[Free]

Seems like a lot of peaks for quite frequent hand types...

[jtfanclub]

OK, I changed my chart above to include the fomulas used by Tysen.

I then changed three things (two, really) about Pavell's formula:

1. I added .25 tricks (one point) per doubleton.
2. I added 1.5 tricks, instead of 1.25 (six points instead of five) per void.
3. I guessed that Pavell's formula overvalued the 4333 distribution by .1. Note that I'm not saying that he should add in 'subtract 4/10 of a point if you're 4333'. This is for calculation purposes only: I made 4333 worth -.1, which since 4333 is the baseline for the formula, it added .1 to all of the other values. I included the error in the 4333 calculation in with the other errors.

Another way to think of #3 is instead of using 4333 as the baseline, I used 5332.

I have no idea if my math is right, but if it is the total is now .470. Not only better than Zar, but better than TSP. The over-valuation of 4333 made the rest of the calculations look off. It would probably look better still if you figured out the exact error of 4333 relative to the other distributions, but I think that's unnecessary.

Notice that 4441, 5332, 5422, and 5431 all have an error of .005 or less. That means it's correctly calculating the difference between the trick taking potential of a 4441 hand and a 5431 hand, for example. Neither Zar nor TSP come anywhere near that accuracy for the more frequent hands.

It looks to me like using Pavell with 6-3-1 instead of 5-3-0 is the best of the bunch. I'm very curious to know how this modification would affect the overall system in terms of correctly evaluating hands.

[Pavell]

GijsH, on Nov 2 2004, 12:12 PM, said:

The principles presented here sound very similar to MIDMAC.
It also seems that not a Bidding System is described but a Hand Evaluation System. Am I right Pavell?

MIDMAC has the advantage that standard HCP count is used (ace=4 etc), see http://www.midmac.tk for more details.

Maybe Pavell can explain where his system differs from other systems?

- Probably you will be right if you read only first 2 Chapters. Wait to post the next two Chapters, and the difference between the Hand Evaluation System and the ForcePoint Bidding will appear. Then you will see how to discover your patrtner's EXACT shape below 4♣ (you will do that dynamically, almost not using static bids, even after opps preemtive overcalls), and how to discover your partner's Controls with One Only question, all that leaving more than enough Bidding space for discovering their suits, along with the Queens and the Jack in the longer partner's suit (of course, if you need that information). First two Chapters alone (although they are included in all versions of ForcePoint), are for players who want to streighten their assurance for the final contract's Level, using their bidding system. The original Fp bidding will start with Chapter 3. That will be enough for everyone to understand why this system differ from all other systems.

[Pavell]

Flame, on Nov 2 2004, 10:06 AM, said:

Pavell, on Nov 2 2004, 08:39 AM, said:

I'm not a Player, I'm a theorist. Probably you know who was Goren and what he did for the bridge world ? But do you know his players qualification ? I'll leave you to discover it alone

Dont leave pls
You just have to understand that many can say thier system is the best, by saying that you dont add anything to how i see your system.
You can be just a kid who think he invented the world and no one else know anything, i dont know you.

- It is not necessary to know me, simply try the Counting for now, with all hands face up (you can download such a Demo from my site), despite what system you use. Or forget about me, it is OK, too.

[Pavell]

Flame, on Nov 2 2004, 10:06 AM, said:

Pavell, on Nov 2 2004, 08:39 AM, said:

I'm not a Player, I'm a theorist. Probably you know who was Goren and what he did for the bridge world ? But do you know his players qualification ? I'll leave you to discover it alone

Dont leave pls
You just have to understand that many can say thier system is the best, by saying that you dont add anything to how i see your system.
You can be just a kid who think he invented the world and no one else know anything, i dont know you.

I'm 55 and have made the Bidding of ForcePoint for 12 of the last 15 years

[Pavell]

jtfanclub, on Nov 2 2004, 02:17 PM, said:

Pavell, on Nov 2 2004, 04:14 AM, said:

Probably you are not. But I'm not working with 0.5 tricks, nor you can bid 3.5 ♥. 

Sure you are. You're taking a point total, and dividing by 4. You end up not only with half tricks, but with quarter tricks. You can arbitrarily declare that a 3/4 of a trick is a full trick, but that's still dealing with them.

I didn't do anything fancy- you add up all your numbers and then divide by four, I just divided by four first. That made the numbers equal to tricks, and therefore (for me) easier to understand.

Let me challenge you on two things, and have you refute them. That might get the ball rolling:

You claim that a singleton ace is worth 9 points, or the full value of the ace plus the full value of the singleton. You claim that a singleton king is worth 3 points, or only the full value of the singleton.

