Blackjack tip betting EV?

Hey WoV, I've lurked these forums for a little bit now but I finally have a question I'm curious enough about to ask. Sorry if this is the wrong spot for it.

I work as a tablegames dealer, and recently got in an argument with some other dealers about toke bets. If a player gives me the option to either take a tip or bet the tip I always bet it, and they think I'm losing money in the long run to the house edge. But what they fail to think about is that over 90% of the time if they player is playing correctly and doubles/splits all hands they should, they typically double/split your toke bet along as well. How much does this affect the EV of betting tokes? Its basically free splits/doubles, with no real downside.

I read that there is a 12.15% chance of getting a double/splitworthy hand in blackjack, and that means there's a 12.15% of having your toke bet doubled, possibly even more than doubled with resplits and DAS. Am I overlooking something here? I'm not sure how to fully calculate this but it seems like it has to be +EV in the long run.

Great question. I've always though it was crazy how dealers would always want you to gamble their tip but you bring up a good point with the doubles and splits. I think most people double the tip on doubles but I dont necessarily always see them put out another tip on splits. I'll be curious to see the ev on this.

Quote: glhf

Hey WoV, I've lurked these forums for a little bit now but I finally have a question I'm curious enough about to ask. Sorry if this is the wrong spot for it.

I work as a tablegames dealer, and recently got in an argument with some other dealers about toke bets. If a player gives me the option to either take a tip or bet the tip I always bet it, and they think I'm losing money in the long run to the house edge. But what they fail to think about is that over 90% of the time if they player is playing correctly and doubles/splits all hands they should, they typically double/split your toke bet along as well. How much does this affect the EV of betting tokes? Its basically free splits/doubles, with no real downside.

I read that there is a 12.15% chance of getting a double/splitworthy hand in blackjack, and that means there's a 12.15% of having your toke bet doubled, possibly even more than doubled with resplits and DAS. Am I overlooking something here? I'm not sure how to fully calculate this but it seems like it has to be +EV in the long run.

You are correct. If the player doubles the dealer bet when he doubles his own that definitely is +EV for you. If he doesn’t add a chip for you when he doubles that is a -EV for you, as an example, if he doubles a 9 against a 6 and gets a deuce your single bet is stuck at 11.
If he splits and has to put an extra chip for you that is big +EV.

I think that there is greater value in playing with the player. I think he is more likely to tip more frequently if you bet it rather than just picketing it. But I have no proof.

I also think that taking the toke as a bet is better for the dealer; for the reasons mentioned, and because the player will be more likely to re-bet to make sure that the dealer is taken care of if the original toke loses.

Yeah greasyjohn, I agree with that for sure, there are definitely mental aspects to this as well. Appreciate the input :)

Quote: Greasyjohn

I also think that taking the toke as a bet is better for the dealer; for the reasons mentioned, and because the player will be more likely to re-bet to make sure that the dealer is taken care of if the original toke loses.

This. Someone who bets for the dealer usually puts up far more over the course of a down than they ever would have handed in.

Greasyjohn has the best point in the thread so far about re-betting. I can say that after toke betting for a dealer if it loses there's "some" small % that I will rebet it because it loss and I want them to win too, with me =).

However, as far as the original premise, I agree with your coworkers. As the OP said, assuming the player is playing correct basic strategy (splits/doubles/etc). The original bet when placed is directly subjected to the house edge of the game while playing basic strategy "well." Thus, before the cards are dealt if you have a $10 bet on the table, and a .5% generic HE game, the EV of that hand is EV = 10*(-.005) = -$0.05. Why would this math change if the bet is a $1 tip?

The tip is the same thing as adding $1 to the top of the player bet. Thus, it's obviously not advantageous for the player to bet more, as he's exposing more to the house edge (assuming a non-counting basic strategy player). From the dealers perspective, your EV of the tip unplayed is $1. From the $1 tip being played it's EV = 1*(-.005) = $0.995. There IS value being lost here to the house edge. in the long run.

Also, hope they're doubling A-7 to 2-6 in H17 game =)... most probably aren't!

Quote: Romes

Also, hope they're doubling A-7 to 2-6 in H17 game =)... most probably aren't!

They're ever so slightly more likely to make that double, because some of them ask the dealer "It's you're money too, what are we doing?", and if the dealer knows what they're talking about, they're saying "Double."

Using Wizards probability for each outcome, and assuming player play perfect basic strategy and always split/double for dealer as well.
The dealer will hower never lose more than 1 unit.

Result player	Result dealer	Probability	player	dealer
-8	-1	0,00000019	-0,00000152	-0,00000019
-7	-1	0,00000235	-0,00001645	-0,00000235
-6	-1	0,00001785	-0,0001071	-0,00001785
-5	-1	0,00008947	-0,00044735	-0,00008947
-4	-1	0,00048248	-0,00192992	-0,00048248
-3	-1	0,00207909	-0,00623727	-0,00207909
-2	-1	0,04180923	-0,08361846	-0,04180923
-1	-1	0,40171191	-0,40171191	-0,40171191
-0,5	-0,5	0,04470705	-0,022353525	-0,022353525
0	0	0,0848329	0	0
1	1	0,31697909	0,31697909	0,31697909
1,5	1,5	0,04529632	0,06794448	0,06794448
2	3	0,05844299	0,11688598	0,17532897
3	5	0,00259645	0,00778935	0,01298225
4	7	0,00076323	0,00305292	0,00534261
5	9	0,00014491	0,00072455	0,00130419
6	11	0,00003774	0,00022644	0,00041514
7	13	0,00000609	0,00004263	0,00007917
8	15	0,00000066	0,00000528	0,0000099
EV			-0,0028	0,1118

The player return show the well known -0,28% EV while the dealer have an positive EV of 11,18% (If the player keeps the bet after a surrender "dealer EV" would be 8,95%)
The difference is of course payed by the player.

Quote: Torghatten

Using Wizards probability for each outcome, and assuming player play perfect basic strategy and always split/double for dealer as well.
The dealer will hower never lose more than 1 unit.

Result player	Result dealer	Probability	player	dealer
-8	-1	0,00000019	-0,00000152	-0,00000019
-7	-1	0,00000235	-0,00001645	-0,00000235
-6	-1	0,00001785	-0,0001071	-0,00001785
-5	-1	0,00008947	-0,00044735	-0,00008947
-4	-1	0,00048248	-0,00192992	-0,00048248
-3	-1	0,00207909	-0,00623727	-0,00207909
-2	-1	0,04180923	-0,08361846	-0,04180923
-1	-1	0,40171191	-0,40171191	-0,40171191
-0,5	-0,5	0,04470705	-0,022353525	-0,022353525
0	0	0,0848329	0	0
1	1	0,31697909	0,31697909	0,31697909
1,5	1,5	0,04529632	0,06794448	0,06794448
2	3	0,05844299	0,11688598	0,17532897
3	5	0,00259645	0,00778935	0,01298225
4	7	0,00076323	0,00305292	0,00534261
5	9	0,00014491	0,00072455	0,00130419
6	11	0,00003774	0,00022644	0,00041514
7	13	0,00000609	0,00004263	0,00007917
8	15	0,00000066	0,00000528	0,0000099
EV			-0,0028	0,1118

The player return show the well known -0,28% EV while the dealer have an positive EV of 11,18% (If the player keeps the bet after a surrender "dealer EV" would be 8,95%)
The difference is of course payed by the player.

Yep, multiply that by the actual average bet for flat betting 1.xxx and it's even more.
I.e. 1$I nitial toke bet means 1.xxx$ average bet with splits and doubles