Announcement

Collapse
No announcement yet.

Kelly System vs Flat Betting

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Kelly System vs Flat Betting

    There are so many points that were covered in the "JR Miller's Record" thread that a
    number of them deserve their own threads. Here is my try:

    JR Miller has his "flat bet" approach (which sounds much better to me than any other
    system), and although JR has been very stern (OK, perhaps arrogant, too) with SLAMDUNK's assessment of the Kelly Criterion, I really enjoyed to see SLAM comment about what it takes to many any system work and that is the learning experiences of a gambler from his own mistakes. (see page 3 7/21 7:11pm) One must have stumbled through enough mistakes to actually reach a point where he can make use of any money management approach. What good is any system if the player does not have the capability,patience or discipline to implement it and stay with it?

    If you read that entry, SLAM concedes that, "Over the years I have made many, many bets that I have since learned should not have been made." Who among us hasn't? (There were many bets I have made that I knew ten pitches into the game I was dead.)

    It is only with this kind of long-term experience that one can develop the self-control, talent and instinct for betting sports. How often have we made plays that were handicapped after we had a fight with our girlfriend/wife, or after having only 3 hours sleep the night before, or we bet just to have "action", or because we had a grudge against a team or player we didn't like or whatever? (Maybe after a couple or six beers, laying 15 points didn't seem so scary after all!) What are our winning percentages from bets made in those situations?

    The point here is that we as non-professionals, many times don't know when to bench ourselves, to back off, when we're not in the best frame of mind to make a solid assessment of a game to bet or whether to bet at all. How many times have we suffered a bad-beat, and then turned around and said, "F--K It" and come back with a larger amount on the next game (usually the
    same day). How many of us have started with a modest bankroll betting games for $100-$200 a pop, and after finding ourselves in a hole down the road, we land up betting $800-$1,000 on a game we didn't even like nearly as much as some of the $100 plays we made earlier?

    True, gambling is not for those who lack self-control, but let's be honest, how many of us have found ourselves in a big hole with the games at some point in our betting careers? (Ok, everybody raise their hands...)

    This is why JR MILLER's preferred "flat-bet" system gets my vote. One can argue mathematics all one wants, but don't look past the real-life situation of why we find ourselves involved in betting the sports; the two main reasons for most of us are GREED
    and EGO (you can probably throw in LAZINESS, too - I plead guilty to all three!) The
    "flat-bet" theory is sound enough to force the player to think long-term, not bet games on a "whim" during any kind of streak, and does not encourage the player's internal propensity for greed (to make a "killing" during a win streak, thinking that one is on a "high").

    Overall, it is a smoother ride no matter how the streaks are running, and with some discipline applied, can keep the player from his own self-undoing.

    With the "flat-bet" approach, one starts with a bankroll, bets no more than 1%-2% of that bankroll on any given play, and risks EXACTLY the same amount on every straight bet. In working it this way, you treat your bankroll like a bank deposit earning interest, except the interest that can be earned with even a so-so season can dwarf the interest offered by any bank or other short-term investment. (You absolutely, positively bet the same amount on every game, yes even if you are playing a +2.40 underdog! Because if you play less on the dogs, you are costing yourself profits when they win, and at some point they will win.)

    If you are like JR MILLER and choose predominately underdogs, lay no more than 20 in juice in most cases (including parlays) - then at season's end - you do not need to win 52.3% to see some kind of profit. If you were to play almost all dogs at +1.05 or higher, if you hit even 51% of those games, you would still see a decent profit!

    It is this kind of long-term thinking approach that JR MILLER is trying to educate people with at his website www.professionalgambler.com The school of thought that says, "I need to increase my bets by a certain percentage as I go along, so that I can maximize my winnings by season's end" to me is not the best way to go, because you can get hurt in the short run, and in the long run it will play on your own greed factor when you start winning because it can change the "mental value" of your bankroll (What fun is it to play $200 games, after you have had a few $1000 or $1500 bets?). How many of us have the mental stamina to stay consistent with such a system over the long haul? The "flat-bet" approach takes enough discipline let alone a snowballing effect of increasing plays.

    To me, the biggest drawbacks to the "flat-bet" scheme are that you need to make a lot of plays to earn good cash (2-3,000 plays a year), and if you lay juice on most games you can juice yourself to death with only average winning results.

    In my opinion, the biggest factor that is self-undoing of gamblers is GREED. More than
    anything else, greed allows players to allow themselves to distort the value of money, to
    over assess their own capabilites to handicap an event, and to get involved in games they never should have bet because they're intoxicated with the thought of getting richer quicker than the laws of chance dictate - or to chase they money the just lost.

