Variance: Video BJ vs Video Poker

Hey all,

I was hoping someone could help me understand the numbers on this one.

There is an (I believe) IGT video BJ game at my local casino, with the following rules:
Single-deck reshuffled every hand, dealer hits S17, no doubling or splitting, BJ pays 1:1.
I believe the house edge is ~4%, but would appreciate a more precise calculation.

I think that puts the house edge of this game pretty similar to 6/5 JoB, but that the variance of the BJ would be much lower since every results is either -1, 0, or +1 while any video poker game has the opportunity for some much bigger wins, while more commonly getting losses.

But what is the actual variance of the video BJ game? I'm specifically thinking about, suppose I'm close to reaching a tier credit bonus (for example) and want to get a nominal amount of points to achieve the goal. Of course a different video poker game, particularly 9/6 DDB (also available in this casino) will have a lower house edge, but the much bigger variance would expose me to much more losses, potentially, while the video BJ game would have more predictable outcome.

For example, if I were to play $5/hand on this video BJ game (vs $1 5-coin 9/6 DDB), how much would I expect to lose in an hour of play - assuming same playing speed?

Thanks so much for your input!

I had some massive short-lived beginner's luck on the VBJ game. Come back the next day, it was all bad luck. I should plan on a 20-25 hand buy-in per session and see how high I can get, it could be substantial, or it could be a one way trip to $0. If I won enough on this machine, I should just cash out and go to a real $15 or $25 BJ table.

With VP only the quads are getting me ahead substantially and so far I can't expect more than 3 in half an hour, and sometimes 0 in 2 hours. I would expect $10 per 25 cent per hand would be a good session buy-in; $50 per 5 coin quarters, or $200 per 5 coin $1, or $1,000 per 5 coin $5, or $2,000 per 5 coin $10, or $5,000 per 5 coin $25. Again, this money could all go down the drain in 15-30 minutes. Quads will extend your play, but eventually the bottom will fall out. Cash-out appropriately.

It looks the hose edge is 2.936% but that is not including the no double rule (which I have not seen on Game Kings).
http://www.bjstrat.net/cgi-bin/cdca.cgi

When available, sometimes Casino Wizard craps may be a good option if it earns points. The local Caesars has a 3x odds craps that offer $50 per tier point (regular video poker is $10 per point). It's a 0.34% house edge on don't pass which compares favorably to the 1.56% house edge on 9/5 jacks or better, given that there is no strategy involved other than properly playing the don't pass odds or a three point dolly/molly. Since about 2% of the return on video poker is through the royal flush, this would also avoid W2-G's. The problem is that playing on a Casino Wizard machine is very boring.

How many dollars per tier credit for slots?
How many dollars per tier credit for video blackjack?

It may be better to play an amusing slot, from an expected loss per credit earned standpoint. (Yes, variance would be higher.)

Thanks to the replies so far!

This is a Caesars property where the video BJ or other video poker is $10/tier credit, but some of the highest return payables will be $50/tier point. Slots would be $5/tier point.

I fully understand that the lower the house edge is, the lower the expected loss will be "in the long run." But in a finite-length session, and particularly for arbitrarily "short" sessions, variance has a big impact on the theoretical range of likely results in a particular session.

I know that playing a session in a negative-EV game will ultimately, eventually lead to a loss. I hope we all know that. But, low variance will make it more likely to "lose slowly," which leads me back to my original question: what would the variance of the video BJ game be in my original post, and how would it impact the likely results of playing for just 1 hr compared to the "more favorable" 9/6 DDB option?

I have only the coarsest estimate:
Variance of Blackjack is around 1.3
Variance of DDB is around 42.

This particular "blackjack" probably has an even lower variance, because of the 2 for 1 pay on player natural 21.

Thank you for these variance numbers, super helpful!

For anybody else who is curious, I used the gambling session calculator on wizardofodds to crunch the numbers based on these variances.

After estimating the house edge of the BJ game to be ~4%, assuming $5 per hand and 500 hands per hour for one hour, and a standard deviation of ~1.14 (which might be "too high" for the game in reality), there is a 99% chance of the result landing somewhere between losing $396.51 and winning $196.51 with the 50/50 point at losing $100.

For the 9/6 DDB game, using a house edge of ~1.02%, assuming $5 per hand and 500 hands per hour for one hour, and a standard deviation of ~6.48, there is a 99% chance of the result landing somewhere between losing $1710.91 and winning $1659.91 with the 50/50 point at losing $25.50.

Also interesting, the chance of having a winning session in the DDB game is about 48.5%, but only 21.6% for BJ.

Thanks all for your help!

If I played VBJ and won 5,000 ($50 of) points, that'd be $15,000 of coin-in @ $3/point. Play gets extended when I'm not on a losing streak and the points can pile up fast while I'm ahead. But it is really sad to see that I've got a 1 in 5 chance of having a winning session on it. On the upside, it does what I want, only a bit less, or maybe if I was playing longer, it'd get there alright.