Advanced Euchre QUIZ - PART 2

Ask questions, discuss and debate your strategies, euchre polls and more
irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Re: Advanced Euchre QUIZ - PART 2

Unread post by irishwolf » Mon Apr 25, 2022 4:54 pm

Ya, you gotta love it!

And the 'key' to ensure not just an exercise in futility is to hard code (and rewire if need be) ALL the hands or at least the summation of the similarities so one can retrieve it on demand in real time! And that includes any possible variations and the exceptions. And you can't just read it and think you got it! Good euchre players have thousands of hands in Working Memory. This is way above Social Euchre!

It's what makes euchre interesting.



raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Tue Apr 26, 2022 8:42 am

Moving on to Wes' Quiz Question #12:
Your team is up 8-1, and you're sitting in S2 with 9S + Q-10H + A-KC (9H turned). What do you do?

I simulated 100,000 hands, and found:
bid H, R1: (12,222 / 53,072 / 31,663) [sweep / 1 pt / euchred] EV = +0.15
pass, R1 (bid C, R2): EV = +0.09

Here's the breakdown if you pass:
S3 calls 4,871 hands, R1, for an EV of -0.98 [perspective of S2/S4]
S4 calls 45,732 hands, R1, for an EV of +1.20
S1 calls 36,785 hands, R2, for an EV of -1.08 [again, from perspective of S2/S4]
S2 calls 9,574 hands, R2, for an EV of -0.18 [if S2 passes, R2, EV drops to -0.65]

So, by a small but statistically significant margin, S2 should call this hand.

But, there are the usual caveats. My program is making assumptions about when S3, S4, S1 and S2 will call (if you pass). With a margin of just 0.06, it's possible that changing those assumptions could alter the result. And my program may not be playing all the hands optimally (I am continually working on improving it when I discover poor play).

That said, I do believe I have corrected all the egregious errors, and, while difficult to judge, I think that the remaining errors are numerous but individually quite infrequent, so that their cumulative effect on a particular scenario may move the needle by 0.04 or 0.05 in EV, in extremis, but unlikely more. [Case in point, a previous hand from this post where I was finding a higher euchre rate bidding wp than bidding alone. I found a couple of clear mis-plays and corrected them, which moved the EV by 0.04. Important but not critical - i.e., doesn't change the outcome of the analysis].

In this particular hand, my program passes, so I will now investigate if I can correct that error. There is a trade-off between complexity of program and optimization of result, so it's possible I will conclude this is an outlier case and accept the sub-optimal result - loss of 0.06 EV in a scenario which may arise once in 100,000 hands. But it may also prove to be an easy fix.

I was thinking of investigating whether this hand should be bid if the score is 8-8 or 8-9, but that very much changes the bidding dynamic, and one thing my program does not yet address is how score affects the bidding decision. CRITICALLY important in actual play, but I'm looking at the "law of large numbers" result, as Irish calls it. That's the core; I can consider adjusting for score once I get the basics down.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Tue Apr 26, 2022 2:16 pm

And Wes' Quiz Question #13:

You are dealer, with your team winning 9-0. You are dealt 9S + K-QH + K-9D, with the 9C turned. Do you bid or pass?

Here's what I found: (100,000 hands)

bid: [0 / 0.002 / 0.131 // 0 / 0.867 / 0] EV = -1.60
pass: [0.010 / 0.103 / 0.157 // 0.471 / 0.147 / 0.112] EV = -0.81

The 6 numbers separated by / are the odds, respectively, of scoring +4 / +2 / +1 // -1 / -2 / -4 pts. (negative pts. means the opponents score those points). In other words, it is the probabilistic distribution of outcomes, in terms of who scores what points. To interpret the first row of numbers as an example (when S4 bids), they have a 0 chance of scoring 4 pts., a 0.2% chance of getting a sweep, a 13.1% chance of earning a point, and a 86.7% chance of getting euchred. The opponents have 0 chance of scoring 1 or 4 pts.

So clearly with this hand, "in the long run", it is better to pass (from the EV values). The only reason to bid is to prevent the opponents from scoring 4 pts. Unfortunately, my analysis shows that you are paying too high a price for that protection, and are actually reducing your very high probability of winning the game. This is something I have brought up before, but I can't seem to convince Wes, Irish, and I think perhaps Ed or the validity of this point (my apologies in advance if I am misrepresenting your point of view).

Looking at the above figures (which I will assume you agree with - though perhaps we will squabble about the 3rd decimal place), when bidding you have 13.2% chance of scoring a point and ending the game, but a very high chance of simply giving the opponents 2 pts (that's why it's called donating). When you pass, you have a 27% chance of scoring a point and ending the game - twice as high! - and the opponents have only a 11.2% chance of scoring 4 pts. (and indeed only a 25.9% chance of scoring 2 or more pts). In my eyes that's a far more favorable outcome.

But you don't have to share my assessment. We need only define one more point probability distribution, that for a completely random hand (in other words, all subsequent hands in this game), and with these three data sets we can calculate the exact probability of winning the game. It's just arithmetic, though easier done with a computer program due to the large number of paths possible to arrive at a score of 10, 11, 12 or 13 pts. by one team.

[Actually, I have refined the calculation further by figuring out the point probability distribution for 3 addition cases: when one or the other or both teams are within 2 pts. of winning, and so don't bid alone.]

Here are the point probability distributions for a random game, calculated by looking at the outcome of 10 million randomly dealt hands: (viewed from the perspective of the dealing team).

normal: [0.067 / 0.141 / 0.459 // 0.187 / 0.121 / 0.025]
dealing team doesn't bid alone: [0 / 0.223 / 0.448 // 0.187 / 0.117 / 0.025]
non-dealing team doesn't bid alone: [0.067 / 0.139 / 0.049 // 0.184 / 0.151 / 0]
neither team bids alone: [0 / 0.222 / 0.449 // 0.182 / 0.147 / 0]

Analyzing the above hand using these point probability distributions, as well as those from bidding or passing the original hand, the odds of winning the game are 98.5% if you pass and 98.2% if you bid. Or, from the perspective of your opponents, they have a 1.5% chance of winning the game if you pass, but a 1.8% chance of winning the game if you bid. So you increase their chances by 20% by bidding! Don't do them that favor!

Tweaking the various point probability distributions will have an effect of the calculated game-winning chances, but it will be tiny. I can really see no justification for bidding here.

Just to play around, I tried changing the initial score.

odds of dealing team winning the game
1) winning 9-6:
bid: 67.9%
pass: 74.2%

2) tied 9-9:
bid: 13.3%
pass: 27.0%

3) tied 6-6:
bid: 28.6%
pass: 39.0%

It's looking to me like it's always better to pass with this hand, whatever the score.
Last edited by raydog on Sun May 01, 2022 10:38 pm, edited 1 time in total.

