12-13-21 OE Monday - got euchred #1

[Dlan]

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

[raydog]

I don't understand the bidding in this hand. I simulated the hand, all cards fixed, and my program has S1 calling H, R1 (and earning a point, in this particular scenario). I also tested 10,000 hands with just S1's cards and the turn card fixed - so the as hand viewed by S1 - to compare calling H, R1, and calling S, R2 (by S1 both times). I found an EV of +0.20 when calling H, R1, and an even better EV of 0.27 when "bagging" and calling S, R2. [in the latter case, partner in S3 gets some very good results, R1, when they happen to call H, R1; S2 / S4 about break even when they happen to call H, R1; and S1 has an EV of 0.36 on the pure Spades call, R2 [so just looking at the hands when that happens]).

So I don't see why it ever came to S2 bidding C, R2 (or why they made that bid!)

Two takeaways for me:
1) My program assumes that every player at every point in time is playing optimally. But if players are passing biddable hands and reducing their chances of winning pts., I can't possibly simulate that. Finding the proper subset of possible hands remaining when R2 rolls around, for example, is crucial, and will be skewed if players are erroneously passing hands. But that's real life, and I don't see any way around it. In the long run, if opponents are playing poorly, one's odds of winning the game improve. [I don't mean to cast aspersion on whoever played this particular hand; there may have been a good reason, and at any rate, I make plenty of mistakes when I play. This is just a general observation vis-a-vis the reliability of using statistics and probability to determine best practice.]

2) My program missed the more favorable "bag H, R1, and bid S, R2" option because I currently only check out the "bag R1 to call next R2" option, and Spades here is not next. Another update and improvement I will make to my program!

[Wes (aka the legend)]

I don't understand the bidding in this hand. I simulated the hand, all cards fixed, and my program has S1 calling H, R1 (and earning a point, in this particular scenario). I also tested 10,000 hands with just S1's cards and the turn card fixed - so the as hand viewed by S1 - to compare calling H, R1, and calling S, R2 (by S1 both times). I found an EV of +0.20 when calling H, R1, and an even better EV of 0.27 when "bagging" and calling S, R2. [in the latter case, partner in S3 gets some very good results, R1, when they happen to call H, R1; S2 / S4 about break even when they happen to call H, R1; and S1 has an EV of 0.36 on the pure Spades call, R2 [so just looking at the hands when that happens]).

So I don't see why it ever came to S2 bidding C, R2 (or why they made that bid!)

Two takeaways for me:
1) My program assumes that every player at every point in time is playing optimally. But if players are passing biddable hands and reducing their chances of winning pts., I can't possibly simulate that. Finding the proper subset of possible hands remaining when R2 rolls around, for example, is crucial, and will be skewed if players are erroneously passing hands. But that's real life, and I don't see any way around it. In the long run, if opponents are playing poorly, one's odds of winning the game improve. [I don't mean to cast aspersion on whoever played this particular hand; there may have been a good reason, and at any rate, I make plenty of mistakes when I play. This is just a general observation vis-a-vis the reliability of using statistics and probability to determine best practice.]

2) My program missed the more favorable "bag H, R1, and bid S, R2" option because I currently only check out the "bag R1 to call next R2" option, and Spades here is not next. Another update and improvement I will make to my program!

I dont get it either raydog. Passing a S1-R2 3 trump call when one only blocks 1 out of 3 suits is just flat out poor euchre play. This is a classic example of a Hoyle loyal pass. S1 is afraid to jump the fence, so he passes a clear biddable hand instead. What S1 needs to understand is his #1 job is to play to win the game not go with Hoyle. What suits he has blocked or not blocked should be the primary driver of his action in these marginal spots (unless we havehard data that suggests otherwise), not Hoyle. Hoyle should certainly be a factor in S1's decision making but not the primary one.
Hoyle concepts can make a beginning player good, but more often than not Hoyle prevents a good player from becoming great. The best players in the world don't play by Hoyle, they play to win. Always.

Moving on, S2 makes what I consider another Hoyle Loyal blunder. Why is he calling clubs there with just 10-9 when he has R+1 in diamonds? I would be shocked if a good simulation showed that club call beating out a that diamond call.

Hoyle concepts were never meant to be abused like this. The idea is if all your decision points are close than go with Hoyle, but if they are not close then you got to play your hand. This game is not that complicated. People are overthinking things man.

[irishwolf]

It's a EUCHRE PROVERB - YOU MUST CALL ON BIDDABLE HANDS TO WIN IN THIS GAME!

Some good comments Ray, but one thing about your program, using this hand as an example, to me hands down a better decision for S1 to Pass R1 and call Spades in round 2, leading trump to the first trick.

So does your program not weigh the odds of calling one suit against another? And take the best route?

And to S2/S4 getting euchred? It was a 'blind' lead of trump to that 2nd trick when a Diamond should have been led for a point. But the issue is one vs the other?
Learn when to pull the trigger!

IRISHWOLF

[Wes (aka the legend)]

It's a EUCHRE PROVERB - YOU MUST CALL ON BIDDABLE HANDS TO WIN IN THIS GAME!

Some good comments Ray, but one thing about your program, using this hand as an example, to me hands down a better decision for S1 to Pass R1 and call Spades in round 2, leading trump to the first trick.

Yep. Gotta lead trump on first trick from this configuration. Ray's numbers would seem to suggest that a two trump hand like this would be a 1st rd call (vs a 9H upcard):



Cool stuff.

And to S2/S4 getting euchred? It was a 'blind' lead of trump to that 2nd trick when a Diamond should have been led for a point. But the issue is one vs the other?
Learn when to pull the trigger!

IRISHWOLF

As you noted, the crux of the hand comes down to that "blind lead" on 2nd street. Should S2 lead trump in that spot or not? That is the question and really the most interesting aspect of this hand (obvious mistakes aren't really that interesting to me). If S2 had an off ace to promote I would lead trump in that spot. Or if S2 simply had no fresh voids I would also lead trump in that spot. But in this hand with a fresh void to work with but no off ace I think leading the Diamond is best but I'm not like positive on that especially with the two lowest trump and thus no possibility of overtrumping the enemy later in the hand. If S2 had say Ac9c instead of Tc9c, then I would feel strongly that leading trump is bad when we have a fresh void to work with, with no off ace. So yeah I'm in the camp that lead trump in this spot is wrong. If a simulation proved me wrong on Tc9c I wouldn't be surprised. If a simulation proved me wrong on Ac9c I would be shocked.

[Wes (aka the legend)]

Also, it's worth noting that not only did S1 make the mistake of passing a biddable hand, he also led incorrectly after the 2 seat called. S1 should not be leading the turned down suit. Lead your off ace. Either it forces out enemy trump or it walks. Both are good results for your team. And sometimes S2 will get overtrumped on this first lead which is a coup for your team. So many euchres happen becuz the maker got overtrumped somewhere.

[raydog]

Irish, to answer your question, when considering S1, R1 bids, my program currently only looks at potential next suit bids, R2, not green suit bids. An easy thing to change, and I will do so.

For other seats, potential R2 bids are not considered, as there is a high probability that opportunity will never come (as S1 will very often bid, R2). Though I realize that, depending on the score and the hand, having a potential suit to bid in the 2nd round could be a factor in the decision to bid or not, R1. A degree of complexity I'm not ready to tackle just yet.

[irishwolf]

My experience on R2, S1, about 70% S1 will call about 70%, either Next or Cross suit. Pass 30% has nothing or a stopper in green/euchre hand. That's SWAG and highly variable due to many variables, TNTC.

For what it's worth.

But I think your program for R1 is realistic and on track to evaluate hands.

Cool!