Euchre, Bagging with a Next Order

[Richardb02]

The “Bag” works perfectly, but if Dealer passes, you have a Next order in reserve! This is an excellent example!

Comments?





https://worldofcardgames.com/#!replayer ... %3A1%7D%5D

[Wes (aka the legend)]

I don't like it. In theory, one's Next call does not have to be as strong as one's 1st round call given the fact that waiting for Next gives S1 the added benefit of sometimes euchring his opponents. But the disparity of the two hands is just too large here. R+2+A is just too much better than calling Next with L+1. Of course this is just my opinion. One doesn't really KNOW until they put it to the test.

[raydog]

I tried putting this to the test (10,000 hands) and found it was much better for S1 to call up the KS R1.

Get euchred 13% of the time, make 2pts 13% of the time, make 1pt. the rest. Very good +EV.

If pass R1 and bid C R2, do a nice job euchring opponents when they do call R1 (they call about 25% of the time, and you euchre then 40-45% of the time), but when you call next R2 (75% of the time), you get euchred almost half the time, make very few 2pts., and so barely eke out a +EV.

No contest.

[jspectre]

It's tempting to try and bag this hand at that score, as getting a euchre with your upcoming deal gives you great winning chances, but it doesn't seem worth it. There's only 3 trump left unaccounted for, and the dealer isn't likely to call without having 2 out of those remaining 3. I would rather call here than hope for the dealer to pick up, and then bet it all on next if they were to pass, especially because next loses value the more I have in the original suit. I can still get a march on my deal or there's potential to still win on the opponents deal at 9-9, make the statistically stronger play here, and keep yourself in the running.

[irishwolf]

Raydog, you said, ..."when you call next R2 (75% of the time), you get euchred almost half the time, make very few 2pts., and so barely eke out a +EV."

How was the hand set up? What did S1 lead to the first trick? I don't believe S1 should be euchred 50% of the time.

He has three options on what to lead and they are not all equal.

IRISH

[raydog]

Apologies, Wolf - 50% is indeed way too high. I mistakenly compared "points lost by being euchred" with "points won by winning the hand (1 pt)", and since the latter was only slightly lower than the former I said almost 50%. Clearly, the point values of the two cases are different.

So let me give the correct stats here:

S1 calls next (clubs) R2 and leads :
2pts: 464 hands (6%)
1 pt: 4466 hands (57%)
-2pts: 2904 hands (37%)
EV = 0.046 [includes the hands where another player bid R1, not shown here]

same as above, S1 leads :
2 pts: 516 (7%)
1 pt: 4233 (54%)
-2 pts: 3085 (39%)
EV = -0.001 [includes the hands where another player bid R1, not shown here]

same as above, S1 leads the :
2 pts: 449 (6%)
1 pt: 3816 (49%)
-2 pts: 3569 (45%)
EV = -0.19

Note that I also ran the scenario for 10,000 hands here - I had only run it for 1,000 hands before (another error) and it turns out bidding R2 is even worse than I suspected. This run of 10,000 hands shows everyone passing R1 78% of the time.

You are always very skeptical of my data, as am I, because my program may not be playing correctly. For example, if expert players would bid or pass a different subset of hands R1, that would leave a different subset to be played by S1 in R2, altering the results. And, obviously, the play of the hand when S1 does bid R2 (including the first lead) also matters [my program did by default lead the 9H, which seems to be the best choice].

I am only reporting my findings here because they overwhelmingly point to one play (i.e., calling the KS into the dealer's hand, R1) being better to the other (i.e., bidding next, R2). I don't pretend to be able to tease out subtle differences in play and say which is better.

[irishwolf]

Raydog,

We are in agreement that Eldest (S1) should order up the KS with the hand he has, and not knowing that the Dealer in this particular situation is better than bagging the Dealer or Calling next. The EV will be higher than bagging.

It's simply easily calculable, no simulator needed as probability of combinations of the dealer having (3 unknown) no trump ~35%, one trump is ~47%, 2 trumps ~16% and so on. That compared to calling next with only Left/Ace. Yes, the dealer might also an ace as well of the two unknown to S1.

