Are there any benefit of sitting in 1st/3rd seat?

[myfish]

I am interested to know about it(just for fun :angry: )

[Free]

Yes there is! Especially if you lie :P It's a matter of fooling opps vs fooling partner.

- In 1st seat you have only 1/3 chance your partner has the good hand, while in 2nd seat you have 1/2.

- In 3rd seat you already know partner passed and won't go anywhere (unless you open really solid and he passed with a nice 12 count). 3rd seat however is less efficient, because many people play very light openings there, and most psychs happen in such situations. I wouldn't have my first partner telling me: "3rd seat everything goes" :angry:

- In 4th seat you need a solid opening imo, otherwise you give the contract away and you get a negative score instead of a 0.

[jillybean]

Nice explanation :rolleyes: :D :P

[pclayton]

3rd seat is the equivalent of playing on the button (dealer) in hold 'em. You tend to control the action with the appropriate hand.

 

1st seat is good, but not quite as much as 3rd. But uou get to establish the tempo of the auction right away.

 

2nd seat is like playing the big blind. Big 'variance' plays generally don't work.

[helene_t]

An additional advantage when opening (especially preempting) in first seat is that RHO, who is in ballancing seat, can still have opening strength, so the range of his overcalls/doubles become very wide.

[tysen2k]

This is a question I posed to rgb several months ago and I got a lot of "that's a cool idea, I'll look into that" but it never went anywhere. I wanted to try to actually quantify how much is the "dealer's advantage."

 

If anyone on this forum is good at running this kind of simulation, I'm sure everyone would be interested in seeing the results.

 

Have GIB or Jack run a team match of a few thousand hands with the computer playing in every seat. At one table North is always dealer and East is always dealer at the other table. This has the effect of one team always having the advantage of being dealer on every hand. How much does this team win by?

[Gerben47]

I recently 'discovered' that Jack can play PBN deals so I can let it play deals from a PBN file with north dealer and then the same hands with east dealer.

[Cascade]

This is weird ...

 

I have simulated 200 hands using Jack with two sets of identical hands except that in the first set North is always dealer and in the second set East is always dealer.

 

After 200 hands the score is 329 - 460 in favour of 2nd/4th over 1st/3rd.

 

I have no idea what is going on here.

 

The average 0.655 Imps per board to 2nd seat.

 

The sample standard deviation is 6.01 Imps per board.

 

This is getting close to 95% significant that if 1st seat has an advantage we would not be seeing this data.

[Free]

I think I can understand the reason why. If you open in 2nd seat, you have some information about your RHO (passed), while in 1st seat, you don't have that information. Perhaps that's why the program can play the hands easier and win some imps.

However, I think if you would give the program a VERY frequent 2-level structure, the results might change in favor of 1st/3rd seat, but I'm not sure.

[mikestar]

I think I can understand the reason why.  If you open in 2nd seat, you have some information about your RHO (passed), while in 1st seat, you don't have that information.  Perhaps that's why the program can play the hands easier and win some imps.

However, I think if you would give the program a VERY frequent 2-level structure, the results might change in favor of 1st/3rd seat, but I'm not sure.

If your idea for the cause is correct, frequent two bids won't make so much difference. Second hand opneings will be less frequenct, but when they do occur, the inferences about first hand weakness will be more significant.

[Free]

Wayne, do you have the scores of these hands? On what kind of hands does the program gain so much?

[bestguru]

What if two openers can be weak or strong? 2cnd seat won't know as much. Something like:

2♣ 1 suiter

2♦ majors

2♥ spades and a minor

2♠ minors

 

All force a response so opener has a chance to rebid with a strong hand.

[Free]

What if two openers can be weak or strong? 2cnd seat won't know as much. Something like:

2♣ 1 suiter

2♦ majors

2♥ spades and a minor

2♠ minors

 

All force a response so opener has a chance to rebid with a strong hand.

Yes, but 2nd seat has 2 rounds to bid when the opening was weak (most frequent)...

[Cascade]

Wayne, do you have the scores of these hands?  On what kind of hands does the program gain so much?

I do have the scores and I can potentially get the results but not the bidding or play. Unfortunately it is quite onerous to get the results. But I might be able do that (but not the bidding and play) for some of the extreme results.