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Tue Apr 26, 2022 2:27 pm

raydog wrote:
Tue Apr 26, 2022 8:42 am
Moving on to Wes' Quiz Question #12:
Your team is up 8-1, and you're sitting in S2 with 9S + Q-10H + A-KC (9H turned). What do you do?

I simulated 100,000 hands, and found:
bid H, R1: (12,222 / 53,072 / 31,663) [sweep / 1 pt / euchred] EV = +0.15
pass, R1 (bid C, R2): EV = +0.09

Here's the breakdown if you pass:
S3 calls 4,871 hands, R1, for an EV of -0.98 [perspective of S2/S4]
S4 calls 45,732 hands, R1, for an EV of +1.20
S1 calls 36,785 hands, R2, for an EV of -1.08 [again, from perspective of S2/S4]
S2 calls 9,574 hands, R2, for an EV of -0.18 [if S2 passes, R2, EV drops to -0.65]

So, by a small but statistically significant margin, S2 should call this hand.

But, there are the usual caveats. My program is making assumptions about when S3, S4, S1 and S2 will call (if you pass). With a margin of just 0.06, it's possible that changing those assumptions could alter the result. And my program may not be playing all the hands optimally (I am continually working on improving it when I discover poor play).

That said, I do believe I have corrected all the egregious errors, and, while difficult to judge, I think that the remaining errors are numerous but individually quite infrequent, so that their cumulative effect on a particular scenario may move the needle by 0.04 or 0.05 in EV, in extremis, but unlikely more. [Case in point, a previous hand from this post where I was finding a higher euchre rate bidding wp than bidding alone. I found a couple of clear mis-plays and corrected them, which moved the EV by 0.04. Important but not critical - i.e., doesn't change the outcome of the analysis].
Good stuff. Up 8-1, even if calling was slightly -EV vs passing, I would still recommend calling with this hand to control variance. We block nothing in the 2nd round. Don't give S1 the chance to be a hero. But as I suspected this call is straight up +EV so it's just a standard call at almost any score except when down 3+, then I think the argument for passing hoping our P has that magical loner is compelling but that's difficult if not impossible to prove. The other possible exception is what you mentioned below:
raydog wrote:
Tue Apr 26, 2022 8:42 am
I was thinking of investigating whether this hand should be bid if the score is 8-8 or 8-9, but that very much changes the bidding dynamic, and one thing my program does not yet address is how score affects the bidding decision. CRITICALLY important in actual play, but I'm looking at the "law of large numbers" result, as Irish calls it. That's the core; I can consider adjusting for score once I get the basics down.
8-8 and 9-8 are the mindf**k scores where I'm not sure what to do. If you solved that spot I would be grateful. Down 8-9 I feel pretty good about making this call. It's the best we got. Our P passing will usually = game over.

One more thing I wanted to add:
raydog wrote:
Tue Apr 26, 2022 8:42 am
In this particular hand, my program passes, so I will now investigate if I can correct that error. There is a trade-off between complexity of program and optimization of result, so it's possible I will conclude this is an outlier case and accept the sub-optimal result - loss of 0.06 EV in a scenario which may arise once in 100,000 hands. But it may also prove to be an easy fix.
I agree that the tradeoff may not be worth it. I will say this much tho, whenever S2 has 2 trump + a green ace + a void & S2 doesn't block much, it's probably a +EV call.

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Tue Apr 26, 2022 2:35 pm

raydog wrote:
Tue Apr 26, 2022 2:16 pm
And Wes' Quiz Question #13:

You are dealer, with your team winning 9-0. You are dealt 9S + K-QH + K-9C, with the 9C turned. Do you bid or pass?

Here's what I found: (100,000 hands)

bid: [0 / 0.002 / 0.131 // 0 / 0.867 / 0] EV = -1.60
pass: [0.010 / 0.103 / 0.157 // 0.471 / 0.147 / 0.112] EV = -0.81

The 6 numbers separated by / are the odds, respectively, of scoring +4 / +2 / +1 // -1 / -2 / -4 pts. (negative pts. means the opponents score those points). In other words, it is the probabilistic distribution of outcomes, in terms of who scores what points. To interpret the first row of numbers as an example (when S4 bids), they have a 0 chance of scoring 4 pts., a 0.2% chance of getting a sweep, a 13.1% chance of earning a point, and a 86.7% chance of getting euchred. The opponents have 0 chance of scoring 1 or 4 pts.

So clearly with this hand, "in the long run", it is better to pass (from the EV values). The only reason to bid is to prevent the opponents from scoring 4 pts. Unfortunately, my analysis shows that you are paying too high a price for that protection, and are actually reducing your very high probability of winning the game. This is something I have brought up before, but I can't seem to convince Wes, Irish, and I think perhaps Ed or the validity of this point (my apologies in advance if I am misrepresenting your point of view).

Looking at the above figures (which I will assume you agree with - though perhaps we will squabble about the 3rd decimal place), when bidding you have 13.2% chance of scoring a point and ending the game, but a very high chance of simply giving the opponents 2 pts (that's why it's called donating). When you pass, you have a 27% chance of scoring a point and ending the game - twice as high! - and the opponents have only a 11.2% chance of scoring 4 pts. (and indeed only a 25.9% chance of scoring 2 or more pts). In my eyes that's a far more favorable outcome.

But you don't have to share my assessment. We need only define one more point probability distribution, that for a completely random hand (in other words, all subsequent hands in this game), and with these three data sets we can calculate the exact probability of winning the game. It's just arithmetic, though easier done with a computer program due to the large number of paths possible to arrive at a score of 10, 11, 12 or 13 pts. by one team.

[Actually, I have refined the calculation further by figuring out the point probability distribution for 3 addition cases: when one or the other or both teams are within 2 pts. of winning, and so don't bid alone.]

Here are the point probability distributions for a random game, calculated by looking at the outcome of 10 million randomly dealt hands: (viewed from the perspective of the dealing team).

normal: [0.067 / 0.141 / 0.459 // 0.187 / 0.121 / 0.025]
dealing team doesn't bid alone: [0 / 0.223 / 0.448 // 0.187 / 0.117 / 0.025]
non-dealing team doesn't bid alone: [0.067 / 0.139 / 0.049 // 0.184 / 0.151 / 0]
neither team bids alone: [0 / 0.222 / 0.449 // 0.182 / 0.147 / 0]

Analyzing the above hand using these point probability distributions, as well as those from bidding or passing the original hand, the odds of winning the game are 98.5% if you pass and 98.2% if you bid. Or, from the perspective of your opponents, they have a 1.5% chance of winning the game if you pass, but a 1.8% chance of winning the game if you bid. So you increase their chances by 20% by bidding! Don't do them that favor!

