Announcement

Collapse
No announcement yet.

Reverse Middle Scalps: Putting the House Percentage in Your Favor.

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

  • #16
    Yes, there are scalping opportunities in soccer. Hard to learn? Nope. Easy enough to swap lessons if you want.

    Comment


    • #17
      Ronbets...

      I was not born to be anyone's "yes man", but for the period of time that I've been in this forum, Reno's words and tales of his experiences resonate well with me. His insights frequently remind me of my own similar experiences, which is why I can relate to much of what he has to say.

      I didn't begin scalping lines until late last year, although I had been thinking about doing this for a number of years. My biggest regret is that I didn't drop everything sooner and move to Vegas before everything changed.

      So, maybe subconsciously, I admire guys like you and Reno for having the guts and courage to pursue your dreams, despite the personal conflicts that one encounters in that pursuit (e.g. - explaining to family & friends what you do for a living, etc.).

      Now that I've chosen this path for myself, I come into this forum to share my thoughts and gain insights from the experiences of others.


      P.S. - My other mentor is Homer Simpson

      [This message has been edited by Bette Midler (edited 06-07-2000).]

      Comment


      • #18
        BTW Ron....


        I do disagree with Reno on a few things:

        1) I almost NEVER put myself in a position to be heavy on one side of a game, and

        2) Ben & Jerrys is still my favorite ice cream. (Sorry Reno )

        Comment


        • #19
          Bette Midler,

          Your well on your way.

          As far as personal conflicts?? This game we play is not conducive to married life. You have to go to Rosewell,New Mexico to find the yielding woman.

          Comment


          • #20
            Boomer:

            Is the percentage of all games that have landed on (to use your example) 10 runs really what’s relevant here? What if 7% of all games have landed on 10, but some much larger percentage, say 15%, of games with a run line of (approximately) 10 have landed on 10? Presumably comparatively few games projected to be unusually high scoring or unusually low scoring land exactly on 10, but are those games relevant? The way you’re analyzing things seems to me to underestimate your risk of being middled, as you are including past performances of dubious similarity.

            Comment


            • #21
              zippy,

              so how often is a total exactly correct??????

              Comment


              • #22
                I don’t know, AussieVamp2. I’m just speculating that it seems mathematically suspect to count all past games equally in estimating the chances that a specific upcoming game will land right on a certain number. Boomer’s case was a game where the projected total runs was in the 9.5-10 range. He reasoned that since 7% of baseball games played in recent years have ended up at exactly 10 runs, he needed to make sure that his expected payoff was high enough to offset a 7% chance of his being middled on this game. I’m just saying that if one were to look instead only at baseball games played in recent years where the projected total runs was in that same 9.5-10 range, very likely the number that ended up landing exactly on 10 would be quite a bit higher than that 7%. Just an educated guess. I haven’t done all the number crunching that would be necessary to test such a hypothesis. But I’m curious to hear from those who have, or from those who see a logical error I’m making that would render such an investigation pointless.

                Comment


                • #23
                  It is the chance that a game with a 10 total will actually end on 10 that matters, not the chance that any given game will end on ten.
                  Yet also, the sample sizes needn't be as large as people might think. This isn't football, where different totals, because of the nature of the scoring, result disproportionately.
                  In baseball totals, it's important to do a chart of "exact scores by total size." That is, how often exactly 12 runs are scored when the total is 12. After that you need to find the AVERAGE likelihood of this for all the different values. This is why you don't need such a large sample (though the larger the better, certainly). Your results from games with totals of 10 are relevant, that is they have predictive value, for games with totals of 9 or 11. Use the average likelihood, not the #-specific actual results. For example, and just pulling #s out of the air here, if 9's hit at 8% and 11's hit at 8% but 10's hit at 11%, unless you have a reason to believe 10's are a specifically more likely total (which they aren't in baseball) you should expect your average likelihood to be 9%. The difference between your 9 and 11-total results relative to your 10-total result is just random distribution.
                  What you'll find is that the smaller the total, the slightly greater the chance of landing right on it, so you may want to actually use a little calculus here, though you can rough it out and be close enough. This is to be expected, a 1-run variance from a Qualcomm 8 being a greater number than the same variance off a Coors field 14. Yet, after accounting for the scoring increase in the past decade, there's no reason to ignore data from any recent era.

                  I did some research into all this once, but unfortunately I can't remember the exact numbers. What I do remember is deciding that my EV is still much higher playing poker than sitting in front of the computer all day scalping. But if I WAS going to scalp, I'd learn it. In my talks with bettors, for example, I've yet to meet one who knows how to calculate the money line value of a 1/2 run, which is all we're really talking about. Ask most bettors which is better, if you're betting the over: 10,+100 or 9.5,-115?
                  It's sad, but good for the BM's I suppose, that so few can answer it. Or just as bad, ask most bettors how to compare a run line to a money line. They can't do it, yet they make decisions between run line betting and money line betting all the time. On what? On air.
                  Good luck, all.

