WHAT TO DO ON THIS HAND

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

WHAT TO DO ON THIS HAND

Unread post by irishwolf » Fri Dec 31, 2021 10:25 am

It is always good to have a hand to discuss. This one can be quite controversial. This hand we a some Definitive answers, once and for all.

Thus this is really for Raydog's simulator.

You sit in 3rd seat, (S3) and the up card is (Card_A-H) (AH), your hand is (Card_J-D) (Card_K-H) (Card_Q-H) (Card_A-C) (Card_J-S) (JD KH QH AC JS).

With the following questions:

1) Do you order (S3)? If so what is the Euchre rate, Sweeps & 1 pointers? The EV?

2) What is rate of the Dealer (S4) ordering the AS? If the dealer picks it up the Euchre rate, 1 pointers, and EV.

3) What is MOST likely to occur after the Dealer passes, R1 (S1, S2, S3,S4 on R2?

While (that's if Ray decides to bite) what are the OE readers answers to these questions? And of course, why? Don't be shy you won't learn anything!

IRISHWOLF



XaviRonaldo
Posts: 84
Joined: Thu Aug 16, 2018 11:41 am

Unread post by XaviRonaldo » Sun Jan 02, 2022 12:15 am

Depending on the score I'm ordering that up.

If the opponent is 2 points from winning I'm probably passing. If my partner has nothing then we are probably in trouble. If my partner has another suit to call I have a definite trick in any suit. But a euchre is too much of a risk. Best to play safe knowing there is no possible way the opponent can get 2 points if they call. If the spade wasn't the Jack though I would probably still order up. I am vulnerable to a black call and with the opponent 2 points away my partner isn't going to call thin. So the opponent will probably call black and quite possibly get the 2 points anyway.

On any other score it's worth calling up IMO.

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

Unread post by raydog » Mon Jan 03, 2022 8:01 pm

I simulated this hand over 1,000,000 random iterations (S3 and turn card fixed).

Since I developed my program without STD (since that's the way I learned to play), I first checked what S3 should do, R2, if they pass R1.

(100,000 hands)
S3 bids Clubs, R2: 1,011 / 4,110 / 1,357 [2 pts. / 1 pt. / euchred] EV = +0.55
S3 passes, R3 (STD): S4's results
0 / 138 / 17 || 0 / 2,674 / 3,649 [4 pts. / 1 pt. / euchred || 2 pts. / 1 pt. / euchred]
(1st 3 numbers are for bidding alone; last 3 numbers are for bidding w/ partner)
EV = -0.70, which means an EV of +0.70 for S3.

So better for S3 to pass, R2, if STD is in effect. UNLESS S2/S4 are at 9 pts, in which case it is better to bid Clubs (S2/S4 only earn points 1,357 games instead of 2,812, or about 1/2 as often)

Here's the data from the main simulation:
S3 bids, R1: 80,367 / 531,321 / 254,085 EV = +0.21
865,773 games make it to S3 for a bid
S2/S4 score at least one point 254,085 games
S1 bids about 1.6% of games for an EV of +1.57; S2 bids about 11.9% of games for an EV of -0.91.
I didn't include those games in my calculations because they don't vary and so are simply dilutive.

S3 passes, R1:
S4, R1: 0 / 25,331 / 71,140 || 0 / 115,366 / 120,131 EV = -0.73; 33.2% of games
S1, R2: 19,362 / 14,352 / 208 || 110,151 / 197,754 / 24,227 EV = +1.26; 36.6% of games
S2, R2: 0 / 10,240 / 736 || 0 / 47,826 / 43,485 EV = =0.30; 10.2% of games
S3, R2: (passes)
S4, R2: 0 / 1,344 / 197 || 0 / 26,680 / 37,244 EV = -0.72; 6.5% of games
overall EV for S1/S3 = +0.90
S2/S4 score at least one point 251,222 games