Assuming S1 ordered the KS, I would lead the 9S to the first trick. I would anticipate my euchre rate to be around 14% (+/-3%). The EV on ordering should be about .65. My partner will have the AS or JC a good percentage of the time. I can't get that high of a EV calling next or bagging.

My projection is similar to you simulator results:
"I tried putting this to the test (10,000 hands) and found it was much better for S1 to call up the KS R1. Get euchred 13% of the time, make 2pts 13% of the time, make 1pt. the rest. Very good +EV." I agree.

If I keep my euchres below 36%, I will always have a + EV.

IRISH

[Tbolt65]

I tried putting this to the test (10,000 hands) and found it was much better for S1 to call up the KS R1.

Get euchred 13% of the time, make 2pts 13% of the time, make 1pt. the rest. Very good +EV.

If pass R1 and bid C R2, do a nice job euchring opponents when they do call R1 (they call about 25% of the time, and you euchre then 40-45% of the time)


Apologies, Wolf - 50% is indeed way too high. I mistakenly compared "points lost by being euchred" with "points won by winning the hand (1 pt)", and since the latter was only slightly lower than the former I said almost 50%. Clearly, the point values of the two cases are different.

So let me give the correct stats here:

S1 calls next (clubs) R2 and leads :
2pts: 464 hands (6%)
1 pt: 4466 hands (57%)
-2pts: 2904 hands (37%)
EV = 0.046 [includes the hands where another player bid R1, not shown here]

same as above, S1 leads :
2 pts: 516 (7%)
1 pt: 4233 (54%)
-2 pts: 3085 (39%)
EV = -0.001 [includes the hands where another player bid R1, not shown here]

same as above, S1 leads the :
2 pts: 449 (6%)
1 pt: 3816 (49%)
-2 pts: 3569 (45%)
EV = -0.19

Note that I also ran the scenario for 10,000 hands here - I had only run it for 1,000 hands before (another error) and it turns out bidding R2 is even worse than I suspected. This run of 10,000 hands shows everyone passing R1 78% of the time.

You are always very skeptical of my data, as am I, because my program may not be playing correctly. For example, if expert players would bid or pass a different subset of hands R1, that would leave a different subset to be played by S1 in R2, altering the results. And, obviously, the play of the hand when S1 does bid R2 (including the first lead) also matters [my program did by default lead the 9H, which seems to be the best choice].

I am only reporting my findings here because they overwhelmingly point to one play (i.e., calling the KS into the dealer's hand, R1) being better to the other (i.e., bidding next, R2). I don't pretend to be able to tease out subtle differences in play and say which is better.

Hi Raydog,

With everything you said and shown. Your numbers alone speak to playing for 2 points and take that gamble that dealer could pick up.

Your team is down 7-9. Where is it most advantageous to get two points here? By ordering or by euchre and hoping they pick up? Of course its going to be by a euchre and hoping they pick it up.

Passing here and calling next show's 63%, 61% and 55% chance of either Making point or getting two points. Which make bagging even more profitable when you are favored to make points on an ensuing pass/turn down.

Also another factor to consider is the type of players your playing against. Aggressive players? Passive players? Expert, Advanced, Good, Intermediate, average or bad? These are also considerations that factor in, when deciding to make a pass or a call. Math alone is good to have guide and a sense of things but there are other factors like score, opponents, partners , upcard, various leads on various calls or orders ect.... ect... That go into when deciding what to do at any given decision making point. Plus regardless of if you make 1pt or 2pts on either the bag or your second round call. You still have the deal. So you get two chances at getting it 9-9 with the pass or second round call and then getting your deal at that score which makes your chances of winning the game that much more. It is important and well worth the risk to get your deal and 9-9. It is less favorable to be at 8 to 9 your deal but still is advantageous. But now you need two points to go out instead of just 1point.


Tbolt65
Edward

[raydog]

Edward, you make an excellent point. I have a habit of not considering the score when I make my bidding decisions (pure mathematics, and optimizing EV), which is not optimal. Something I need to work on.

