values" and what they really want is to continue using the same values. People are just used to be confused with those values and unable to understand them. The reason that they don't ask for a change doesn't mean that they are happy with the values. I was confused with those stats for years and I didn't tried to ask for a change , and if you make a poll about the topic of how it is possible to have POSITIVE river calling efficiency $ and 0.87 river calling efficiency I'm sure that 100% of the users will answer "I have no idea" Everyone that I asked high about that (including high stakes professionals), they also say that they have no idea. People need stats that they are able to interpret the exact meaning of their values and not stats with meaning: "I think RCE between 1.2 and 1.9 is probably OK for me I don't know why but after years of playing poker I show that those values looks giving good results" to get an answer "No I don't agree I beleive that RCE is ok between 1.7 and 2.2" and after that to don't be able to answer the quetion : "HOW IT IS POSSIBLE TO HAVE RCE 0.87 and positive RCE $ values"

If you make the software do what you say that he does it should be better For example using your definition displayed on HM2 and Leakbuster also, RCE 0.87 means that I have lost 0.13 cents for every dollar I had put on calling the river, and my RCE $ value should be NEGATIVE! But it is positive ...

[fountouris]

I just checked what the RCE stat means in PT4 and found an answer to their forum about how the define RCE in their sotware

In PT4 RCE is the value is the quotient:

RCE=Amount of Money Called on the River and Won / Amount of Money Called on the River

Is there the same on HM2?

If yes, that is much different definition from "how much dollars I got for every single dollar I have put on the pot by calling the river" that you give in HM2 and Leak buster software and it gives me the answer how can someone with RCE < 1 can have RCE$ > 0




If yes, I finally found an answer to my question after years...

We still need an answer from the support if that holds and if yes, then you should REPLACE the definiton on HM2 software that "RCE IS how much dollars I got for every single dollar I have put on the pot by calling the river" to the way that you really calculater RCE is so the users can finally understand the values that they see as their stats.

Example and question:

I have played one single hand in my life: The pot is 1000$ the opponents bet 500$ I call 500$ and I won the pot

With the PT4 definition "Amount of Money Called on the River and Won / Amount of Money Called on the River" my RCE is 1, I have put 500 to call the river and 500 of them won"

With the definition ""how much dollars I got for every single dollar I have put on the pot by calling the river" that yu display to your software as RCE stat explanation my RCE is 1500/500=3

Which of these ways are you uset o calculate RCE? How can I calculate RCE by my own?

Give me an answer of what is the RCE that you think that the current version of HM2 calculates on the above hand example:RCE= 1, RCE=3, or RCE=something else? If RCE=something else then how much it is and how I can calculate that?

[Patvs]

HM's RCE definition/formula by Roy: River Call Efficiency - Micro Stakes Full Ring Games - Micro Stakes Poker Strategy Forum