 

Deal  N Dlr E Dlr   Diff   IMPs
85    -100   1190   -1290   -15
44    -400   500    -900    -14
63    -800   100    -900    -14
35    -650   200    -850    -13
95    -100   750    -850    -13
196   -200   660    -860    -13
34    -100   600    -700    -12
42    -100   620    -720    -12
90    -600   100    -700    -12
99    -50    590    -640    -12
103   -100   630    -730    -12
122   -100   600    -700    -12
127   -100   620    -720    -12
130   -100   620    -720    -12
46    -50    480    -530    -11
53    130    660    -530    -11
75    -100   400    -500    -11
112   -100   400    -500    -11
156   -480   50     -530    -11
169   -100   420    -520    -11
24    -500   -50    -450    -10
97    -420   50     -470    -10
105   -630   -180   -450    -10
107   -430   50     -480    -10
179   -50    400    -450    -10
186   -630   -240   -390    -9
61    -200   140    -340    -8
89    100    420    -320    -8
48    200    500    -300    -7
163   -170   100    -270    -7
178   -140   130    -270    -7
18    -420   -170   -250    -6
25    -150   110    -260    -6
115   -50    170    -220    -6
162   -90    170    -260    -6
182   180    430    -250    -6
189   -110   130    -240    -6
49    -50    140    -190    -5
58    -90    110    -200    -5
70    -50    140    -190    -5
106   -100   110    -210    -5
167   -100   110    -210    -5
177   -150   50     -200    -5
41    -650   -500   -150    -4
168   -100   50     -150    -4
176   -50    110    -160    -4
6     100    200    -100    -3
26    -200   -100   -100    -3
30    -400   -300   -100    -3
74    80     170    -90     -3
129   -170   -50    -120    -3
132   0      100    -100    -3
33    50     100    -50     -2
51    460    510    -50     -2
59    420    480    -60     -2
9     400    430    -30     -1
11    -460   -430   -30     -1
23    650    680    -30     -1
56    420    450    -30     -1
65    400    430    -30     -1
68    -660   -630   -30     -1
69    -510   -480   -30     -1
71    620    650    -30     -1
131   -630   -600   -30     -1
133   620    650    -30     -1
136   -140   -110   -30     -1
145   -450   -420   -30     -1
149   -430   -400   -30     -1
155   200    230    -30     -1
173   -680   -650   -30     -1
191   680    710    -30     -1
195   -690   -660   -30     -1
198   90     110    -20     -1
2     -90    -80    -10     0
4     -100   -100   0       0
10    -1460  -1460  0       0
14    -460   -460   0       0
15    -200   -200   0       0
16    450    450    0       0
21    650    650    0       0
31    690    690    0       0
32    -710   -710   0       0
37    -200   -200   0       0
39    -200   -200   0       0
40    450    450    0       0
47    620    620    0       0
50    620    620    0       0
52    170    170    0       0
54    -620   -620   0       0
64    -650   -650   0       0
67    110    110    0       0
77    -630   -630   0       0
78    -920   -920   0       0
81    100    100    0       0
82    -200   -200   0       0
83    -50    -50    0       0
84    -90    -90    0       0
87    100    100    0       0
91    450    450    0       0
93    660    660    0       0
96    -150   -140   -10     0
101   -90    -90    0       0
108   110    110    0       0
113   -100   -100   0       0
116   630    630    0       0
119   110    110    0       0
120   50     50     0       0
123   -50    -50    0       0
124   -300   -300   0       0
125   660    660    0       0
126   -490   -490   0       0
128   480    480    0       0
134   -140   -140   0       0
135   -120   -120   0       0
137   -600   -600   0       0
139   450    450    0       0
140   170    170    0       0
143   -150   -150   0       0
144   -50    -50    0       0
146   -100   -100   0       0
150   120    120    0       0
152   -170   -170   0       0
154   600    600    0       0
157   -110   -110   0       0
159   -140   -140   0       0
160   430    430    0       0
161   420    420    0       0
164   -140   -140   0       0
165   -450   -450   0       0
166   -680   -680   0       0
170   -1370  -1370  0       0
171   -100   -100   0       0
175   -460   -460   0       0
180   100    100    0       0
184   -430   -430   0       0
187   140    140    0       0
190   400    400    0       0
192   140    140    0       0
193   -450   -450   0       0
199   -90    -90    0       0
200   -450   -450   0       0
5     -130   -150   20      1
12    -420   -450   30      1
13    650    620    30      1
28    620    600    20      1
29    710    680    30      1
43    480    450    30      1
73    -140   -170   30      1
80    -600   -630   30      1
88    -100   -140   40      1
109   630    600    30      1
110   90     50     40      1
117   90     50     40      1
121   480    450    30      1
197   -430   -460   30      1
57    -50    -100   50      2
62    -100   -150   50      2
94    -460   -520   60      2
118   -50    -100   50      2
142   200    150    50      2
148   690    620    70      2
158   -450   -500   50      2
172   150    100    50      2
174   -450   -510   60      2
7     -100   -200   100     3
17    380    280    100     3
36    200    100    100     3
72    50     -50    100     3
76    -90    -200   110     3
138   210    110    100     3
60    100    -100   200     5
66    110    -100   210     5
147   -300   -500   200     5
20    120    -120   240     6
38    100    -140   240     6
104   450    200    250     6
1     110    -170   280     7
98    500    200    300     7
181   -130   -400   270     7
27    450    100    350     8
79    150    -200   350     8
45    -200   -620   420     9
100   100    -300   400     9
102   500    110    390     9
151   -200   -620   420     9
19    420    -50    470     10
22    -140   -620   480     10
111   50     -420   470     10
114   600    150    450     10
188   630    150    480     10
194   300    -140   440     10
3     450    -50    500     11
92    620    50     570     11
141   750    150    600     12
153   100    -600   700     12
8     500    -300   800     13
86    140    -650   790     13
183   130    -620   750     13
55    750    -200   950     14
185   1210   200    1010    14