Tweaking the various point probability distributions will have an effect of the calculated game-winning chances, but it will be tiny. I can really see no justification for bidding here.

Just to play around, I tried changing the initial score.

odds of dealing team winning the game
1) winning 9-6:
bid: 67.9%
pass: 74.2%

2) tied 9-9:
bid: 13.3%
pass: 27.0%

3) tied 6-6:
bid: 28.6%
pass: 39.0%

It's looking to me like it's always better to pass with this hand, whatever the score.
You're definitely not misrepresenting my view. :) I'll be honest. There is no simulator in the world that can sway my decision point here. I'm never passing that hand in this spot. This is a MUST dealer donate for me. If god came down and told me I was wrong then I would change. Problem is I'm an atheist!

Either way, I think it's perfectly rational for others to find your actual analysis more compelling than my bravado.

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Tue Apr 26, 2022 3:17 pm

raydog wrote:
Mon Apr 25, 2022 4:29 pm
Ah, Wes, for every response you always have 3 new questions!
:-) Always keep in mind tho that I am merely posing those questions theoretically to give others an idea what we should be thinking about next. I'm not requesting you actually solve them. I mean yes I absolutely love when you do this work and I get really excited to see the answers but at the same time I know you have a life. I only want you to do this stuff when youre inspired, not when I ask you to becuz then it will become work and I hate work!!!
raydog wrote:
Mon Apr 25, 2022 4:29 pm
Regarding the hands you proposed in you latest post: S = simulator; W = Wes

S4, A-10H + A-KD + AC (JS)
pass: EV = +0.40 S
bid: EV = +0.35 W
Good stuff. I accept that outcome. I still think this is a good hand to call with if we have a decent lead but that's hard to prove. Either way, you now have a strong argument for passing R+0+3A since if a hand that blocks no suits is a -EV call, it pretty much follows any other hand from this configuration will be a -EV call also.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S4, A-10H + A-KD + 9C (JS)
pass: EV = +0.06 S
bid: EV = -0.06 W
I accept this answer too and basically same story as above.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S4, Q-10-9H + AD + 9C (JS)
pass: EV = -0.44 S
bid: EV = -0.38 W
Hey I found one where calling beats out passing! Sweet! I see no inconsistency with this hand being a call and the others above being a pass either. With the former hands, having multiple aces gives us better defense when we pass and that alone can explain away this apparent paradox.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S4, JS + KH + JD + 9-10C (AS)
pass: EV = +0.24
bid: EV = +0.33 S
[I should correct a previous error by saying here that my program will pass with R+1 trump and all suits blocked IF the 2nd trump is K or lower, otherwise bid; so here bids because the 2nd trump is an A]

S4, JS + KH + JD + 9-10C (KS)
pass: EV = +0.26 S
bid: EV = +0.23
This finding is very important to me. I've been in arguments in the past over this configuration. I'm very glad this can finally be put to rest. Passing is correct with R+1 when we have all suits blocked except when we have J-A. Awesome. Really appreciate this Ray.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S3, K-Q-10S + A-KC (9S)
pass: EV = +0.49 S, W
bid: EV = +0.40
Awesome. As I suspected. When S3's outside suited ace is in Next, S3 should pass from this configuration. Awesome.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S3, K-Q-10S + AD + KH (9S)
pass: EV = +0.21 S, W
bid: EV = +0.12
Another important finding. If we don't have that suited ace, we should pass.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S4, A-KH + A-KC + 9D (QH)
alone: EV = +0.74 W
wp: EV = +0.81 S
I'm a little surprised by this result but I accept it. You have now basically solved all 3 trump non-bower possible loner hands. Why did I use the qualifier "basically"? Because you've established that this hand is a loner:

(Card_A-H) (Card_K-H) (Card_Q-H) (Card_A-C) (Card_A-S)

And I'm assuming that also means this hand is a +EV loner:

(Card_A-H) (Card_K-H) (Card_9-H) (Card_A-C) (Card_A-S)

But we still have to test this hand to rule out other possible loners:

(Card_A-H) (Card_Q-H) (Card_10-H) (Card_A-C) (Card_A-S)

I'm betting that hand is not a loner but that's pretty much the last hand you'd have to test to have this spot completely solved. Unless of course that hand IS a loner, then you'd have to test AhTh9hAcAs.

Great work Ray. I've always been going alone with AhKhQhAcKc. That's a leak I can clean up.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
S4, A-KH + A-9C + 9D (QH)
alone: EV = +0.49
wp: EV = +0.81 S, W
I would say that you didn't have to test this one given that once we know AhKhQhAcKc isn't a loner it follows that AhKhQhAc9c is not a loner BUT I am very glad you did! That's a pretty large gap in EV!! I never would've predicted that. Sometimes I get tempted to gamble and "go for it" even at a neutral score but I will never get tempted to go for it with this holding again unless my team is down a lot.
raydog wrote:
Mon Apr 25, 2022 4:29 pm
All this analysis basically shows that my program is generally internally consistent - it mostly bids correctly according to how it sees the results turning out. But the assumptions made as to how those bids are made by each seat - and whether they reflect how human players bid in real games - is, as always, debatable. Though I contend they SHOULD usually be bidding the way I have programmed (score and specific knowledge of other players' tendencies notwithstanding). And, my program still has its flaws.
Your program is unequivocally awesome as far as I'm concerned. Now it is true that the more you optimize your program the less it will reflect reality becuz frankly must humans suck at euchre, but I would not worry about that "problem". Your approach is extremely useful and important as it will give us many insights into this game and help guide us through those many grey areas. And once we know what the correct play is, it's not hard to make adjustments on our own for when we have a weak P or are up against weak opponents. So even tho your program is not simulating a weak game texture it will still prove invaluable in helping us guide our play in such a game.

irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Unread post by irishwolf » Tue Apr 26, 2022 11:16 pm

RAY,

I find your post on this topic (Dealer KH QH KC XX 9S - 9C UP) most interesting: (you did not say what the slough would be?) But that is not my question.

Basically your Post was about the 'donate' from the Dealer.
But let's change the scenario, from S1 seat's perspective (GIVING HIM THE WEAK HAND). S1 is now presented with two issues is that no one has passed yet. S2 is one and then the Dealer and is discard advantage. How much does this change the outcome 9 to 0?

Now put the score at 9 to 7 and then 9 to 6 - as standard practice to donate. Analyze that hand but give it to S1 with first, Jack up? What's the probabilities? Compared this to any other Card up? Is it wise to donate? I think it is with the Jack up but not otherwise in the long run. This goes against tradition but I have often differed that it was good strategy. Especially, putting the Opponent to 9 & 9. And I suggest 9 & 8 is slightly different. The comeback for opponents when putting them to 9 is around 30% from my analysis. That is all with you loser hand one King doubleton.

