Red 7 vs Kiss 3
  • Had a question about the red 7 vs the kiss 3 counts. Have been using the red 7 count for a couple of years and have just started using 20 as the "mean" instead of 0 to avoid the hassell of negetive numbers.Using the red 7 I start the count for a dd game at 16. But the kiss 3 starts the count for a dd game at 17.Since they are similar systems how come the kiss 3 doesn"t start at 16?Is there a math reason for this?Thanks in advance for your replies.
  • I wondered the same thing. I have been using Kiss III and now changed to red 7. The reason i changed i got hit hard the other day w/Kiss III, and also i am playing on my computer with Hoyl's BJ and it has a counting system Hi LO i think any way at a count on kiss III of like 27+ my computer showed low card heavey instead of high card.
  • Red 7 and Kiss III are virtually identical systems with regard to style of design, complexity and performance. The only thing is that Kiss III is more completely structured and user friendly when it comes to defining IRC's (initial running counts) and index numbers.

    If you have decided to begin your Red 7 double deck count at "16", then at what running count do you begin ramping up your bets? An RC of "19" would give you an average "true count" of +1.2. An RC of "20" would give you a +2.0 true count. But you don't need a +2 true to begin betting multiple units! In a typical double deck game, you have a 0.25% edge at +1 true (counting what's gleaned from basic strategy departures at that point).

    That's why Kiss III calibrates the starting point for double deck at "17" and begins ramping up the bets at a nice round "20" -- a rise of three points.

    If you prefer the Red 7, there nothing at all wrong with using it with all the Kiss III IRC's and index numbers. Arnold Snyder had a heck of an idea when he came up with the Red 7 Count over 20 years ago. I just modified it into something I could call my own, then took the ball and ran a little further downfield with it.
  • Renzey:
    Would you comment on the ramp and bet spread of Red 7 vs. KISS?
    I have looked at both systems and feel Kiss is a lot more aggressive. Using KISS, I have hit some low spots in the BR which Red 7 might not produce.
    Thanks.............
  • Renzey said:
    Red 7 and Kiss III are virtually identical systems with regard to style of design, complexity and performance. The only thing is that Kiss III is more completely structured and user friendly when it comes to defining IRC's (initial running counts) and index numbers.

    If you have decided to begin your Red 7 double deck count at "16", then at what running count do you begin ramping up your bets? An RC of "19" would give you an average "true count" of +1.2. An RC of "20" would give you a +2.0 true count. But you don't need a +2 true to begin betting multiple units! In a typical double deck game, you have a 0.25% edge at +1 true (counting what's gleaned from basic strategy departures at that point).

    That's why Kiss III calibrates the starting point for double deck at "17" and begins ramping up the bets at a nice round "20" -- a rise of three points.

    If you prefer the Red 7, there nothing at all wrong with using it with all the Kiss III IRC's and index numbers. Arnold Snyder had a heck of an idea when he came up with the Red 7 Count over 20 years ago. I just modified it into something I could call my own, then took the ball and ran a little further downfield with it.

    Thanks Renzey those are the numbers I was looking for. Was hoping you would weight in on this. Thinking of possibly switching to the kiss 3. Finding it easier to identify the black 2 as a + card than the red 7 as a + card.
  • Sage 01 -- If you're talking about the recommended betting ramp for the Red 7 Count in Arnold's latest edition of Blackbelt in B/J (2005 printing), it certainly is quite conservative.

    In the shoe game, it reaches its top 8 unit bet at about a true count of about +7. In double deck play, it reaches its top 6 unit bet at about +5 true. Those are s - l - o - w ramps.

    When I establish a betting ramp, I look for the smallest ratio between expected hourly win and hourly standard deviation. For example, in the 6 deck game, using KISS III with a spread of $25-to-$250 and reaching that $250 bet at an RC of "23" (about +3 TC), the expected hourly win is $39 and the hourly standard deviation is $1000 (via simulation). That's a ratio of about 25-to-1. The smaller the ratio, the better.

    Now, slowing down the ramp to reach the $250 bet at that which is recommended for the Red 7 ("33" RC in this case) yields an expected hourly win of just $6 with an hourly standard deviation of only $450.

    The first thing we can see here is this slower ramp can use much larger bets to arrive at the same hourly standard deviation (incurred volatility). So we can increase the bets to be $56-to-$560 and now we'll experience the same volatility during our playing sessions (a $1000 hourly SD). However, our yield will be an expected win of just $13 per hour (a 76-to-1 ratio).

    If we wanted to play it safer with our bankroll and have only a $450 hourly standard deviation, we could've just used the faster betting ramp (max bet @ "23"), reduced the bets to $11-to-$110 and had an expected hourly win of $17 instead of $6.

