Join Bridge Winners
All comments by Benoit Lessard
1 2 3 4 5 6 7 8 9 ... 285 286 287 288
You are ignoring the author of this comment. Click to temporarily show the comment.
Your post is a textbook example of fighting point schmoints with more point schmoints.

QTxxx
J9
AJxx
Tx

is almost a GF/2Nt jacoby but

xxxxx
xx
Axxx
xx

is a preemptive hand ?

“AKxxx, ATxx, x, Axx. That's only 15 HCP”
April 20
You are ignoring the author of this comment. Click to temporarily show the comment.
I think you should focus more on the than the count aspect of this hand. If im west and discourage what do I expect my partner to do ? I think the odds that his switch is is very high compared to him finding a switch especially after i didnt make a negative double. So if think in term of obvious switch or something similar you should discourage mostly when you think is better than club.

HHxx
xx
JTxx
852

My experience is that when I discourage with a hand like above I often regret it. I guess me and my partners dont rely too much on finding a switch in a “3rd suit” if dummy is not weak in it.
April 13
Benoit Lessard edited this comment April 13
You are ignoring the author of this comment. Click to temporarily show the comment.
I assume Richard line is strip the and play trump to get to dummy and play low to 7 or cover if the Q/T pop up. But for that you need to keep 2 entries on dummy.
April 11
Benoit Lessard edited this comment April 11
You are ignoring the author of this comment. Click to temporarily show the comment.
Its our first bad week now, from 2-3 deaths last weekend to a cumulative of 29 today mostly in seniors home. AFAIK all dead were either 70+ or had a medical predisposition.
April 4
You are ignoring the author of this comment. Click to temporarily show the comment.
Obviously Elo is more accurate but its also a way to lose customers. Masterpoint is obviously close to worthless as a skill evaluator but its having the worse effect that some stronger players with few master point and some weaker players with a lot of masterpoint prefer to stay home rather than being forced to play in their own masterpoint bracket.

I personally dont want to pay 20$ a session for a field weaker than my bbo games where I can play for free (I don't play vs random players on bbo)

I personally I prefer a good old price reduction rather than all the marketing guizmo from the Airmiles, coffee border cup and bogo. However if you have to do one to keep your customer you have to make some effort to make one that at least try to achieve some goals.

As a rating for a little game and for a game with a fair amount of luck as bridge I can assure you that its more than good enough. I see the skill difference between 20 and 30 the same way I see the skill difference between 1500 and 1650 at chess. Obviously 20 now wont be the same as 20 in 5 years but same goes for our currncy and we live with it.
April 4
You are ignoring the author of this comment. Click to temporarily show the comment.
I don't know how to explain it to you but the order of magnitude is not close at all, the play of the card and the simulation of possible holding with 52 cards is trivial (like in minutes not hours) compared to a game like Starcraft 2. What the machine did at what the machione did for S2 its very impressive. The problem with S2 is that the machine cannot play vs itself at super high speed like she ca do for simpler games.
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
The type of game doesnt matter that much for example you could make a game where a robot need to cross the street without getting hit by cars. You could have some cars no drivers they are just physical objects (think Atari asteroid), cars that are driven totally randomly, Tesla cars and cars that are driven by humans. The AI doesnt need to know there is 4 types of cars, he doesnt need to know the laws of physics, he doesnt need to know how humans drives all he need is a lot of data on possible events and a clear goal. For him its not a game hes relying on the laws of large numbers and is just crunching numbers.

In that regard bridge is quite simple compared to a game like shogi or to a video game where you have to take account pixel lenght and hundreds of units.
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
Its the last version that beat AG 100-0 not the 2017-2018 version.
March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
https://www.youtube.com/watch?v=rkt51PTCQG4&t=2201s
game that make no sense even for the strongest GM I was told.

https://www.youtube.com/watch?v=nPexHaFL1uo&list=PLtwgYi5EtlSc39zoTVl_oAYCvU0MY5Q_Q&index=3&t=834s
Another Star Trek game
Same game with better commentary IMO
https://www.youtube.com/watch?v=wui0YweevtY
March 29
Benoit Lessard edited this comment March 30
You are ignoring the author of this comment. Click to temporarily show the comment.
You are totally right that perfect information or not isnt relevant they just weight outcome is a statistical way, basically when you are far fron the computionnal horizon it doesnt matter at all if its a game of luck or a game of perfect information or just calculating weather forecast. When you are in the computationnal horizon the game is over if the AI is winning and if the AI is in a losing position hes simply freerolling, weither or not he can anticipate wich mistake the opponents is more likely to make is the weakness of the AI (winning a position that he would lose if playing against himself) but since hes not behind that frequently it doesnt matter that much. As for the psychology there is a lot of stuff that we consider trick or ruse that are just technique for this type of AI.