From this data it seems clear that S3 should pass, R1, with this particular hand as their EV is much higher. But we do need to think about score. As XaviRonaldo pointed out, S2/S4 can't possible score 2 points, so if the opponents have 8 points or less then passing is a good strategy: better EV AND opponents can't win the game on this hand. But what if they have 9 points? When S3 passes, R1, S2/S4 STILL earn at least a point less often than if S3 bids. And even if S4 decides to bid more often, R1, to try and win that last point, they will only be marginally successful, and give up chances at euchring S1 in R2. On the whole, S4's best strategy would be to NOT bid more aggressively (but since they can't see S3's hand they of course WILL bid more aggressively, and this particular case it will hurt them).

Of course, 254,085 vs. 251,222 is perhaps too close to be significant. I concede that. But I did do a simulation over 100,000 hands, comparing if S4 bid normally, R1, vs. calling EVERY hand. In the latter case they were euchred mercilessly BUT were able to eke out an extra 6,200 hands where they scored the much needed point. Unfortunately, this meant that no games went to R2, so they missed out on 9,400 potential games where they would have scored a point. So, indeed, better to not be aggressive.

Is this what you were looking for, Irish?

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

Unread post by irishwolf » Mon Jan 03, 2022 8:55 pm

EXCELLENT RAY!

Reviewing now and have only got as far are S3R1 which is what I wanted to see how it compared with my test of 300 hands. That euchre rate you got of 29%, I got 31% euchre rate, but also got 12 - 14% sweeps. Concluding (me that is) that the EV will always be positive ordering up the AH. I gave every opportunity for S2/S4 to euchre S3 as I wanted a conservative estimate, i.e., like leading Spades even though diamonds could have been led to S3 void. That makes a difference with euchre rate. But I found, concluded in my test that you can't go wrong ordering, S3R1. I found that S2/S4, their euchre rate was not that high because it was almost the same hands that when S3 ordered, he got euchred. However, R2 I do agree with you overall but need more time to review the details of your 100,000 hands. So I need to digest the rest of your test. I did peak at the EV for R2, WOW. Much higher than I thought, but not a total surprise with S3 EUCHRE HAND. S3 can play anything.

I did not get as definitive results as it so variable what S1 & S2 would do. S4 is different if STD or not. But the kicker here is that S3 if all pass R1, now has a powerful R2 Euchre hand. The AH went down, he now has KH/QH, ace equivalent plus the AC. Two aces if Spades or Diamonds is called with the Right bower in either suit. Weaker in clubs and that is where S2/S4 could score if they had the JC with aces.

But the issue is, as you noted, is this STD? If STD, obviously he MUST pass R1. Euchre rate will be high for S2/S4 calls and sweeps for S1 calls. But what happens in R2 is that S2/S4 also passes (my test) unless this is STD. Any calls by S1 gets great results, including weak one, including next. So R2, ALL pass again if STD is not in force for the Dealer.

The issue I had in R2 was how aggressive S1 and S2 would be. So variable, I could only make estimates. Definitely though S3 EV will exceed R1.

This has some neat implications.

Thanks for doing this test and I will probably have more comments after reviewing in more details.

This is good stuff Ray.

Thanks,

IRISHWOLF

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

Unread post by irishwolf » Tue Jan 04, 2022 12:29 am

Ray, got through the rest of your test. JUST WOW on how many hands you did! Over the top.

So obviously S1 has to be calling with any 3 trumps but seldom with just 2 Diamonds in next or his calling would exceed 45%. But with any 3 trumps results in a great EV (S1). Mostly, as a result of the strong hand S3 has. Having 3 clubs, spades or diamonds will actually be quite high due to S3's hand. Crossing suit (review of my test) was quite high. Says a lot for Next!

If I had a magic wand, it would be nice to know the suits called by S1 and S2 on R2. That's okay, I know roughly those suits by review of my test data.