    I haven't gone thru the double deck comparison, and it'll probably be less dramatic.
  • Betting Ramps (continued) -- With these comparisons having been done, there still was no mention made of how many units are in the player's bankroll. In Bluebook II, it was recommended (pg 183) that the player have 120 basic units for a session stake and 1000 basic units in his master bankroll. These are the criteria I use when I play. During prolonged losing periods, I either replenish my master bankroll up to its original level or shrink the size of my units fit the 120/1000 format.

    Arnold's Blackbelt in B/J (pg 154) recommends 150 basic units for a master multi-deck bankroll, and I don't necessarily disagree with that. Over the years, there have indeed been two or three times when a majority chunk of my 1000 units has eroded away during a 200 or 300 hour losing streak. If your bankroll is unreplenishable from another source when needed, then you need a considerable extra measure of conservatism to your bet sizing.

    However, I don't think that should be done in slower betting ramps -- but in smaller betting units! My reason again for saying that, is the superior "hourly win-to-hourly SD" ratio with the quicker ramp.

    In closing, Arnold's latest Blackbelt edition contains two very educational chapters (ch 12 & 14) on bet sizing and risk of ruin. I consider him to be the foremost source of reference for real world card counters and have learned a ton from reading his stuff.

    Correction: I errantly said that Blackbelt's recommendation for betting 6 units with the Red 7 in double deck play would come at a true count of about +5 -- but it would actually be at about +9! That would occur only around 1% of the time if 60% of the cards are dealt out. I'm sure that's got to be too slow.
  • Fred,
    What is your view on using Kelly optimum betting with Kiss III? For instance, assuming a player with a 400 unit bankroll in a 0.5% initial house edge 6D game:


    RC TC Edge Bet
    19 +1 0.0% 0u/1u
    21 +2 0.5% 2u
    23 +3 1.0% 3u
    26 +4 1.5% 5u
    29 +5 2.0% 6u
    30 +6 2.5% 8u
    31 +7 3.0% 10u

    The preceding values are at 80% Full Kelly. For example, at Running count of 30 the true count is approximately +6, thus

    Optimum Bet= [(Player's Edge) x 80%] x Bankroll
    = (2.5% x 80%) x 400u
    = 8u

    Edit: I realize that since a count higher than the pivot is only an estimation of the true count, the bet may not be so optimum afterall so this ramp may not be very effective. However, the problem still remains with the player here in this hypothetical situation with a 400 unit bankroll: he does not have enough units to play the KISS III ramp comfortably and lowering his bet unit value may not be possible (assuming that 1 unit = 5 dollars, and that five dollar tables are the lowest limits available). Considering Red 7's ramp as being too slow, is there an alternative ramp you would suggest that falls somewhere between K-III's standard ramp and Red 7 for our 400 unit player?
  • Deviance - Running a quick 10 million hand simulation with your proposed betting ramp using $5-to-$50 bets yielded an EV of +0.42%, an hourly win of $3 and an hourly standard deviation of $113.

    The $50 bets came less than 1% of the time and a $0 bet was used 8% of the time due to poor counts. The $3 per hour/$113 SD is a 37-to-1 ratio - a poor yield.

    You could get a higher hourly win with the same hourly SD by shrinking the units and speeding up the ramp -- but smaller units are not a possibility, so I see your problem.

    Unless you're doing it just for the love of playing, I think I would bite the bullet, speed up the ramp and bet at about 1.25 Kelly while you're down at these stakes. Remember however to bet at a lower Kelly if your bankroll shrinks significantly. If your bankroll grows, keep your stakes the same but speed up your betting ramp gradually. Your goal would be to get to a 1000 basic unit bankroll, then reach a 10 unit bet by no more than "25" RC.
  • Deviance -- Ran another sim using the following ramp:

    4 or less, bet $0
    21 $12
    22 $15
    23 $25
    25 $30
    26 $40
    28 $45
    29 $50

    Results were an EV of +0.59%, hourly win of $5 with an hourly SD of $148. That's about a 30-to-1 ratio; still poor but better. The chance of losing an entire $2000 bankroll anytime during the first 300 hours was 20% -- providing you never slowed down the betting ramp.

    You could probably be yet a little safer off by betting $0 at "6" or less. It would raise your overall EV as well as your hourly win and lower your SD -- all modestly.
  • Interesting results, thanks for posting them. It seems that 1.25 Kelly is better suited in comparison to 0.80.

    An inquiry: Assuming the player wongs in at TC +2 and wongs out on all lower counts, what would be the optimal spread and ROR? What EV would this carry?

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!