    We all know the tremendous mental pounding we take when we are in an extended losing
    streak. That dreadful feeling when you can't hit the broad side of a barn with your plays. Watching the players you bet on make turnover after turnover, or squander solid scoring chance after scoring chance, or blowing solid leads. Using a "flat-bet" system makes this stress an easier part of the cycles we all go through. It doesn't allow you to get too high when you win, or too low when you lose. It doesn't put you under extra-pressure to win a game or games just because you have more at risk on them. It can keep you focused and looking at the big picture, and in the long run, with even decent handicapping skills, WILL make you money. Will it be less money than another system? (Please don't tempt me, I trying like hell to not to be GREEDY!)

    Unlike, JR, I think SLAM makes a good case for what he believes works the best. But, the Kelly Criterion or other systems won't get my vote because:

    A) I believe in the long run it is better to play the same amount on every game, it keeps most players from fooling themselves that some plays or price values are "better" than others. In the short run one can make a score by picking right, but what happens if a few "big plays" are wrong here and there, is that bettor sharp(or tough)enough to adjust his play accordingly? And even if he can, do you realize the extra stress imposed on him by having to turn it around? The
    kind of stress that often times leads to disasterous outcomes? Why put too many eggs in one basket? (I plead guilty there, too!)

    B) For the Kelly Criterion especially it would appear the that the onus is on the player to assess an actual number as to the percentage chance of a team/player winning an event. Maybe SLAMDUNK or possibly the Kosher Kids can do that, but to be honest I can't. (By the way, what is that percentage number based on? I could see in the case of the Computer Boys knowing when they had a big
    percentage edge, but how do others arrive at a 58% probability of a team winning a game?
    Could somebody teach me that?)

    c) Having to calculate a different betting amount on each game may cause an undue distraction to the player, because it might negatively effect his choice of teams based on what he might win, rather than searching out the board to focus on the play(s)that WILL win.

    All anyone needs to do is to look at JR Miller's website and hold it up next to any other tout/advisor's site and you can see at first glance that JR is a top-notch act. Like his direct (arrogant) writing style or not, his site is the greatest effort I have come across to educate gamblers on what it takes to win in the long run. I believe on the site he showed his house being built in Tennessee. I don't think he earned the money for that by returning Pepsi bottles for 5 each. I have not met him nor have I subscribed to his services, but after visiting his site, there is so much food for thought there, that most will find themselves
    going back often, if for anything else just to see what games he picked tonight.

    Gentlemen, we can debate mathematics till the cows come home (I would give the nod to
    the "flat-bet" system there also), but like SLAMDUNK himself you need to have the
    experience and personal fortitude to make any system work, and I would say for most of us seeking to make a living from the games, or a consistent side income, the "flat bet"
    approach is the ONLY way to go. It makes for the smoothest ride possible along with the most realistic chance of financial gain.

    For me the right combination of mathematics along with an approach to help offset the
    insanity of gambling.

    Just because your holding Babe Ruth's bat in your hands, doesn't mean you can hit a home
    run.

    Good Luck to All! Larry-LB/FCC

    [This message has been edited by Landmark95 (edited 07-22-2000).]

    [This message has been edited by Landmark95 (edited 07-22-2000).]

    [This message has been edited by Landmark95 (edited 07-22-2000).]

  • #2
    Landmark95, in my opinion, for someone who has no idea what kind of win percentage to expect, flat-betting or betting a very low-percentage of bankroll makes sense. However, if someone has a long-term track record of consistently averaging a certain win percentage, in order to maximize profit potential, it makes sense to bet either a pure or modified Kelly system. Despite what J.R says, the mathematical proof of the efficacy of Kelly is irrefutable.

    Comment


    • #3
      .....Wow.....Larry, you have defined subjective reasons to use flat betting that I never even remotely would have thought of.

      I must say that I certainly did not invent the "flat betting" system described in my book and on my website. When I was doing research for the first edition of my book, "How Professional Gamblers Beat the Pro Football Pointspread" back in 1989, I enlisted the help of Dr. Jeffery Allbritton, Professor of Mathematics at Middle Tennessee State University. I wanted someone more qualified than I to double-check all the math in the book and to verify my conclusions. Dr. Allbritton's help was invaluable in establishing the fact that flat betting is the "best" betting system against sports.