My take is that with a good partner or even an average player at S1 (and against aggressive players), I can be more conservative, he he! It's all about adjustment to score and situation.

IRISHWOLF

P.S.
You have earned my respect for your Simulator // programming. Readers of this Forum should take extra notice. When are you going to put it on the market, lol?

If I were you, DO NOT try and change your programming to accommodate STD. Actually, that is a variant of standard euchre and R2 players adjust accordingly (more conservative by all but S4).

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

Unread post by raydog » Tue Jan 04, 2022 1:48 pm

I did the simulation you requested, Irish, just to test my intuition. Here is what I predicted:

S1R2: D 40-45%; S 30-35%; C 25-30%
S2R2: D 25-30%; S 35-40%; C 30-35%

And here is what I found (100,000 hands, 53,528 made it to R2):

S1, D: 1,765 / 22,861 [alone / w/p] 67% of S1 bids
S1, S: 724 / 6,485 20% of S1 bids
S1, C: 852 / 4,033 13% of S1 bids
36,720 bids by S1

S2, D: 722 / 1,869 25% of S2 bids
S2, S: 218 / 3,702 38% of S2 bids
S2, C: 158 / 3,552 36% of S2 bids
10,221 S2 bids

The success of these bids is in line with what I reported earlier for 1,000,000 hands - simply divide those figures by 10 and you'll be close.

I wasn't so far off on S2 bids, but very much underestimated the proportion of Diamond bids by S1R2.

It makes sense that C bids are slightly less than S bids as there are only 5 of those cards remaining outside of S3's hand, as opposed to 6 S. S1 will favor a D bid (next) - much more than I expected, in fact - but if S1 passes, R2, they must 0 or 1 D, meaning S2 likely has them and is willing to defy convention and bid next (which is why S2R2 D bids are not lower).

As I explained earlier, my program simply uses a point system to determine which is the best suit to bid, and there is a bonus added for a next bid by S1R2. So even when that suit doesn't seem as strong as the other color suits it may still get the higher point value and be preferred. I of course can't guarantee that S1 always bid the best suit, I can only say that I compared thousands of possible hands to determine my point values, and do seem to work when I run these various scenarios for hands presented on this forum. I'd guess it picks the right suit at least 90% of the time, and 5% of the time it's too close to have a solid preference.

Of course, KNOWING what cards S3 holds could change which is the best suit to bid, but that's cheating! The decision can only be based on the cards a player sees and the fact that everyone before them has passed.

How does this compare with what you predicted, Irish?

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

Unread post by Wes (aka the legend) » Tue Jan 04, 2022 2:28 pm

raydog wrote:
Mon Jan 03, 2022 8:01 pm
I simulated this hand over 1,000,000 random iterations (S3 and turn card fixed).

Since I developed my program without STD (since that's the way I learned to play), I first checked what S3 should do, R2, if they pass R1.

(100,000 hands)
S3 bids Clubs, R2: 1,011 / 4,110 / 1,357 [2 pts. / 1 pt. / euchred] EV = +0.55
S3 passes, R3 (STD): S4's results
0 / 138 / 17 || 0 / 2,674 / 3,649 [4 pts. / 1 pt. / euchred || 2 pts. / 1 pt. / euchred]
(1st 3 numbers are for bidding alone; last 3 numbers are for bidding w/ partner)
EV = -0.70, which means an EV of +0.70 for S3.

So better for S3 to pass, R2, if STD is in effect. UNLESS S2/S4 are at 9 pts, in which case it is better to bid Clubs (S2/S4 only earn points 1,357 games instead of 2,812, or about 1/2 as often)

Here's the data from the main simulation:
S3 bids, R1: 80,367 / 531,321 / 254,085 EV = +0.21
865,773 games make it to S3 for a bid
S2/S4 score at least one point 254,085 games
S1 bids about 1.6% of games for an EV of +1.57; S2 bids about 11.9% of games for an EV of -0.91.
I didn't include those games in my calculations because they don't vary and so are simply dilutive.

