Fun 21 (Princess, HAL, Carnival,...) - Early Surrender Optimal Strategy

[Big_G_111]

Let me get started with the tl;dr version:

 

Fun 21 on Carnival plc ships - you should take early surrender in the following cases:

16 v dealer 10

8-8 v dealer 10

7 vs dealer A

13-17 vs dealer A

7-7 , 8-8 vs dealer A

 

—-

 

Long version :

 

On the Carnival plc ships, they offer a blackjack variant called “Fun 21”. You can find the details elsewhere - this game is essentially Spanish 21 in the US or Pontoon in Australia.

 

However, one difference is that early surrender is offered - you can toss in your hand after seeing your cards and the dealer’s up card, but BEFORE he/she checks for a dealer blackjack, and get half your wager back. This is different than the more common late surrender, where you can only pull out after the dealer has checked for blackjack (and taken all the bets if he/she has it). 

 

You can find the optimal strategy for Spanish 21/pontoon online. What you CAN’T find is the optimal early surrender strategy, because AFAIK, the Carnival plc ships are the only places on earth that run this particular game, and no one has bothered to compute it.

 

So I ran the numbers myself, using an infinite deck approximation. I saved myself some trouble by only checking those hands that would call for early surrender in normal BJ (14-16 vs 10, 5-7 and 12-17 vs A).

 

Here are the numbers I got for expected return - anything below -0.5 means you should early surrender:

16 v 10 -0.52647

15 v 10 -0.48701

14 v 10 -0.44426

5 v A -0.46416*

6 v A -0.48183*

7 v A -0.51335**

12 v A -0.49465

13 v A -0.53353

14 v A -0.56941

15 v A -0.60253

16 v A -0.63311

17 v A -0.68005

 

I also ran some edge cases - 16 v 9, 17 v 10, 18 v A - to make sure I wasn’t missing any additional surrender cases - they all cleared the -0.5 hurdle. 

* This number excludes the possible gain from drawing to a hard 11, then doubling. I didn’t compute the gain because these already are good enough to not surrender.

** This number also excludes the gain from drawing to hard 11 (via a 4 or a pair of 2s), then doubling. I ran the gain from that times the odds of getting to that spot, and the gain was still below -0.5, but not by much. 

As far as splitting 8s v 10 or A - doing that puts you into -0.6 territory before accounting for the chance of doubling, and there’s no way the double option comes close to making up that much money. Toss it in. 

I might have made some calculation errors, but I’m pretty confident these are the right plays. 

 

The 16 v 10 and 7 v A are pretty close to even, so no big deal if you want to take your chances with them. But you should definitely be giving up on those stiff hands v A. Interestingly, the worst hand in the game is 17 v A, not 16 vs A or 16 v 10. 

 

All in, this looks to trim about 0.2% from the house edge, bringing it down to the 0.6-0.7% range. Not as good as 3:2 blackjack, but better than 6:5 blackjack. 

 

Hope this helps. I enjoy playing this game, and y’all should give it a go. Just be sure to print off the strategy before you sail - it’s NOT the same as for regular BJ. 

 

 

 

[Big_G_111]

And, of course, I realized some errors in my calculations 5 minutes after posting. 🙂

 

Specifically, I need to account for hitting 17 v A (believe it or not, that’s the optimal play), and soft 18 v. A.

 

I will revise and post later. Spoiler alert: this might make the difference on 7 v A early surrender.

[Big_G_111]

tl;dr version:

 

Corrected Early Surrender Optimal Strategy for Fun 21 on Carnival plc ships

Hard 16 v 10

Hard 13-17 v A

(Includes 7-7 & 8-8, excludes soft hands)

 

——

Long version:

 

OK, here are the corrected numbers. (I realize few people here will care about the specifics, but I’ll leave them here anyway for those who stumble upon this post years later.)

 

5 v A -0.46159

6 v A -0.47934

7 v A -0.49838

 

12 v A -0.49233

13 v A -0.53138

14 v A -0.56743

15 v A -0.60070

16 v A -0.62142

17 v A -0.66625

 

As I guessed earlier, this DOES make it so you should NOT early surrender the 7 v A. 

 

—

 

 

 

[Big_G_111]

One last clarification before I sail - 

 

The hard 16 v 10 is only early surrendered if it’s 10-6 or 9-7, not 8-8, which should be played out as a split. 

 

—

 

Let me explicitly list every play.

 

One should early surrender the following:

 

Hard 16 v 10 (but not 8-8 v 10)

 

Hard 13 v A

Hard 14 v A (including 7-7 v A)

Hard 15 v A

Hard 16 v A (including 8-8 v A)

Hard 17 v A

 

Edited February 9, 2019 by Big_G_111

[Petersonfcu]

On 2/9/2019 at 2:34 AM, Big_G_111 said:

 

 

Corrected Early Surrender Optimal Strategy for Fun 21 on Carnival plc ships

 

 

(I realize few people here will care about the specifics, but I’ll leave them here anyway for those who stumble upon this post years later.)

 

—

 

 

 

Hey Big-G, and along I come five years later looking for this EXACT information!   Thanks!!!!!  And if you are still on the boards, can you provide me a link to the infinite deck software you used to run these calculations??

 

Paul

[Big_G_111]

All right! I actually found one other person that could use this info!!!

 

As for the software - I did it myself in Excel. :)  (I was an actuary for a short time in my ancient history.)

 

I looked at the commercial BJ analysis programs, but it wasn't clear that I could modify them to incorporate the Spanish 21 deck probabilities and rule mods. 

 

So, I read a couple of papers online on how to do infinite-deck regular BJ calculations. (Over the years, there have been a few math/statistics majors here and there choosing this for their senior thesis.) I then set up a set of spreadsheets in Excel, only dealing with the 10-12 cases I listed in the first post. 

 

Took me about 10 hours for the initial setup. The key was to first run them using standard BJ, so I could compare the results to the readily-available results of others, then fix any errors I made. 

 

Once they were 100% working for standard BJ, I first shifted the probabilities to the 48-card deck of Spanish 21, then adjusted the payouts and player rules to those of Spanish 21. (Dealer rules don't change in Spanish 21.) That took me another 5 hours or so, then 2-3 hours to fix the omissions, as seen in posts 2, 3 & 4.

 

I saved maybe $100 total from these surrenders on my cruise, so I guess I'll have to go on more cruises to fully recoup my time investment. :) 

[SunsetCruizen]

Hey Big_G_111 - Thanks for working this out.  I will make use of this on my Carnival cruise next April.

[knagl]

Thank you for the interesting post and all the number crunching. I like to get the best value out of my gaming dollar, so I'll for sure bring this info with me on my upcoming cruise.

[Buffettfan 225]

following, any updates?

[Big_G_111]

On 8/10/2022 at 3:24 AM, Buffettfan 225 said:

following, any updates?

Nope - the Fun 21 game, as dealt on the various Carnival lines, hasn’t changed, as far as I know.

 

I wish one of the mods would edit the first post, so that the final, accurate info is front and center for this thread.  Absent that, I’ll just repeat the full set of early surrender situations :

 

—

 

Hard 16 v 10 (but not 8-8 v 10)

 

Hard 13 v A

Hard 14 v A (including 7-7 v A)

Hard 15 v A

Hard 16 v A (including 8-8 v A)

Hard 17 v A

 

—

 

Also, don’t forget to use the Spanish 21 chart for play decisions, Specifically, do NOT play this game the same as regular BJ - that will give the house an extra 2-3% house edge.