Looking again at this particular hand, with the score in mind, and considering my two options (#1: bid R1; #2: bid C R2), I calculate the following "raw" odds:

#1) get 2 pts.: 13%
get euchred (game over): 13%

#2) get 2 pts.: (22% X 45%) + (78% X 7%) = 15%
opponent scores a pt.: (22% X 55%) + (78% X 37%) = 41%

The way I read, even allowing for aggressive (or inept) opponents pumping up that "15% euchres for 2pts" number (when I bag), there is no escaping that fact that my opponents score at least a point and win the game there and then 3 X as often if I bag. And I certainly want to avoid that at all costs!

So ultimately, I stick by my decision to call S R1.

[Tbolt65]

I think you under value the importance of receiving 2pts to get it to your deal at 9 to 9. The risk is well worth it. You don't give up too much by taking it. You still have an good chance of making point in round 2. By passing you make it very hard for dealer to make a point. They will not be making point 3x more if they call up when you pass. That's counter intuitive to what you suggest that by ordering up round one you make point 74% of the time.

It is much easier to finesse 1pt at a 9 to 9 score than to do the same finesse at 8 to 9 your deal. You want to make it as easy as possible and make it the highest percentage possible to win on your deal. The highest percentage to do that is to bag it. You are likely to get 1 point on a pass or if you had ordered in either case but you miss out on the high certainty of a euchre if you do order when the dealer is likely to pick up. Sure you have a higher percentage to make point when ordering 1st round. But you still have a strong percentage to make point even if the dealer passes as well. So why not make a smart gamble and allow your opponents to make a mistake?

Tbolt65
Edward

[raydog]

I understand what you are saying, Edward, but when I crunch the numbers I find that S1 bidding R1 gives them a better chance of winning the game than if they bag R1 and bid R2.

Given the current score (and current hand), no one is going to going bidding alone anymore in this game. So the possible point values for this and and subsequent hands (should they be necessary) are 1 or 2 pts for the dealing team or 1 or 2 pts for the non-dealing team.

There are 5 paths to victory for the team sitting N/S, as summarized in this table:
(they obviously can't lose any pts)

current hand next hand 3rd hand
1) +1 pt +1 pt +1 pt
2) +1 pt +1 pt +2 pt
3) +1 pt +2 pt
4) +2 pt +1 pt
5) +2 pt +2 pt

It's simply a matter of calculating the odds of each of these scenarios and summing them. Here are the odds I found of each team scoring 1 or 2 points by simulating 100,000 random hands:

current hand current hand subsequent
S1 bids R1 S1 bids R2 hands
dealing team scores 2 pts: 12.5% 28.9% 22.9%
dealing team scores 1 pt: 0% 12.4% 43.7%
other team scores 1 pt: 75.6% 44.7% 19.1%
other team scores 2 pts: 12.0% 14.0% 14.3%

Plugging the appropriate percentages into the victory paths shown previously, I find that N/S have a 36.3% chance of winning the game if S1 bids R1 in the current hand, but that drops to a 26.1% chance of winning if S1 passes R1 and bids next R2. Despite the fact that they score 2pts a few more times (through euchres) if they pass R1.

But maybe you don't trust the stats of who scores how many points if S1 bids R2. I tried a fake scenario where I boosted the % of 2 pt scores to 30%, and lowered the % of 1 pt scores to 40% (so a hugely better outcome for N/S in the current hand than expected), and using these numbers found that N/S still only has a 35.0% chance of winning the game. Even with all those 2pt scores.

What appears to be MOST important here is not letting the other team score a single point, which would end the game there and then.

P.S.: just to walk you through a sample calculation, the odds of team N/S winning the game by winning 1 pt in each of 3 games [scenario 1) above], if S1 bids R1, is given by (.756)(.437)(.191) = .063. Do this calculation for all 5 scenarios and add together to get 36.3%.

[Tbolt65]

I understand what you are saying, Edward, but when I crunch the numbers I find that S1 bidding R1 gives them a better chance of winning the game than if they bag R1 and bid R2.

Given the current score (and current hand), no one is going to going bidding alone anymore in this game. So the possible point values for this and and subsequent hands (should they be necessary) are 1 or 2 pts for the dealing team or 1 or 2 pts for the non-dealing team.