[tysen2k]

Wayne,

 

Thanks for starting the simulations.

 

What convention card are you using for Jack? Maybe it should be done twice, once with a conservative approach and once with an aggressive one.

 

Actually with only 200 deals, we're not quite at the 95% significance level, but it's getting close. I think we need at least 1000 deals.

 

Tysen

[Cascade]

Wayne,

 

Thanks for starting the simulations.

 

What convention card are you using for Jack? Maybe it should be done twice, once with a conservative approach and once with an aggressive one.

 

Actually with only 200 deals, we're not quite at the 95% significance level, but it's getting close. I think we need at least 1000 deals.

 

Tysen

Yes I said almost 95%. I calculated this as the tail probability 0.061588943

 

I am not 100% sure of the answers to these questions - I will tell you what I did (and why).

 

I have 1000 deals ready to analyze.

 

For standard play Jack's scoring options seem to be Total points, Rubber or Chicago.

 

Therefore to get an IMP score I set up a tournament with the 1000 hands. Actually 2000 hands since each deal is played twice with a different dealer each time. Also there were in fact more than one tournament since Jack would not cope with one 2000 board tournament.

 

Jack ran and created a result for each of these 2000 hands - I think I still have some to run. When this runs in the background I am not sure what convention card it uses but I did not choose the random convention card option so I presume it uses the one that I have loaded for Jack.

 

This is a brief summary of my convention:

 

4-card majors

weak 1NT

multi-2♦ and Weak two suiters

non-solid pre-empts

 

aggressive general bidding and competitive auctions

conservative towards game and penalty doubles

 

To get the results out of GIB I had to replay all of the hands in a teams match (tournament) and see what the other table result was. I did this quickly by passing in every board and manually recording the results. I cannot see an automatic way of finding the result or any way of seeing the bidding and play at the other table and even the list of results that came up would not let me cut and paste.

[bestguru]

Could you use the command line version of GIB and redirect the output to files for later parsing? I will probably be doing somethign similar soon, so let me know if you want help with the scripts. I can probably set something up over the weekend.

[Cascade]

Can you give me an example of the script that would be needed and the hand format so that it would play multi hands.

 

TIA

[han]

An additional advantage when opening (especially preempting) in first seat is that RHO, who is in ballancing seat, can still have opening strength, so the range of his overcalls/doubles become very wide.

Good comment, except I disagree with the "especially preempting". It seems to me that the ranges for overcalling in direct seat or in balacing seat are much wider over 1-level openings than over preempts. In some ways, low level openings (like 1C) are excellent preempts.

[bestguru]

scripting bridge.exe (the command line version of gib) isn't working out for me. I am missing files and the command line switches do not seem to match the documentation I found on the web.

 

Here is the doc I found. maybe you will have more luck with it than me.

 

http://www.kolumbus.fi/sackab/kgb/all_htm/gibdos.htm

[tysen2k]

Has anyone run more hands? Is the dealer still at a disadvantage? How about trying different convention cards?

 

This seemed like a real puzzle to me and I'm interested to see if it gets sorted out.

[Gerben47]

It would be interesting to see if 1st and 3rd do better on the hands where 1st seat passes or on the hands 1st seat bids something. Now don't tell me you know the answer - we thought we knew the answer before as well.

[MickyB]

Even better, a breakdown by the call made in first seat.

[Gerben47]

Then you get in the small number statistics again. Besides those numbers depend a lot on the system used.

[Cascade]

Has anyone run more hands? Is the dealer still at a disadvantage? How about trying different convention cards?

 

This seemed like a real puzzle to me and I'm interested to see if it gets sorted out.

I am away from home at the moment. I will try and investigate further when i get home. I might need to write some code to automate GIB. If anyone has some experience with this I would accept help. I remember I had some problems last time I tried this.