      To tell you the truth, I expected derision and rejection of the entire book because of my own rejection in the book of the "bet-more-when-you're-winning-and-less when-you're-losing" philosophy, my rejection of the "1-star, 2-star, 3-star" philosophy, and my rejection of any form of the so-called "Kelly" criterion.

      I figured the book would be a failure.

      However, after my book came out, I was invited by Jack Painter and Bob McCune to speak to their classes on sports betting at Clark County Community College in Las Vegas. It was there I learned - to my absolutely gleeful surprise - of what Jack Painter called his "plateau" betting system, and Jack's system was virtually identical to the system described in my book. I have to say I was ecstatic. Jack Painter and Bob McCune were highly respected teachers of sports betting and it was unbelievably fulfilling to learn they already supported the same money management theories as in my book.

      Essentially, the rule is, risk no more than 2% of your beginning bankroll per bet and do not "jiggle" the size of your bet until your bankroll increases (or decreases) by roughly 50 percent. The same bet size should be used against 2-bet parlays paying no less than 13-to-5 and 3-bet parlays paying no less than 6-to-one. (Using parlays does not mean you can win more money, but you can limit your downside risk when you have multple bets. It is this factor that causes bookmakers to restrict their advertising the availability of parlays. It is not in the bookmaker's best interest for you to limit you downside risk.)

      Landmark, I believe the mathematical aspects alone are enough to "prove" the advantages of flat betting over using some sort of progressive betting scheme, but your shockingly insightful examination of the subjective factors involved (greed, etc.) are also enough - in themselves - to prove the value of avoiding progressive betting systems.

      Comment


      • #4
        .....Wow.....Larry, you have defined subjective reasons to use flat betting that I never even remotely would have thought of.

        I must say that I certainly did not invent the "flat betting" system described in my book and on my website. When I was doing research for the first edition of my book, "How Professional Gamblers Beat the Pro Football Pointspread" back in 1989, I enlisted the help of Dr. Jeffery Allbritton, Professor of Mathematics at Middle Tennessee State University. I wanted someone more qualified than I to double-check all the math in the book and to verify my conclusions. Dr. Allbritton's help was invaluable in establishing the fact that flat betting is the "best" betting system against sports.

        To tell you the truth, I expected derision and rejection of the entire book because of my own rejection in the book of the "bet-more-when-you're-winning-and-less when-you're-losing" philosophy, my rejection of the "1-star, 2-star, 3-star" philosophy, and my rejection of any form of the so-called "Kelly" criterion.

        I figured the book would be a failure.

        However, after my book came out, I was invited by Jack Painter and Bob McCune to speak to their classes on sports betting at Clark County Community College in Las Vegas. It was there I learned - to my absolutely gleeful surprise - of what Jack Painter called his "plateau" betting system, and Jack's system was virtually identical to the system described in my book. I have to say I was ecstatic. Jack Painter and Bob McCune were highly respected teachers of sports betting and it was unbelievably fulfilling to learn they already supported the same money management theories as in my book.

        Essentially, the rule is, risk no more than 2% of your beginning bankroll per bet and do not "jiggle" the size of your bet until your bankroll increases (or decreases) by roughly 50 percent. The same bet size should be used against 2-bet parlays paying no less than 13-to-5 and 3-bet parlays paying no less than 6-to-one. (Using parlays does not mean you can win more money, but you can limit your downside risk when you have multple bets. It is this factor that causes bookmakers to restrict their advertising the availability of parlays. It is not in the bookmaker's best interest for you to limit your downside risk.)

        Landmark, I believe the mathematical aspects alone are enough to "prove" the advantages of flat betting over using some sort of progressive betting scheme, but your shockingly insightful examination of the subjective factors involved (greed, etc.) are also enough - in themselves - to prove the value of avoiding progressive betting systems.

        Comment


        • #5
          I think we can all agree that, if the choice is between deciding how much to bet by emotion or by math, math, no matter the system, is preferable. Whether to bet a flat amount, a flat percentage of BR (my preference), or a varying (by supposed advantage) percentage of BR doesn't matter nearly so much as the initial decision to be consistent in your approach.

          Comment


          • #6
            Well one point

            Pretty hard to use Kelly type stuff when you are starting out on betting on a sport, as you have no idea if you even have an edge, let alone how big it is.....

            Comment


            • #7
              Good point, AV. So the flat bet system certainly is the one that handicappers should start with.

              Later (years later) it MIGHT be that one can show an ability to put an expected win percentage to different kind of bets. I would guess that this is a very, very small percentage of bettors. For this case there are other possibilities.