There are 5 paths to victory for the team sitting N/S, as summarized in this table:
(they obviously can't lose any pts)

current hand next hand 3rd hand
1) +1 pt +1 pt +1 pt
2) +1 pt +1 pt +2 pt
3) +1 pt +2 pt
4) +2 pt +1 pt
5) +2 pt +2 pt

It's simply a matter of calculating the odds of each of these scenarios and summing them. Here are the odds I found of each team scoring 1 or 2 points by simulating 100,000 random hands:

current hand current hand subsequent
S1 bids R1 S1 bids R2 hands
dealing team scores 2 pts: 12.5% 28.9% 22.9%
dealing team scores 1 pt: 0% 12.4% 43.7%
other team scores 1 pt: 75.6% 44.7% 19.1%
other team scores 2 pts: 12.0% 14.0% 14.3%

Plugging the appropriate percentages into the victory paths shown previously, I find that N/S have a 36.3% chance of winning the game if S1 bids R1 in the current hand, but that drops to a 26.1% chance of winning if S1 passes R1 and bids next R2. Despite the fact that they score 2pts a few more times (through euchres) if they pass R1.

But maybe you don't trust the stats of who scores how many points if S1 bids R2. I tried a fake scenario where I boosted the % of 2 pt scores to 30%, and lowered the % of 1 pt scores to 40% (so a hugely better outcome for N/S in the current hand than expected), and using these numbers found that N/S still only has a 35.0% chance of winning the game. Even with all those 2pt scores.

What appears to be MOST important here is not letting the other team score a single point, which would end the game there and then.

You Imply here that S1 has a greater chance of making a mistake when passing a Round 1 a hand that could make a point and gives up points when they bag the same hand, thus ending the game. If the hand is strong enough to make a point conversely, it's also strong enough to get two points if the dealer picks up. Correct? So, if it does go pass-pass, and seat 1 now has the call in the second round and only just gets the 1 point they would have had if they had ordered Round 1. Nothing is lost. Correct? Hence the numbers in the long run should point to more euchres on a pass. Which you do say will get more euchre's but also say the chance of winning goes down significantly. How is that? When you have the deal, you'll get 9-9 more often when passing as you stated but how are you losing more often in this spot with your deal? This is what you want. This is the whole premise of bagging and getting it your deal and the score of 9-9. I believe Wes says it's like a 65% or 67% favor for the dealer at this point and I'll take that all day and every day. This is the best scenario you want with the hand we currently have being at a score of 7-9 on the previous hand. The only thing better is to go alone. This is not a hand you can go alone with, in any round. I think you may confuse the numbers because that 35% would be about right for E/W winning the game on N/S 9-9 deal.

I believe the hand generator is not playing optimally or near optimal in the Seat 1 and dealer scenarios or on the ensuing hand when S1 becomes the dealer on the next deal.

Tbolt65
Edward

P.S.: just to walk you through a sample calculation, the odds of team N/S winning the game by winning 1 pt in each of 3 games [scenario 1) above], if S1 bids R1, is given by (.756)(.437)(.191) = .063. Do this calculation for all 5 scenarios and add together to get 36.3%.

[irishwolf]

Down 7 to 9, you have to pass and strongly suspect the Dealer will pick it up to protect his score! Even though if S1 orders he will make just one point about 82% chance (sweeps are hard to come by, statistically speaking).

You can't bank on a sweep with with what S1 has - dealer will have at least one other trump and now has two to catch a trick and block a March. And the AH is 56.7% with opponents.

If he S1 orders and only makes 1 point and now 8 to 9. 65% chance to score 1 point (as sweeps are about 16%). Now the score is 9 to 9 and their deal. If all that happens I have about a 35% chance of winning at S1.

I (R1) am passing at S1 even though I have a better chance of scoring a point - I need a euchre. But if I am at Dealer seat, I am making trump unless I have everything blocked. (Sounds like Eric is present, lol!)

Just how I see this!

IRISH

[Wes (aka the legend)]

Down 7 to 9, you have to pass and strongly suspect the Dealer will pick it up to protect his score!