S3 passes, R1:
S4, R1: 0 / 25,331 / 71,140 || 0 / 115,366 / 120,131 EV = -0.73; 33.2% of games
S1, R2: 19,362 / 14,352 / 208 || 110,151 / 197,754 / 24,227 EV = +1.26; 36.6% of games
S2, R2: 0 / 10,240 / 736 || 0 / 47,826 / 43,485 EV = =0.30; 10.2% of games
S3, R2: (passes)
S4, R2: 0 / 1,344 / 197 || 0 / 26,680 / 37,244 EV = -0.72; 6.5% of games
overall EV for S1/S3 = +0.90
S2/S4 score at least one point 251,222 games

From this data it seems clear that S3 should pass, R1, with this particular hand as their EV is much higher. But we do need to think about score. As XaviRonaldo pointed out, S2/S4 can't possible score 2 points, so if the opponents have 8 points or less then passing is a good strategy: better EV AND opponents can't win the game on this hand. But what if they have 9 points? When S3 passes, R1, S2/S4 STILL earn at least a point less often than if S3 bids. And even if S4 decides to bid more often, R1, to try and win that last point, they will only be marginally successful, and give up chances at euchring S1 in R2. On the whole, S4's best strategy would be to NOT bid more aggressively (but since they can't see S3's hand they of course WILL bid more aggressively, and this particular case it will hurt them).

Of course, 254,085 vs. 251,222 is perhaps too close to be significant. I concede that. But I did do a simulation over 100,000 hands, comparing if S4 bid normally, R1, vs. calling EVERY hand. In the latter case they were euchred mercilessly BUT were able to eke out an extra 6,200 hands where they scored the much needed point. Unfortunately, this meant that no games went to R2, so they missed out on 9,400 potential games where they would have scored a point. So, indeed, better to not be aggressive.

Is this what you were looking for, Irish?
This is basically the same hand from my Advanced Quiz #16:

viewtopic.php?t=111

I strongly advocated passing this hand. I'm glad your data back me up. Once again your simulator impresses me.

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

Unread post by irishwolf » Tue Jan 04, 2022 4:25 pm

Ray, we are close: Of course when close between Ds and either, Next is called. So a slight bias, AGREED! And just the reverse for S2, favoring green. S3 I find because S1 & S2 pass - can call anything he wants. And if it gets to S4 - Next is favored on all ties.

I had S1R2: D 46%; S 37%; C 17%

S1R2: D 40-45%; S 30-35%; C 25-30%

IRISH

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

Unread post by raydog » Tue Jan 04, 2022 5:30 pm

Wes, thank you for the link to your quiz from 2019 - I wasn't on the forum then. Good fodder to test my program! I'll be posting about that soon (I'll use the thread you or Irish started on March 11, 2019).

Ray

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

Unread post by Wes (aka the legend) » Tue Jan 04, 2022 7:14 pm

raydog wrote:
Tue Jan 04, 2022 5:30 pm
Wes, thank you for the link to your quiz from 2019 - I wasn't on the forum then. Good fodder to test my program! I'll be posting about that soon (I'll use the thread you or Irish started on March 11, 2019).

Ray
Just wanted to point out that I now disagree with 9). S1 should just go alone in hearts, lead the Right and then sell out and lead the Ace of trump whether the Left is revealed or not. This is of course a desperation loner situation.

6) is an interesting hand to test being that it's a marginal 2 trump loner.

17) and 26) are probably too ambitious when it comes to desperation loners in an attempt to win the game, and those hands are a complete no-go in tough game where S1 donates.

I have a perhaps crazy hypothesis that 23) and 25) are simply standard loners except when one's team has 8 or 9 points or when one's team has 7 pts when they are tied or in the lead. In the latter scenario a loner sweep is now only worth 3 points so that changes the math.

Post Reply