This is very different from most methods. Most consider A singleton to be worth 2 tricks- 1.5 for the A, only .5 for the singleton, because the ace isn't helping lower cards. The King is worth 1.5 tricks- the singleton king may help your partner take tricks, and may win a trick on defense.

In contrast, in Midmac a singleton generally ends up worth 2 more than a doubleton (they count shortness and length, and it adds one to both sides). A singleton king subtracts one point. from a normal king. A singleton ace ends up being worth about 6, a singleton king about 4.

So, in your system, a singleton A is worth three times as much as a singleton king. In MIDMAC and most other counting systems, the difference is half that: a singleton ace is worth 1 1/2 times as much as a singleton king.

That's a heck of a big difference.

Can you explain, mathematically or otherwise, the reason for this disparity?

I'm trying to compare your system to Zars, which is tough because you use 4 points per trick, he uses 5. However, it appears the biggest difference is that you value HCP a great deal more than he does- he tends to evaluate distribution stronger. You're both using 6421, but you're dividing by 4, and he's dividing by 5.

Have you read the math on Zar points? What do you think are its biggest flaws?

http://www.bridgeguy...rPoints2003.pdf

- We are friends with Zaro. It was a pleasant surprize when we have discovered each other 1-2 years ago with almost equal Countings. I have tried his 5 pts step and I have 5-6 versions of Demos on my http:/bullbridge.com site with 5 step for a trick Counting. They proved to me that the simple Counting method I'm using (read for the later added Negative pts, I forgot to mention them in the 1st post's issue, sorry about that) is enough stable. I do not know why a Counting difference for the Singleton Controls occured. The Double Dummy's checking showed me that. I'm still unaware of MIDMAC (recently I didn't find time to go there after your reply, but I'll do that in the future). By the way, those ForcePoint Singleton's Countings immediately become unstable with 5 pts step for a trick. I have tried to find their whole numbers value, but it proved to be impossible (I'm against the overloading even with the use of 0.5 values). You told me that I'm using even 0.25 value when I divide by 4 and have a reminder, but that's not really true, because Im watching ONLY if I'm inside the Base's range, approximating the division result without remembering the part values for future use. This is easier than the counting of the part's values, and you can do it absolutely mechanically. You say we both, I and ZAR, use 6421 Counting for the Honor's value, that's right, but if I didn't make a mistake, I think, 2 year ago, before we met, ZAR didn't count the Control's value, and I was pleased recently when I visited his page (I remember he had tried to evaluate the hands by some other very difficult, by my opinion, methods, I even do not want to remember them.
- Now, about the difference between ForcePoint sytem and ZAR "system". ZAR doesn't have a system at all, at least for now, because he doesn't have any Bidding (recently he tryed to make something in this dirrection). When I publish Chapter 3, you will be able to check the original ForcePoint's Bidding, if still interested (you can find it in separate GUI programs, if you download the free Demo, simply start the file "Sysnotes.exe"). But do not forget, it is for computers and Experts, although the regular players can use it, too. It is very easy for the Answerer, because s/he bid what s/he see, but require a great concentration by the partner, who has to imagine the EXACT Answerer's shape. The other systems work with partially shape's explanatons, ForcePoint doesn't, that's why with it, you would refuse to bid a Slam (or NEVER will miss one). It is not easy to explayn everithing to you, you have to try it... Download the free Demo, it will do all the Counting for you (there is not an "auto" ForcePoint Bidding, enter your bids manually, using your own system and compare your shoosen Play Level (PL) with the PL which Fp will suggest to you by adding the both partners B# and using some Adjustments, if necessary. All of it is on the http://bullbridge.com

[Pavell]

tysen2k, on Nov 2 2004, 02:09 PM, said:

Here's how Pavell's hand evaluation system stacks up compared to some others:

             ERROR    SCORE
HCP         1.23    -0.49
HCP+321     1.07     0.00
HCP+531     1.05     0.07
Zar         1.05     0.08
Pavel       1.04     0.11  <--
BUMRAP+321  1.03     0.14
BUMRAP+531  1.02     0.21
TSP         1.02     0.21
Binky       0.99     0.32

My usual disclamers for those who know me...

ERROR is the average # of tricks there is in difference between how many tricks we think we can take and how many we actually take.  This is calculated using a double dummy evaluation of over a million random hands.

SCORE is an estimation of the IMPs/board we expect to gain against a team that uses a simple HCP+321 evaluation method. It’s a measure of how much payoff there is for using a better evaluation system.