Then what I call the real Acid test: What if you had the hand at S1 singleton Ace and give S1 another Ace but no sure stopper and even give him one trump. Two potential stoppers. So then using your simulator is it wise to Donate? I think borderline to donate even with Jack up here. What you say?

You said (see below 9 to 0) and I agree with your results. However, the rationale for those in the lead is that I would rather give away 1 point, no big deal against the probability of scoring 4. Understanding that I am going to win anyway. How many times having observed two loners in a row?

---------------

" You are dealer, with your team winning 9-0. You are dealt 9S + K-QH + K-9C, with the 9C turned. Do you bid or pass?

Here's what I found: (100,000 hands)

bid: [0 / 0.002 / 0.131 // 0 / 0.867 / 0] EV = -1.60
pass: [0.010 / 0.103 / 0.157 // 0.471 / 0.147 / 0.112] EV = -0.81

Just to play around, I tried changing the initial score.

odds of dealing team winning the game
1) winning 9-6:
bid: 67.9%
pass: 74.2%

2) tied 9-9: (Little surprised here.)
bid: 13.3%
pass: 27.0%

3) tied 6-6:
bid: 28.6%
pass: 39.0%

It's looking to me like it's always better to pass with this hand, whatever the score.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Wed Apr 27, 2022 3:07 pm

Irish, I haven't yet read your full post, I stopped when I found a CRITICAL error in MY post! Dealer is dealt the K-QH + K-9D + 9S (9C turned). That's the original hand from Wes' quiz and the one I analyzed; the slough is obviously the 9S.

Apologies for that typo, which completely changes the hand!

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Wed Apr 27, 2022 5:28 pm

Irish, I'm going to assume for the moment that you understood I made a typo, and that the real hand was K-QH + K-9D + 9S (9C).

1) "you did not say what the slough would be?" 9S

2) "let's change the scenario, from S1 seat's perspective (GIVING HIM THE WEAK HAND)" So I give the above hand to S1, and test if she should donate. I suppose I also need to switch the score, so that S1/S3 are now winning 9-0? Otherwise donation doesn't make sense.

3) and then I should test the scenario in 2) but with the score 9-6 and 9-7 (in favor of S1/S3), first w/ the JC turned, then with a different trump. (this one I think is clear)

4) "the real Acid test": how about I give S1 AS + A-9H + KD + QC, with either the JC or the 9C turned (I'll test both), donate or not. Score 9-6 or 9-7 (S1/S3 winning). Is that OK?

[BTW, this isn't just "moving the goal posts", it's changing stadiums! But it's a valid corollary question.]

5) you noted you were surprised at the results I found when the score was tied 9-9, but that simply reflects the point probability distribution I showed earlier. When bidding, 0.2% chance of sweeping and 13.1% chance of scoring a point = 13.3% chance of winning. When passing, sum of the chances of winning 4pts, 2pts or 1pt = 27.0%. Someone will win on this hand, and S2/S4 simply score a point twice as often if S4 passes, R1.

irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Unread post by irishwolf » Wed Apr 27, 2022 11:26 pm

RAY

My comments in Blue:

Irish, I'm going to assume for the moment that you understood I made a typo, and that the real hand was K-QH + K-9D + 9S (9C). Did not go back but understood with what S4 was left with. So all good.

1) "you did not say what the slough would be?" 9S

2) "let's change the scenario, from S1 seat's perspective (GIVING HIM THE WEAK HAND)" So I give the above hand to S1, and test if she should donate. I suppose I also need to switch the score, so that S1/S3 are now winning 9-0? Otherwise donation doesn't make sense. CORRECT

3) and then I should test the scenario in 2) but with the score 9-6 and 9-7 (in favor of S1/S3), first w/ the JC turned, then with a different trump. (this one I think is clear) YES

4) "the real Acid test": how about I give S1 AS + A-9H + KD + QC, with either the JC or the 9C turned (I'll test both), donate or not. Score 9-6 or 9-7 (S1/S3 winning). Is that OK? YES - MOST IMPORTANT. IF YOU POST WHAT I THINK IS COMING, YOU TURN OLD RULES ON ITS EAR!

[BTW, this isn't just "moving the goal posts", it's changing stadiums! But it's a valid corollary question.] AGREED

5) you noted you were surprised at the results I found when the score was tied 9-9, but that simply reflects the point probability distribution I showed earlier. When bidding, 0.2% chance of sweeping and 13.1% chance of scoring a point = 13.3% chance of winning. When passing, sum of the chances of winning 4pts, 2pts or 1pt = 27.0%. Someone will win on this hand, and S2/S4 simply score a point twice as often if S4 passes, R1. YES - UNDERSTOOD.

I am on the road so I made it short. All good stuff.

irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Unread post by irishwolf » Thu Apr 28, 2022 10:36 pm

RAY,

You might want to clarify Post results for 13.

You are dealer, with your team winning 9-0. You are dealt 9S + K-QH + K-9C, with the 9C turned. Do you bid or pass?

Here's what I found: (100,000 hands)
S4 - 9S + K-QH + K-9C, with the 9C
bid: [0 / 0.002 / 0.131 // 0 / 0.867 / 0] EV = -1.60
pass: [0.010 / 0.103 / 0.157 // 0.471 / 0.147 / 0.112] EV = -0.81

Was this correct for: KC 9C KD KH QH (9S SLOUGH)? However, if passing for this hand results then 13. KH QH KD 9D 9C (9S SLOUGH) has to be the same results (Pass) as well.

However, switching things around I want to see what the results would be.

IRISH

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Sun May 01, 2022 10:37 pm

So, I ran your scenarios, Irish. My methodology was as follows. I set the hand for S1 and the turn card, all other 18 cards random (100,000 hands). Then I compared S1 bidding, R1, to passing. I noted the point probability distribution in each case, then plugged those numbers into a 2nd program to determine the probability of eventually winning the game.

Here are the results:
S1 has K-QH + K-9D + 9S (9C turned)

At the following scores, here are the odds of winning the game if passing or bidding.
9-0 (S1/S3 winning)
pass: 98.57%
bid: 98.50%
call this a virtual dead heat.