                  Comment


                  • #24
                    Boomer, Skybook had the total of Phil/TB 10u65 the last few minutes. Every 5cents helps.

                    Comment


                    • #25
                      Thank you for an informative post, pay2play. I was just using common sense; you’ve clearly thought it through and done the necessary number crunching.

                      I’m more of a football bettor, so I’ve mostly thought about this issue in terms of football. As important as people realize the number 3 to be in football, I suspect many bettors still calculate its precise importance incorrectly. Typically they note that roughly 14% [that might not be the right figure—I don’t feel like looking it up to check my memory] of games end up decided by a 3 point margin, and so for a game with a 3 point spread they conclude that there is about a 14% chance of a push.

                      This seems to me to be wrong on two counts. First, you’re really only concerned with the case where the favorite wins by 3 points, not all cases where the game ends at a 3 point margin. In that sense, people overrate the chances of the game falling right on the number. But second, as I brought up in this thread, presumably one should be looking only at past games with approximately a 3 point spread, not all past games. In that sense, people underestimate the chances of the game falling right on the number.

                      This type of calculation is relevant for non-straight bets as well. If you’re considering betting a teaser with a team favored by 9 or 10 points or so, and you’re trying to decide what size teaser to get based on the chances of the game falling exactly on 3 or 4 or some specific number, then you’ll want to look at a different subset of past games (namely those with about a 10 point spread).

                      Comment


                      • #26
                        What's the difference between scalping and middling. Damn, I always thought they were exactly the same thing...

                        There's a soccer prop that I found which is +100 at one book, +175 at the other. Is that a scalp or a middle? How are the two different?

                        Comment


                        • #27
                          Zippy,

                          Bullseye! Your right on. A fair calculation of the probability of a specific game landing on 10 would be the total landings divided by the # of games projected by the oddsmaker between 9.5 and 10. We are measuring the linemakers accuracy, not the total meaningless average of all the rest of the irrelevant games.

                          I have no stats for a projected # landing. An educated guess woud be 12% for dead even #s and approximately 17% for for live odd #s.

                          Comment


                          • #28
                            Zippy,
                            Logical Approach's website has an article with results from a study on football games that addresses your exact question. I don't have the URL, but it's probably just their name .com
                            Wait, I do have it: www.thelogicalapproach.com

                            Anyway, after thinking about it, I've decided to post a formula for comparing value between total lines.
                            The key variable here is the multiplier, which I found with a bit of research to be 16. But it really wasn't much research, and anyone with a better number, please weigh in. I wouldn't argue too much with anyone telling me the number should be 17.2 or 15.3 or whatever. Furthermore, I don't have a size-of-total factor involved because I seem to remember the variance as being surprisingly small. In any case, Reno, Bette, Boomer, anyone at all, if you have a better formula, please share it.

                            This is a formula for calculating a betting advantage between your line and the BM's.
                            "#diff" is the difference between your line and his, "cost" is the line odds, expressed as a negative (so if it's actually +105, write it as -95).

                            (#diff*16)+((cost+100)/2)

                            For example, if you make a line 11.3, and the BM has it 10,-110, you'd calulate
                            11.3-10=1.3 #diff
                            1.3*16=20.8
                            +
                            (-110+100)/2)
                            =
                            15.8% betting advantage (if you're right, of course).

                            To compare it to another BM's line, say, 9.5, -125, you'd find the advantage to be 16.3%, so you'd want to take it there. Then again, at 9.5,-130, you'd have to pass.

                            I'm not selling this formula so I don't have to guarantee it, I'm just suggesting it to help other players. If you've got a better approach, don't flame just share, because I'd love to hear it.

                            PAY2PLAY

                            Comment


                            • #29
                              Interesting topic, maybe these stats will help--for the first 2 months of the 2000 MLB season, April 3-May 29 (May 30 & 31 stats not done yet), there were 738 games played. The total number of runs scored in those games were:

                              7 runs -- 72 games -- 9.7%
                              8 runs -- 49 games -- 6.6%
                              9 runs -- 56 games -- 7.5%
                              10 runs -- 46 games -- 6.2%
                              11 runs -- 56 games -- 7.5%
                              12 runs -- 36 games -- 4.9%

                              The mean average total of runs for all 738 games was 10.7, the median average was 10.

                              [This message has been edited by chance_of_rain (edited 06-07-2000).]

                              [This message has been edited by chance_of_rain (edited 06-07-2000).]

                              Comment


                              • #30
                                Very nice, p2p. This is why I come to this forum.

                                Would you mind telling us how one would go about estimating the multiplier, and what the logic is behind it?

                                Thanks.

                                Comment

                                Working...
                                X