This reasoning doesn't work becuz virtually nobody does this. I mean you could be right that bagging is better but it wouldn't be becuz there's a great chance the "dealer will pick it up to protect the score". I've played over 30K games in the last 4 years. I know like 5 players including myself and you that will ever call light from the dealer spot to protect a score. But even with all that said......

But if I am at Dealer seat, I am making trump unless I have everything blocked.

This is the wrong mindset when up 9-7. You've talked about this before and I nodded when I read your post. You said something like, Up 9-6 you can protect, up 9-7 you really can't becuz the cost of getting euchred is too high. So in this spot even expert players' hands are tied here.

[irishwolf]

WES, OOPS! actually I mid-spoke due to the Dealer seat has 9 (9 - 7). So let me clarify. Can't protect the score in this situation as you cannot afford a euchre because Dealer's side does not have the next deal. Yes, i agree with your statement. So scratch my comment ..."I am making trump." Glad you are paying attention!

However, it's true with a hand having everything blocked, - I am Passing as the score worst case is 9 to 8 and their deal. If we hold them to 1 point 9 to 9 and now our deal!

And with the hand as posted, (AS KS QS KH 9H), I would call trump because I will score a point about 75% of the time (better than the alternatives of Passing).

I hope that clarifies this.

WES said, "This is the wrong mindset when up 9-7. You've talked about this before and I nodded when I read your post. You said something like, Up 9-6 you can protect, up 9-7 you really can't becuz the cost of getting euchred is too high. So in this spot even expert players' hands are tied here." YES - AGREED!

IRISH

[raydog]

To Edward: Calling Spades R1 and Clubs R2 are not identical for S1. Much better chances of making a pt. R1 (76%, by my simulation) than R2 (45%, by my simulation). It's those additional times getting euchred on a R2 call that scupper S1's chances of winning the overall game. The stats on R1 calls by S1 are purely how well their hand holds up. The stats on R1 passes / R2 calls by S1 are an amalgam of results from bids by other players, R1, PLUS how well S1's hand holds up on a Clubs bid (OK, but not great), so it gets a bit messy. But I still find it much worse than bidding R1.

Irish: if S4 picks up 100% of the time that S1 passes, R1, then S1 will indeed get a ton more 2pt. scores (from euchres). I just don't think that is a realistic assumption. But if you know your opponent in S4 is going to be hyper-aggressive or -protective and call trump even on extremely thin hands with a very negative EV, than by all means it may well pay off to bag R1. That's nothing I can simulate - in my mind it's poor play which will reduce S4's chance of winning the game (e.g, calling with a 60% chance of being euchred).

[Tbolt65]

To Edward: Calling Spades R1 and Clubs R2 are not identical for S1. Much better chances of making a pt. R1 (76%, by my simulation) than R2 (45%, by my simulation). It's those additional times getting euchred on a R2 call that scupper S1's chances of winning the overall game. The stats on R1 calls by S1 are purely how well their hand holds up. The stats on R1 passes / R2 calls by S1 are an amalgam of results from bids by other players, R1, PLUS how well S1's hand holds up on a Clubs bid (OK, but not great), so it gets a bit messy. But I still find it much worse than bidding R1.

S1 R1 vs S1 R2. At face value. S1 R1 is a stronger hand by itself. That is not debatable. I think if you add in and account for the times you get the euchre. This added percentage is going to make it very close. Much closer than what we have currently of 76% vs 45% in a R1 vs R2 s1 call respectively. With that added %. Whatever that is going to be. I contest will make it a necessary gamble and well worth it. To indeed pass. Especially at this score.

With that said if all we were looking at was the s1 r1 order vs a s1 r2 order. It is clear that a s1 R1 would be the best way to go.

Tbolt65
Edward

[raydog]

While the added euchres achieved when S4 (or S2) pick up do skew things to the advantage of S1, my simulation finds that S2 + S4 only declare trump 20% of the time if S1 passes, R1 [of course, the 20% of times when they have the strongest hands, and get euchred "only" 45% of the time]. Which means S1 ends up playing 80% of the hands with Clubs as trump. And so the two scenarios are not close.