9-6
pass: 73.87%
bid: 76.56%
better to bid (donate)

9-7
pass: 67.32%
bid: 71.34%
better to bid

Same hand, but with JC turned
9-6
pass: 67.22%
bid: 74.38%
better to bid

9-7
pass: 60.44%
bid: 68.68%
better to bid

Then I gave S1 a slightly stronger hand: AS + A-9H + KD + QC (2 off-suit aces + a trump)
At the following scores, here are the odds of winning the game if passing or bidding.

w/ 9C turned
9-6
pass: 82.57%
bid: 79.87%
better to pass

9-7
pass: 75.76%
bid: 75.39%
better to pass (but very close)

w/ JC turned
9-6
pass: 79.57%
bid: 76.75%
better to pass

9-7
pass: 72.07%
bid: 71.57%
better to pass (but close)

Interpretation: With a really poor hand, it's better to donate from S1, especially at scores of 9-6 and 9-7 in your favor. But "really poor hand" is open to interpretation. Just having a couple of off-suit aces and a trump elevates the hand to "bad but not awful", and warrants sucking it up and just passing.

This analysis assumes that my data is correct, and seems quite straightforward. However, if my point probability distributions are incorrect (due to my program not bidding or playing the given hands optimally), that could change the results in some cases (where it is close). I am not expert enough to know what the "old rules" are, but these results seem quite reasonable to me. Perhaps players have been taught that they need to have all other suits blocked to NOT donate in this position; I have always been skeptical of that very draconian threshold, and am quite happy with the rule" truly awful hand: donate; merely bad hand: pass". With the determination of "truly awful" vs. "merely bad" being based on # of aces, # of trump and # of voids (assuming you have a trump to potentially take advantage).

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Sun May 01, 2022 10:40 pm

Also, I corrected my typo in the question 13 results, changing K-9C to K-9D, as it should have been. I hope that addresses your second post, Irish.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Sun May 01, 2022 11:56 pm

Moving on to Quiz questions #14 and #15 (not very controversial).

#14: Score 0-0, in S1 with 10S + 10H + J-9D + JC, 9S turned. Everyone passed, R1, what do you do, R2?

I tested 100,000 hands, and found the following:
pass: EV = -0.15
bid D: EV = -0.53
bid C: EV = -0.69

With all suits blocked here (so opponents can at most score 1 pt, and potentially be euchred), it's clearly best to pass.

As a side note, I simulated the case where, with all cards randomized, the bidding goes to the 2nd round and S1 has 2 of 3 suits blocked, and found it was better to bid by an EV of +0.02. To me, this suggests that there are cases where S1 should still pass when they have only 2 of the 3 biddable suits blocked. But I have not investigated further into the actual circumstances where that is true. The "all suits blocked" rule stands, but there may also be a "sometimes with only 2 suits blocked" rule.
__________________________________________

#15: winning 3-2, in S2 with 10S + K-10D + A-JC, KH turned. Everyone passes, R1, and S1 passes, R2.

I tested 100,000 hands, and found the following:
pass: EV = +0.32
bid C: EV = +0.63
bid S: EV = +0.23

A clear winning strategy here.

irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Unread post by irishwolf » Mon May 02, 2022 11:52 am

Ray,

Thanks for running the test hand. As for your simulator, S1 should lead the AH here and save KD/AS to the end. It stops a Diamond but not AD dbltn as well as a Spade. Sloughing KD on trick 4 tells S3 to saving diamonds if he has it. I think pretty much straight forward for the Dealer's play. So I trust your results.

To me this is confirming and refining the Donation when at 9 to 6/7. And I am in TOTAL agreement! This includes your comments as well. So you are not the only one being Skeptical. ha ha This is critical stuff, and to me one of the most interesting testing because you are running so many hands. I also tested, far fewer hands but I can better see how the cards are played and with the same results. I had bigger spread tho when Jack is up but in the same direction. With 9C up, just not worth the Donating and I have stated here on OE. The two are apples and oranges, IMO.

So now you have set an old Euchre Must do on its ear. Here it is:

When at 9 to 6/7 Eldest, S1 not having SURE stopper has been almost a LAW in Euchre boing back 160 years (Meehan and many others). Only questioned as not sure by Dwight & Leeds (1887). But that was a different time when the game was played with 32 cards and 11 cards to the Stock buried which completely changes the chance of the Opponents having a loner when at 9 to 6/7. It was Natty, see his Columns on Donating when at 9 to 6/7 for the modern game when 24 cards. Old methods of how to play passed on as MUST. Natty says Must donate and that winning with a ~65% chance is good enough for him. But what if you can improve that when you have two aces. Because now the Dealer has to have all trump or that other Ace? I disagree and think you can improve that 65% to 75 - 80% by refining that Donation. Even good euchre players get scared at 9 to 6/7 and donate being satisfied giving away 1 point in theory than losing the game to a loner.

Thanks,

IRISH

________________________________________________________________________________

RAY TESTS RESULTS:
Then I gave S1 a slightly stronger hand: AS + A-9H + KD + QC (2 off-suit aces + a trump)
At the following scores, here are the odds of winning the game if passing or bidding.

w/ 9C turned
9-6
pass: 82.57%
bid: 79.87%
better to pass

9-7
pass: 75.76%
bid: 75.39%
better to pass (but very close)

w/ JC turned
9-6
pass: 79.57%
bid: 76.75%
better to pass

9-7
pass: 72.07%
bid: 71.57%
better to pass (but close)

Interpretation: With a really poor hand, it's better to donate from S1, especially at scores of 9-6 and 9-7 in your favor. But "really poor hand" is open to interpretation. Just having a couple of off-suit aces and a trump elevates the hand to "bad but not awful", and warrants sucking it up and just passing.

This analysis assumes that my data is correct, and seems quite straightforward. However, if my point probability distributions are incorrect (due to my program not bidding or playing the given hands optimally), that could change the results in some cases (where it is close). I am not expert enough to know what the "old rules" are, but these results seem quite reasonable to me. Perhaps players have been taught that they need to have all other suits blocked to NOT donate in this position; I have always been skeptical of that very draconian threshold, and am quite happy with the rule" truly awful hand: donate; merely bad hand: pass". With the determination of "truly awful" vs. "merely bad" being based on # of aces, # of trump and # of voids (assuming you have a trump to potentially take advantage).
Last edited by irishwolf on Tue May 03, 2022 9:23 am, edited 1 time in total.

Tbolt65
Posts: 820
Joined: Sun Jun 16, 2019 9:14 pm
Location: Las Vegas

Unread post by Tbolt65 » Tue May 03, 2022 2:01 am

raydog wrote:
Sun May 01, 2022 10:37 pm
So, I ran your scenarios, Irish. My methodology was as follows. I set the hand for S1 and the turn card, all other 18 cards random (100,000 hands). Then I compared S1 bidding, R1, to passing. I noted the point probability distribution in each case, then plugged those numbers into a 2nd program to determine the probability of eventually winning the game.