HCP is A=4, K=3, Q=2, J=1

HCP+321 is HCP + 3 per void + 2 per singleton + 1 per doubleton

HCP+531 is the same with more points assigned to shortness

Zar is HCP + Controls + twice the length of longest suit + once the length of second-longest suit minus length of shortest suit.
http://public.aci.on.ca/~zpetkov/

BUMRAP is a substitute for HCP: A=4.5, K=3, Q=1.5, J=0.75, T=0.25

TSP is the method described in this article. It’s an attempt to find the best evaluator using simple whole numbers.

Binky is Thomas Andrew’s evaluator:
http://thomaso.best....dge/valuations/

Tysen

- You have to do it 1 more time (I'm sorry, I forgot to add the vital information for the Negative (N) pts. It is already done. And do not forger, that I greatly prefer to sacrifice some explicity by using the whole numbers Honor's Counting, instead to make Human players crazy and very !!!

[hrothgar]

Pavell, on Nov 4 2004, 01:00 PM, said:

Any chance that you would be willing to implement this as a web based application?

I don't install random executables on my PCs and I STRONGLY recommend that others adopt similar policies

[Pavell]

Flame, on Nov 2 2004, 03:00 PM, said:

tysen2k, on Nov 2 2004, 02:09 PM, said:

Here's how Pavell's hand evaluation system stacks up compared to some others:

             ERROR    SCORE
HCP         1.23    -0.49
HCP+321     1.07     0.00
HCP+531     1.05     0.07
Zar         1.05     0.08
Pavel       1.04     0.11  <--
BUMRAP+321  1.03     0.14
BUMRAP+531  1.02     0.21
TSP         1.02     0.21
Binky       0.99     0.32

My usual disclamers for those who know me...

ERROR is the average # of tricks there is in difference between how many tricks we think we can take and how many we actually take.  This is calculated using a double dummy evaluation of over a million random hands.

SCORE is an estimation of the IMPs/board we expect to gain against a team that uses a simple HCP+321 evaluation method. It’s a measure of how much payoff there is for using a better evaluation system.

HCP is A=4, K=3, Q=2, J=1

HCP+321 is HCP + 3 per void + 2 per singleton + 1 per doubleton

HCP+531 is the same with more points assigned to shortness

Zar is HCP + Controls + twice the length of longest suit + once the length of second-longest suit minus length of shortest suit.
http://public.aci.on.ca/~zpetkov/

BUMRAP is a substitute for HCP: A=4.5, K=3, Q=1.5, J=0.75, T=0.25

TSP is the method described in this article. It’s an attempt to find the best evaluator using simple whole numbers.

Binky is Thomas Andrew’s evaluator:
http://thomaso.best....dge/valuations/

Tysen

Nice score by pavell.

- Thank you. And you can add that ForcePoint use ONLY whole Honors numbers, which is easier than the other Countings by the above chart. Also you DON'T need to make any evaluation of your hand, because it is already automaticaly included by the approximation of the division result (when you divide your Total pts by 4 and subtract 2 to find your B#). All this combined, along with the aggresive openings (sometimes even more aggresive than ZAR's) assure a competitive, but very easy Counting even for Beginers. If you use the elementary Adjustments (for Super Fit, Misfit and for Void/Singleton opposite 6 or 7 card suits), you can find your Play Level (PL) with over 83% warranty for all of the billions card distributions.
- The simplicity without loosing the accuracy was the MAIN GOAL for this this system. You will tell if I succeded to achieve it.

[Pavell]

hrothgar, on Nov 4 2004, 06:57 AM, said:

Pavell, on Nov 4 2004, 01:00 PM, said:

 

Any chance that you would be willing to implement this as a web based application?

I don't install random executables on my PCs and I STRONGLY recommend that others adopt similar policies

- You can find all on http://bullbridge.com alond with a ForcePoint's Demo instalation program, which will do all the Counting for you. Sorry, no auto Bidding yet (if you use the "autoBid" button, it will supply wrong bids), and be aware that all hands are face up.

[Pavell]

tysen2k, on Nov 3 2004, 03:55 PM, said:

jtfanclub, on Nov 3 2004, 03:47 PM, said:

I just took Pavell's method and added .25 (one point) per doubleton.  This seems to be much, much closer to the Real results.

Yes, this gives a 5/3/1 system of distribution which I advocated for some time on this forum. I have no idea why Pavell doesn't value doubletons at all.

Quote

I'm not sure what forumula you used for error, so I didn't do that part.

Error = (frequency)*(real-predicted)^2
where frequency is the whole number percent of that hand pattern occurring. 4333 is 10.54, 4432 is 21.55, etc.