You may argue that S4 will pick up the KS far more often - perhaps 50, 60, even 70% of the time, but I would deem that a mistake, as it is ultimately inviting far too many euchres and hurting their chances of winning the game. If their euchre rate is 45% on the BEST 20% of hands, it will be 60%+ on the next 20% of hands, and only getting worse from there. [Granted, S4 doesn't KNOW that their euchre rate will be that high, as they don't know S1 has such a strong hand. But even against a more random sample of cards in the other players' hands, their euchre rate would be unacceptably high]. But as I said in my reply to Irish, I can't properly simulate that kind of play. So I think we simply won't agree on how best to play this hand.

Ray

[irishwolf]

Raydog, I would almost NEVER make trump up 9 to 7 with only ...(e.g, calling with a 60% chance of being euchred). But I would at 65%, and that reason should be obvious because if the score gets to 9 - 9, S2 dealer. You have no better odds of winning. And you still have life if Dealer gets euchred 9 to 7 (now 9 to 9).

I think I clarified what I would do at the Dealer seat - please re-read.


IRISH

[Tbolt65]

While the added euchres achieved when S4 (or S2) pick up do skew things to the advantage of S1, my simulation finds that S2 + S4 only declare trump 20% of the time if S1 passes, R1 [of course, the 20% of times when they have the strongest hands, and get euchred "only" 45% of the time]. Which means S1 ends up playing 80% of the hands with Clubs as trump. And so the two scenarios are not close.

You may argue that S4 will pick up the KS far more often - perhaps 50, 60, even 70% of the time, but I would deem that a mistake, as it is ultimately inviting far too many euchres and hurting their chances of winning the game. If their euchre rate is 45% on the BEST 20% of hands, it will be 60%+ on the next 20% of hands, and only getting worse from there. [Granted, S4 doesn't KNOW that their euchre rate will be that high, as they don't know S1 has such a strong hand. But even against a more random sample of cards in the other players' hands, their euchre rate would be unacceptably high]. But as I said in my reply to Irish, I can't properly simulate that kind of play. So I think we simply won't agree on how best to play this hand.

Ray

I'm talking about this specific hand only. Not any other possibly holdings or hand make up. Because any other holdings, score etc... will possibly have dramatic changes in the way the hand is approached. For the most part you can't put it into one bag. You have to deal with it on the fly and be able to scrutinize and use situational awareness that helps guide one in either making the the proper decision or a properly estimated gamble.


My thinking is not how often dealer will pick up here. However, that is something that does come up but for this particular hand and the score there are other things that take precedent. Like how am I going to best achieve a victory. Can I set the dealer if they pick up? Can I make point if it is passed? Am I more likely to earn two points ordering or hoping dealer picks up? What is the best spot I can put myself in to get it to my deal? All this goes on in a fraction of a second because I have put the time and hands in to know when to gamble in certain spots. This is one of them in my opinion.

I have made calls where I could have easily bagged from first seat but my hand wouldn't been as strong if I had ordered First round. There are times to do it. Most importantly on loner type hands because of the point swing by not ordering and calling when you had a much strong hand to go alone with. Each situation is so very important. When it comes to making decisions.


Tbolt65
Edward

[Tbolt65]

This might help give further perspective to the discussion.

Below is taken from a post from ohioeuchre post : viewtopic.php?f=8&t=745


Tbolt65
Edward


=============



Probably of winning/Losing based on various scores.

(each cell based on 10,000 simulated games and assuming scores of individual hands equivalent to a 100 hand sample size)

Developed by Fred Benjamin, posted on Euchre Science @ Yahoo.com 2007 or 2008.

==========================================================================

• Having the deal in a 0-0 game has a slight 51% win advantage
• Any tie score except 9-9 offers the current dealer only a slight advantage
• Having the deal in a 9-9 game has a huge 2:1 win advantage
• As a rule of thumb, each point you lead by will give you a 14% advantage

==========================================================================

Click the link below to download info in a word format

https://ohioeuchre.com/pdf/FRED_BENJAMI ... Euchre.doc