Here are the results:
S1 has K-QH + K-9D + 9S (9C turned)

At the following scores, here are the odds of winning the game if passing or bidding.
9-0 (S1/S3 winning)
pass: 98.57%
bid: 98.50%
call this a virtual dead heat.

9-6
pass: 73.87%
bid: 76.56%
better to bid (donate)

9-7
pass: 67.32%
bid: 71.34%
better to bid

Same hand, but with JC turned
9-6
pass: 67.22%
bid: 74.38%
better to bid

9-7
pass: 60.44%
bid: 68.68%
better to bid

Then I gave S1 a slightly stronger hand: AS + A-9H + KD + QC (2 off-suit aces + a trump)
At the following scores, here are the odds of winning the game if passing or bidding.

w/ 9C turned
9-6
pass: 82.57%
bid: 79.87%
better to pass

9-7
pass: 75.76%
bid: 75.39%
better to pass (but very close)

w/ JC turned
9-6
pass: 79.57%
bid: 76.75%
better to pass

9-7
pass: 72.07%
bid: 71.57%
better to pass (but close)

Interpretation: With a really poor hand, it's better to donate from S1, especially at scores of 9-6 and 9-7 in your favor. But "really poor hand" is open to interpretation. Just having a couple of off-suit aces and a trump elevates the hand to "bad but not awful", and warrants sucking it up and just passing.

This analysis assumes that my data is correct, and seems quite straightforward. However, if my point probability distributions are incorrect (due to my program not bidding or playing the given hands optimally), that could change the results in some cases (where it is close). I am not expert enough to know what the "old rules" are, but these results seem quite reasonable to me. Perhaps players have been taught that they need to have all other suits blocked to NOT donate in this position; I have always been skeptical of that very draconian threshold, and am quite happy with the rule" truly awful hand: donate; merely bad hand: pass". With the determination of "truly awful" vs. "merely bad" being based on # of aces, # of trump and # of voids (assuming you have a trump to potentially take advantage).

9-6 and 9-7 are crucial times in the game of euchre. With out having Two sides aces, the protected Left-x or The Right bauer Jack. Or say two medium trump through 2 high trump plus an off Ace. You should be ordering all other hands at that score. Realistically you block a lot of hands and like IrishWolf mentions. They need to have that perfect hand to beat you.

If I may add on donations. I believe it is crucial to donate at the proper times and for game management but I have learned to curb on donating every I don't have One of the above mentioned hands I spoke of. I just think doing it too much you give unnecessary points away. Sure you can once in a while donate at various scores on a whim every now and then and it may or may not pay off. These whims like any other plays must be properly gauged and weighted for various scenarios.

Tbolt65
Edward

irishwolf
Posts: 1319
Joined: Tue Apr 24, 2018 9:33 pm

Unread post by irishwolf » Tue May 03, 2022 10:10 am

Yep, I agree here as Ed says, "f I may add on donations. I believe it is crucial to donate at the proper times and for game management but I have learned to curb on donating every I don't have One of the above mentioned hands I spoke of. I just think doing it too much you give unnecessary points away. Sure you can once in a while donate at various scores on a whim every now and then and it may or may not pay off. These whims like any other plays must be properly gauged and weighted for various scenarios."


To add to what Ed says, I donate at any score when I sense the opponent has a loner. When the Jack is up is when you have to be more cautious as it doubles the probabilities to a loner attempt. And as long as I block a successful loner only 1 of 4, and 1 would be 2 point sweep as a minimum, I break even. So the math is: 1 loner, 1 sweep, & 2 - 1 pointers = 8 pts compared to donating 4 times = 8 pts. To win in euchre you must block loners as the team that scores a loner and the other side does not will win the game 80% of the time. It's hard to block 1st seat loners as he gets to bid first. I am not sure how that fits Ray calculator on winning.

I satisfied with my methodology and am not changing.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Mon May 23, 2022 12:54 pm

To keep the discussions flowing (and those euchre neurons firing), I've submitted Wes's euchre quiz question #16 to my simulator, to see what it spits out.

You're in 3rd seat, tied at 6-6, with K-QS + JH + AD + JC (AS turned). What do you bid?

I tested 100,000 hands, here's what I found:

bid: 8,100 / 50,725 / 27,904 [2 pts / 1 pt/ euchred] EV = +0.13
pass: EV = +0.91 (euchre S4 calls often; have excellent support for S1,R2 next calls [R in trump + top card in 2 off-suits]; have decent D call if it gets to you, R2)

No great surprise here, definitely better to pass.
_________________________

I also looked at the amended scenario where S3 doesn't have all the other suits blocked (I substituted the QH for the JH):

bid: EV = +0.15 (logically a shade higher, with the QH being a shade higher than the JH)
pass: EV = +0.69 (also logical, as S2 does better if they get to bid, R2; and S3 does worse if gets to bid, R2)

So, the same result.
_________________________

I noticed that in this 2nd scenario, my program had S3 bidding C, R2 (if it got that far), for an EV of -0.18. Is that the correct call, knowing S1 (partner) is an excellent player and understands the importance of a next call?

I compared this to passing, and found the following (100,000 tested, only 5,300 made it to this decision point):

bid: EV = -0.18
pass: EV = -0.29

The critical observation is that S4 was able to score a successful loner 5% of the time, and "only" got euchred 30% of the time. Without a very good defensive hand, S3 does better to nonetheless bid next with this hand, even knowing that partner in S1 will not be much help. More of a defensive move. Also note that those successful loners lose the game, there and then - so good to preempt!

Albeit, as this is based on 5% of hands, the usual caveat applies: are these REALLY the hands S3 will see, R2? Let me also interject that I have recently reviewed my program to re-assess S1,R2 calls, so feel confident that S1 is playing like a very good player, with a clear understanding of when to bid next, R2.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Mon May 23, 2022 3:16 pm

And, an analysis of Quiz question #17:

You are dealer, down 9-7, holding AS + AH + A-10D + 9C (10S turned). What do you bid?

Based on 100,000 hands, here's what I found:

pass: EV = -0.01
bid wp: EV = +0.05 (2,981 / 40,179 / 21,360) 64,520 hands made it to S4 for a decision
bid alone: EV = -0.41 (1,665 / 30,919 / 31,936)

On an EV bases, it's clearly better to bid with partner. But given the score, it's important to consider which option gives the best chance of winning the game. I plugged these results into a second program to determine what those odds are, given the distribution of outcomes from this hand and and the expected distribution of outcomes form subsequent random hands.

bid wp: 18.1% chance of winning the game
bid alone: 15.5% chance of winning the game

While bidding alone does give you an almost 3% chance of winning the game there in then, it also gives you an almost 50% chance of losing the game there and then. By bidding with partner, you can't win outright, but you have an almost 2/3 chance of making it to the next hand - and thus a better overall chance of winning the game.