Tysen

- I will tell you why I do not want to evaluate the Doubletons. I went all this way, which TSP use. The Doubletons are tricky. You can have AK, AJ, Ax, Qx, Jx, QJ, xx and so on... and have to pay an additional attention to the possible bad suit's comunication. If you are willing to pay a separate attention to each of them, you probably will have better results, but I found it too borring. I simply didn't want to overload the regular players with this additional evaluation. 4 or 5 pts Base is enough wide to swalow the Doubleton's evaluations. The only thing I'm applying to ForcePoint for hand with Void, Singleton or Doubleton is when the reminder of the division's result is 2. For such hands I simply approximate the division resuilt up for the upper B# (I'm doing it also if no pseudo suits appear at all, but only if I have "enough" Control Points (CP) with conjunction with my particular B#. What is "enough" is not published here, and may be discussed only privately). That's my opinion, you are free to disaggree and correct it

[Pavell]

Flame, on Nov 2 2004, 01:55 AM, said:

I advice you to leave the details of your system and explain why you think its better then other systems.

- I will answer to you although I think that there is no big sense in the bare advertising. You have to try the horse, before you buy it

1. You do not need to evaluate your hand, it is automatically done with simply aritmethical aproximation
2. ForcePoint uses whole Honor's numbers
3. Anyone can use the bare Counting and implement it to own system
4. If you decide to use the original ForcePoint's bids, you will discover your partner's EXACT shape below 4♣ with its dynamic OneOnly Convention for all your distributions and Control's needs
5. The system doesn't need additional Conventions, it uses only one
6. It has highly aggresive opening bids for the Majors suits
7. You will NEVER miss a Slam
8. The original Control Points Asking (CPA) will permit you to discover your partner's number and kind of Controls simultaneously with OneOnly Question.
9. You will have enough Bidding space to discover not only the Control's suits, but even the Queens + the Jack in the longest partner's suit (if you need that)
10. The system accuracy is over 83% for the billions of hands distributions (but you have to check at least 0.000001% of them to believe me)

- So, any Beginer will immediately evaluate his/her Bidding to the Expert's level, despite the unability to play good (but that's another question, Augustin Madala has played perfectly even before 16, some players like me continue to try it even over 55)

[Free]

Pavell, on Nov 7 2004, 04:15 AM, said:

7. You will NEVER miss a Slam

I really can't believe this, never say never. There has been hands posted in the past where even with the highest precision possible we couldn't find it. Somewhere in the non-natural system forum there's posted such hand, I'd like to see if you can find slam on that one (something with a ♠ void and perfect fitting honours if I remember correctly).

[Pavell]

Free, on Nov 7 2004, 07:47 AM, said:

Pavell, on Nov 7 2004, 04:15 AM, said:

7. You will NEVER miss a Slam

I really can't believe this, never say never. There has been hands posted in the past where even with the highest precision possible we couldn't find it. Somewhere in the non-natural system forum there's posted such hand, I'd like to see if you can find slam on that one (something with a ♠ void and perfect fitting honours if I remember correctly).

Are you talking for this hand ?

AQxxx.... Kxxx
AKxx......Qxxx
Ax..........Qxx
xx..........AK

If your hand has a ♠ Void, find it, I coudn't.

[Flame]

Oviously theere isnt a system which never miss a slam, saying something like that doesnt encarage me to try your system.
Anyway ppl here are planing a game with different systems, i wish your system could take part in this.

[mikestar]

Pavell, on Nov 7 2004, 03:15 AM, said:

7. You will NEVER miss a Slam

You are a liar. No bidding method never misses a slam--even "Bid a slam on every hand" will sometimes fail by getting you to the wrong slam!

[hrothgar]

> 7. You will NEVER miss a Slam

Interesting claim

---

Can you find the Grand on this hand?

[Gerben47]

4NT p 7♣ dbl p p p

4NT is both minors

[Flame]

7♥ or 7♠ or 7♣

[MarceldB]

> 7. You will NEVER miss a Slam

but also the correct one?

W/-
3 ------------ AKQJ96
AK98753---- Q6
- ------------ T972
AJ743------- Q
7♠?

Axx--------- xx
Axx--------- xxx
AKQx------- xx
xxx--------- AKQxxx
J♦ investigated? If yes 5NT still playable?

Axx--------- xxx
Axxx-------- xx
AKJx-------- Qxxxxx
Ax----------- xxx
Do you know the 6th♦ for 3NT? If yes the Q♦ too?

AKxx-------- xx
AJxx--------- xx
KQ----------- Jxxxxxx
AKx---------- Qx
Q♠ + Q♣ investigated, if yes J♦ too?

To all above questions I can assent with my system BUT claiming "NEVER miss a slem" is an other matter.
So I'm curious to be informed about your sequences for above examples.
Thanks.