              Comment


              • #8
                Even given the premise that one can, indeed, know when he has a 55% probability and a 60% probability, it would still be wrong to risk twice as much on the 60% risks as on the 55% risks. Doing so would mean one of two things: Either the bettor would be risking too much on the 60%ers or not enough on the 55%ers.

                If a bettor is risking, say, $110 on his 55% bets it would, indeed, be okay to risk more on his 60% bets - but only perhaps $120, $130 or so. Certainly nowhere near $220. If the size of your bet will break you with a 55% winning expectation, I can guarantee it will sooner or later break you with a 60% winning expectation.

                Remember, having a 60% winning expectation is not at all the same as actually GETTING a 60% result. Risking 4% of your bankroll with a 60% expectation will sooner or later result in your demise. The 2% you'd supposedly be risking on your 55% bets would not keep you alive.

                Comment


                • #9
                  There's a discussion about binomial distribution on our website at http://www.professionalgambler.com/binomial.html that may help define my point.

                  Comment


                  • #10
                    JR, both the 2% and 4% are CONSERVATIVE NUMBERS. When I run simulations with random numbers(using my strategy), neither approach runs close to tapping out, but in general, it seems that the 2% on 55% games, tends to fall down further than the 4% on 60% games. Your point about expected hit rate, vs actual hit rate is valid. In fact I have some angles that hit such a high rate, I bring down the expected hit rate on them, just because I cannot fathom how they hit so high. I Agree to error in the conservative side, but I do so in my assesment of the games, not in the bet strategy itself.

                    Comment


                    • #11
                      I did some simple stat calculations a few years ago to answer such a question. The assumption was that the player knew his win percentage on each play. Let me stress that this is really unusual to accomplish, and should be backed up with years of actuals for confidence.

                      I solved for the flat bet $ of bankroll that should be risked on each play. The constraint I used was to ensure with some probability (typically 95%)that one's bankroll would not dip below 50% of its original value. In orher words, I didn't want to just barely keep from going broke. My assumption was that if my bankroll dipped below 50% of its starting value, my emotion would take over anyway.

                      The conclusion for me was to risk 2% on plays with an expected win ration of 55%, and 5% on plays at 60%. All these wagers were assumed to be straight bets laying 11:10.

                      Then I applied my own further conservatism, and concluded that 2-4 % was the best range for plays expected to win at 55-60% clip (because of my skepticism about being confident in achieving the 60%).

                      So I think there is in fact a rationale to put more money on plays that will achieve a higher win rate, even up to the point of double the smaller ones.

                      Comment


                      • #12
                        Landmark:
                        Does your book move on air? If not, would you kick a player out for beating you to the steam? By the way, welcome to the forum.

                        Comment


                        • #13
                          I think recommending a flat bet approach makes sense when one is uncertain if they are getting an edge or not on the odds (i.e. a beginner) but if you have a proven track record like JR I don't understand why it isn't constantly changing relative to your current bankroll. Say I start with $1000 and with a base bet of $20 I build it up to $1200. So my base bet is still $20 (I haven't won 50% yet, which is the point where JR says I should raise my base bet). All of a sudden I get an inheritance of a billion dollars but one of the conditions is that I have to stop betting on sports. So I give my $1200 to my twin brother and recommend he uses JR's service. So here comes a brand new client with a bankroll of $1200 who according to the guidelines should make a base bet of $24. But if I never got the inheritance that same money would stay in the old account and be betting only $20. That just makes no sense at all.

                          JR, you say you change your base when your balance has risen or lowered 50% ... how did you arrive at that figure? Mathematically? It just makes no sense to me that starting with $1000 your ideal base bet is still $20 at $1499 and all of a sudden it takes a quantum leap to $30 when you win the extra dollar.

                          BTW, I totally identify with the Stress Factor idea ("Don't let your emotions dictate your bet") but it's just as easy to say "Bet 2% of my up-to-date bankroll balance no matter what" as it is to say "Bet the same amount no matter what".

                          Comment


                          • #14
                            Bucky, we always welcome players who give us real action. We have in the past let go of some players who were nothing but steam, because they became more of a nusiance than anything else. If anyone is mixing steam with other plays, even if they beat us to a number here and there, they normally won't get a problem from us. Steam action is not something we are actively seeking, we have plenty of action to deal with as it is.

                            Comment


                            • #15
                              PerpetualCzech

                              Welcome to the forum. I don't know if you're a beginner or not but you're asking lots of good questions and making very intelligent observations.

                              'mute

                              Comment

                              Working...
                              X