This is not a situation to "swing for the fences". Make the better play, and hope the gods smile on you on the subsequent hands.
Last edited by raydog on Wed May 25, 2022 8:02 am, edited 1 time in total.

Tbolt65
Posts: 820
Joined: Sun Jun 16, 2019 9:14 pm
Location: Las Vegas

Unread post by Tbolt65 » Wed May 25, 2022 12:42 am

raydog wrote:
Mon May 23, 2022 3:16 pm
And, an analysis of Quiz question #17:

You are dealer, down 9-7, holding AS + AH + A-10D + 9C (10S turned). What do you bid?

Based on 100,000 hands, here's what I found:

pass: EV = -0.01
bid wp: EV = +0.03 (3,017 / 39,650 / 21,876) 64,548 hands made it to S4 for a decision
bid alone: EV = -0.41 (3,017 / 39,650 / 21,876)

On an EV bases, it's clearly better to bid with partner. But given the score, it's important to consider which option gives the best chance of winning the game. I plugged these results into a second program to determine what those odds are, given the distribution of outcomes from this hand and and the expected distribution of outcomes form subsequent random hands.

bid wp: 18.1% chance of winning the game
bid alone: 15.5% chance of winning the game

While bidding alone does give you an almost 3% chance of winning the game there in then, it also gives you an almost 50% chance of losing the game there and then. By bidding with partner, you can't win outright, but you have an almost 2/3 chance of making it to the next hand - and thus a better overall chance of winning the game.

This is not a situation to "swing for the fences". Make the better play, and hope the gods smile on you on the subsequent hands.
Not just this hand but there are other scenario's where swinging for the fences vs playing with your partner is not the right play. At these 9-7 scores or 8-7 or even 7-7. You want to take your partner a vast majority of the time and you definitely don't want to be swinging for the fences on skimpy loners. I know there are people who like to swing for the fence but in the long run I had and still now maintain playing for two is very crucial to winning period. Of course I have no math to back up what I say but everyone on this board already knows that.


Edit: Let's say you have Jd-Jh Ac-10c with the 9hearts up. This hand you want to go alone with. If you have Ad-Kd-Kc-Qc with 10 of diamonds. This hand you want to play for two with your partner.


Tbolt65
Edward

Tbolt65
Posts: 820
Joined: Sun Jun 16, 2019 9:14 pm
Location: Las Vegas

Unread post by Tbolt65 » Wed May 25, 2022 12:54 am

raydog wrote:
Mon May 23, 2022 12:54 pm
To keep the discussions flowing (and those euchre neurons firing), I've submitted Wes's euchre quiz question #16 to my simulator, to see what it spits out.

You're in 3rd seat, tied at 6-6, with K-QS + JH + AD + JC (AS turned). What do you bid?

I tested 100,000 hands, here's what I found:

bid: 8,100 / 50,725 / 27,904 [2 pts / 1 pt/ euchred] EV = +0.13
pass: EV = +0.91 (euchre S4 calls often; have excellent support for S1,R2 next calls [R in trump + top card in 2 off-suits]; have decent D call if it gets to you, R2)

No great surprise here, definitely better to pass.
_________________________

I also looked at the amended scenario where S3 doesn't have all the other suits blocked (I substituted the QH for the JH):

bid: EV = +0.15 (logically a shade higher, with the QH being a shade higher than the JH)
pass: EV = +0.69 (also logical, as S2 does better if they get to bid, R2; and S3 does worse if gets to bid, R2)

So, the same result.
_________________________

I noticed that in this 2nd scenario, my program had S3 bidding C, R2 (if it got that far), for an EV of -0.18. Is that the correct call, knowing S1 (partner) is an excellent player and understands the importance of a next call?

I compared this to passing, and found the following (100,000 tested, only 5,300 made it to this decision point):

bid: EV = -0.18
pass: EV = -0.29

The critical observation is that S4 was able to score a successful loner 5% of the time, and "only" got euchred 30% of the time. Without a very good defensive hand, S3 does better to nonetheless bid next with this hand, even knowing that partner in S1 will not be much help. More of a defensive move. Also note that those successful loners lose the game, there and then - so good to preempt!

Albeit, as this is based on 5% of hands, the usual caveat applies: are these REALLY the hands S3 will see, R2? Let me also interject that I have recently reviewed my program to re-assess S1,R2 calls, so feel confident that S1 is playing like a very good player, with a clear understanding of when to bid next, R2.

Don't forget that S1 passing doesn't automatically preclude that they have nothing. Depending on my partner. Especially if it's Wes. With a score of 6-6. I am probably passing and hoping to get a euchre on the dealer when they are forced to make something. With this 6-6 score we can gamble and go for a euchre. At other scores closer to end game, we can not. Other wise if it's a random. I'm probably calling club and praying for a club lead from my seat 1 partner.

Now if this was a different score say 8-8 or 9-8 my team. With Wes as my partner and knowing how he plays I can probably call hearts here. Knowing he is protected in both red and having a boss Ace dia off suit and boss King of spades it is likely we will get 1 point the majority of the time and some of the time get two points. Very rarely would we get euchred. Making this call beneficial.


Tbolt65
Edward

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Fri Jun 17, 2022 1:43 pm

raydog wrote:
Mon May 23, 2022 3:16 pm
And, an analysis of Quiz question #17:

You are dealer, down 9-7, holding AS + AH + A-10D + 9C (10S turned). What do you bid?

Based on 100,000 hands, here's what I found:

pass: EV = -0.01
bid wp: EV = +0.05 (2,981 / 40,179 / 21,360) 64,520 hands made it to S4 for a decision
bid alone: EV = -0.41 (1,665 / 30,919 / 31,936)

On an EV bases, it's clearly better to bid with partner. But given the score, it's important to consider which option gives the best chance of winning the game. I plugged these results into a second program to determine what those odds are, given the distribution of outcomes from this hand and and the expected distribution of outcomes form subsequent random hands.

bid wp: 18.1% chance of winning the game
bid alone: 15.5% chance of winning the game

While bidding alone does give you an almost 3% chance of winning the game there in then, it also gives you an almost 50% chance of losing the game there and then. By bidding with partner, you can't win outright, but you have an almost 2/3 chance of making it to the next hand - and thus a better overall chance of winning the game.

This is not a situation to "swing for the fences". Make the better play, and hope the gods smile on you on the subsequent hands.
I accept these results. I was hero-balling too much back in 2019 in 6-9/7-9 situations + other desperate spots. I'm actually surprised at how close it was (18.1 vs 15.5). I actually thought I was gonna be even more wrong than this.

Of course this begs the question: What is the worst hand in this spot where going alone beats out calling?