However even 5 years ago I was pretty much convinced that Alpha Go can learn bridge well enough to crush top players, the 2017-2018 version Alpha Zero beat the AG version (the one in the movie) 100-0 after just 4 hours of learning. There is also a huge discrepency between the skill level of a partnership game between pro bridge players and pro from games like go and shogi where professionnal train kids at a very young ago, its a bit like comparing today best axe thrower to Urss gymnastics of the golden age. Im pretty much convinced that with today computer it take less than 2 hours to overkill humans learning bridge from scratch but using a already learn bidding system.

The creation of bidding system part is too tough for me to estimate.There need to be discussion on what we will allow the agents to communicate and on how these AI agents will create the bidding system. However once the bidding system is created its will be trivial for the 2 independant agents to overkill humans.Note that they are using some of the most powerful computer on earth to create these AI.

If you know chess you can find Alpha zero vs Stockfish games and its like watching Mike Tyson first 20 fights except you dont see some of the punchs because we dont really understand them.
March 29
Benoit Lessard edited this comment March 29
You are ignoring the author of this comment. Click to temporarily show the comment.
better minor lebensohl type of bid, both minors or diamond to play a direct 5D is a bit slammish.

In this particular auction its no that good because the odds of your side having a good 6D is less than the odds of them being to find a borderline penalty double after the temporized bid of 4NT/5C compared to a direct 5D but for my partnerships its always on.

But if opener bid was a preempt or if we were vul vs nv its a pretty good convention.
March 29
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes 2NT followed by 3D showing only weakish diamonds is a tradeoff.

2452 inst a problem if your have some extra values you bid 3D directly if not you go 2NT-3C-3D. Correct to H should apply in cases like.

P—(2S)—X—(P)
??

3442

with values you bid 3red directly.
without values you go via 2Nt (ask a minor).
if partner pick 3D you pass if he pick 3C than 3H is slighty better than 3D.

older method is 2Nt follow by 3D is D+H
March 24
Benoit Lessard edited this comment March 24
You are ignoring the author of this comment. Click to temporarily show the comment.
I missed the initial responder X but my

my agreement is just better minor lebensohl in those similar auction.
3red = some values. (here I already show values so it show extras values)
2Nt pick a minor or no extra values (opener bid 3D with 1345 3C with 2335/1336)
3C may or may not have extras values.

2nt followed by 3red no extras values. This is the plan even if your weakish have have both reds. Whne you have both reds and opener choose D you pass but if he pick clubs you correct to 3H.
March 23
Benoit Lessard edited this comment March 23
You are ignoring the author of this comment. Click to temporarily show the comment.
If you hand is very strong or very weak it should still work but its not foolproof.
March 23
You are ignoring the author of this comment. Click to temporarily show the comment.
I think the odds that partner is missing a H control and we reach 6D instead of 6S + the odds that 1S-2D-2H-2S allow me to find the best spot are higher than the other way around so ive discarded 2Nt fairly quickly.

Most of the time its going to be a 6 or 7 so implying early that the Q is gold just look normal to me.
March 22
Benoit Lessard edited this comment March 22
You are ignoring the author of this comment. Click to temporarily show the comment.
Ive never understood why strong players dislike 2 way double (except for ethical tempo consideration).

here responder could have 3307 up to 3343 and opener could have 5 to 8D so its kind of clear that 2way X not based on pts might not work here.
March 22
You are ignoring the author of this comment. Click to temporarily show the comment.
I would upgrade since the 5th card is worth a trick quite often but the weaker you are the less you can count on it.

a 19 count kq7xx facing a stiff J can usually enjoy the lenght.
June 2, 2019
You are ignoring the author of this comment. Click to temporarily show the comment.
“undiscussed” was probably just a bit lazy or narrow explanation or their partnership agreement.

Me and many partner probaly didnt discuss this particular auction but we probably did discuss

1- cuebid delayed or not is stayman (in all case or with some exceptions)
2- using 2nt as a way to show or deny stopper.
3- fast or slow denies
4- we play leb over a weak 2 as over 1nt or not

I think the opps are entitled to this even if a particular auction isnt clear.

Im not saying I disagree with the ruling just that as south I would probably ask further explanation knowing that undiscussed is is often just a lazy explanation. As west I would never simply write undiscussed here, if I have a reason to tried 2NT followed by 3H instead of 3H direct than me and my partner have some background knowledge that opps deserve. Maybe west tough that this was obvious and it was south job to ask for further explanation maybe they were late idk.

Maybe they play that style after 1NT opening but its not clear if they play this way over a weak 2.

Maybe they play xfer leb but play reg leb in some case and obviously they didnt discussed that much.

Anyone who think this is awm probably didnt read appeals book/resume lately.
May 17, 2019
You are ignoring the author of this comment. Click to temporarily show the comment.
Like in many poker tournaments its quite likely that there is going to be side games that are tougher and more interesting than the main events.
May 8, 2019
You are ignoring the author of this comment. Click to temporarily show the comment.
Googling ignoramus + ignorami led me to “Take a Wok on the Wild Side”

https://www.youtube.com/watch?v=979g01GPmOw (I assure you no need to listen)
April 14, 2019
1 2 3 4 5 6 7 8 9 ... 285 286 287 288
.

Bottom Home Top