Prediction: This spade hand would meet the loner threshold:

(Card_9-S) (Card_J-C) (Card_A-H) (Card_K-H) (Card_A-D)

But what about:

(Card_9-S) (Card_J-C) (Card_A-H) (Card_9-H) (Card_A-C)

Or even this hand:

(Card_A-S) (Card_K-S) (Card_A-H) (Card_K-H) (Card_A-D)

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Fri Jun 17, 2022 2:04 pm

raydog wrote:
Mon May 23, 2022 12:54 pm
To keep the discussions flowing (and those euchre neurons firing), I've submitted Wes's euchre quiz question #16 to my simulator, to see what it spits out.

You're in 3rd seat, tied at 6-6, with K-QS + JH + AD + JC (AS turned). What do you bid?

I tested 100,000 hands, here's what I found:

bid: 8,100 / 50,725 / 27,904 [2 pts / 1 pt/ euchred] EV = +0.13
pass: EV = +0.91 (euchre S4 calls often; have excellent support for S1,R2 next calls [R in trump + top card in 2 off-suits]; have decent D call if it gets to you, R2)

No great surprise here, definitely better to pass.
_________________________

I also looked at the amended scenario where S3 doesn't have all the other suits blocked (I substituted the QH for the JH):

bid: EV = +0.15 (logically a shade higher, with the QH being a shade higher than the JH)
pass: EV = +0.69 (also logical, as S2 does better if they get to bid, R2; and S3 does worse if gets to bid, R2)

So, the same result.
Good stuff Ray. Makes sense to me. In your 2nd scenario I still think it's defensible to call with R+2+A if you can't trust your P in the 2nd rd. But even if one has a bad P they should never consider calling this hand in the first scenario when we have a euchre hand in the 2nd rd.

The only other interesting question is if the results change for a close-out scenario (9-9 or up 9-X where X = 7 or below). In a close-out scenario the math changes cuz there's no value in euchring the enemy.
raydog wrote:
Mon May 23, 2022 12:54 pm
I noticed that in this 2nd scenario, my program had S3 bidding C, R2 (if it got that far), for an EV of -0.18. Is that the correct call, knowing S1 (partner) is an excellent player and understands the importance of a next call?

I compared this to passing, and found the following (100,000 tested, only 5,300 made it to this decision point):

bid: EV = -0.18
pass: EV = -0.29

The critical observation is that S4 was able to score a successful loner 5% of the time, and "only" got euchred 30% of the time. Without a very good defensive hand, S3 does better to nonetheless bid next with this hand, even knowing that partner in S1 will not be much help. More of a defensive move. Also note that those successful loners lose the game, there and then - so good to preempt!

Albeit, as this is based on 5% of hands, the usual caveat applies: are these REALLY the hands S3 will see, R2? Let me also interject that I have recently reviewed my program to re-assess S1,R2 calls, so feel confident that S1 is playing like a very good player, with a clear understanding of when to bid next, R2.
S3 pretty much only gets a chance to call in the 2nd round in a weak game. In a tough game if the action gets that far someone has everything blocked.

With an amateur P I have been calling Next from S3-R2 with just R+0+2A for the last 5 years and my biased brain says that play is doing ok so I'm not surprised by your results that suggest calling with this hand is the best play.

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Sun Jun 19, 2022 12:04 am

Wes, I looked at the 3 hands you posted to consider bidding alone vs. with partner. [10S turn card in all instances]

1) 9S + A-KH + AD + JC:
My program has S4 bidding alone, and that has a better EV

2) 9S + A-9H + A-JC:
My program has S4 bidding alone, and that has the better EV (1.35 vs. 1.01)

3) A-KS + A-KH + AD:
My program has S4 bidding w/p, and that has the better EV (+0.89 vs. +0.67)
euchre rate is when bidding w/p is 40% of that when bidding alone, and that makes the difference (the extra 4-pt wins when bidding alone can't compensate).

I didn't explore the other scenario further, as it concerns only a small minority of cases, and is too dependent on how the other seats play prior.

BTW, you should be simulating these scenarios on your own, as you have the program!

Wes (aka the legend)
Posts: 1541
Joined: Wed Jun 13, 2018 3:03 pm

Unread post by Wes (aka the legend) » Sun Jun 19, 2022 6:11 am

raydog wrote:
Sun Jun 19, 2022 12:04 am
Wes, I looked at the 3 hands you posted to consider bidding alone vs. with partner. [10S turn card in all instances]
Wait, I'm confused. We're talking about marginal 2 trump spade loners here. The turn card can't be the 10S. Are these results still valid?
raydog wrote:
Sun Jun 19, 2022 12:04 am
BTW, you should be simulating these scenarios on your own, as you have the program!
You are correct. I suck. I will get back to it soon. Life has gotten in the way a bit.

PS: Congratulations to Don and Irishwolf who came to Vegas to play in our partnership tournament. They kicked ass and took first place!!!

raydog
Posts: 260
Joined: Thu Sep 16, 2021 6:56 pm

Unread post by raydog » Sun Jun 19, 2022 10:00 am

Sorry, Wes, I thought you were referring to the original hand.

I ran the scenarios again, had S4 pick up the 9S (or KS in the 3rd case), always discarding the 9D from their hand.

#1: EV pass = +0.29; EV w/p = +0.10; EV alone = +0.15
in this case, the stats were little changed when taking partner along (they didn't help much on balance), so those 4 pts for a sweep (rather than 2 pts) made going alone better than bidding with partner. BUT, I found it best of all to pass. S1 gets euchred a lot, R2, S2 bids some great hands (if he gets the chance), and S4 can bid H successfully if it gets that far.

#2: EV pass = +0.43; EV w/p = +0.13; EV alone = -0.09
in this case, still better to pass and euchre S1, but if you feel compelled to bid, do take partner along

#3: EV pass = +0.09; EV w/p = +0.04; EV alone = -0.14
I find it marginally better to pass here, but it's too close to be definitive. But definitely take partner along if you bid. Interestingly, I found slightly better success with sweeps when going alone, but the absolute success rate (~5%) was too low to make up for the greater chance of getting euchred (43% when going alone, 33% when taking partner along).

Well done, Irish and Don!

Tbolt65
Posts: 820
Joined: Sun Jun 16, 2019 9:14 pm
Location: Las Vegas

Unread post by Tbolt65 » Mon Jun 20, 2022 2:04 pm

Wes (aka the legend) wrote:
Sun Jun 19, 2022 6:11 am

PS: Congratulations to Don and Irishwolf who came to Vegas to play in our partnership tournament. They kicked ass and took first place!!!

Congratulations to Don and Irishwolf. They did kick ass. 8-2. 1st place. However they did not kick our ass, in the tournament 😎


Tbolt65
Edward

Post Reply