JR Miller theory results for me.

[buckeye]

I apologize for those who don't give a chit, but for the hell of it I decided to analyze my $line bets ( mostly golf ) for the month to see if I would have done better betting them all the same, instead of varying units like I do. (a.k.a. JR Miller's flat betting method ) I didn't finish the ROI analysis, which is more important for $lines since I tend to bet dogs at lower unit values and more favs at higher units ( I know this is the opposite of successful capping at many sports, like FB, but on golf I find it works better for me ) so that would lower the winning %'s at lower units, naturally.

Results so far: ROI
.5 unit plays 40-46-3 .465 -.7units -?%
1 unit plays 23-25-3 .479 +1.3units +2.8%
2 unit plays 12-7-2 .632 +8.2units +17.6%
3 unit plays 2-0-0 1.000 +6units +70.2%
5 unit plays 1-0-0 1.000 +5units +100.0%
I eliminated ties, scalps, partial scalps, "line discrepencies", etc. to just look at pure capped plays.

On this small sample it looks like my weighting was correct as my winning % and ROI % is steadily > at higher units. It also looks like I should have skipped the .5 and 1 unit plays as I wasted my time on them. I will keep tracking to see if this is consistent, it may be. I am sure that any given month won't have an exact linear increase like this one happened to have, and I know that every month isn't +, but it is worth looking at.

Have others analyzed their stuff like this?
What I am wondering is if you consider it a 65%, or 60% play, and weight it more, due to "past" trends of that %, or because the plays fall into a category(?) that you know has demonstrated 65% hits for YOU in the past? My weightings are not objective at all, I just weight them by my gut feeling for their strength relative to other plays. But I am no pro, so I'd be curious what others do!

[SLAM DUNK]

Buckeye, your results, are consistent with what I would expect from most cappers. It is only logical(to me-others do not seem to agree), that bets you like stronger should perform better than the ones you do not like that much.

Your sample size is fairly significant on your 1/2 and 1 unit plays. Coincidentally, my weak plays usually perform with the same mediocrity. However, I would not advise abandoning them and this is why. I can only speak for myself, but the only way I LEARN, is by losing. Now not all losing bets are BAD bets, but when I lose I look at the nature of the loss, reanalyze all the variables, and try to learn from the incident if there is something to be learned. All of this helps sharpen my handicapping skills. So even if I drop a few bucks on my weak plays, I will more than make up for it by better decision making on my stronger plays.

In regards to your question about 60% plays...You can go 2 routes here. One would be to continue rating your games as you do, 1/2 unit, 1 unit, 2 unit.... Once the sample gets large enough, you will have a fairly good indication of what you are capable of hitting on your 2 unit plays and 3 unit plays etc. If you find that when you have 100 2 unit plays and 100 3 unit plays in and you have a record of 60-40 on the 2 unit plays and 64-36 on the 3 unit plays, you know that you are capable of hitting 60% on these, but if your record is 54-46, and 57-43, obviously you are not. Remember thought that since you are playing money line, you have to reach a 60% equivalent which would be +16 units, per unit wagered per each 100 games(60-40 on a 11/10 is +16 units). The money line aspect actually makes the process a lot more complex. A more objective approach is to gather up specific guidelines of what you would consider a strong play. For instance in
Basketball it might go like this,
1) the line is 2.5+ points above fair value
2) no key injuries for your team
3) No major emotional reasons why your team might underperform or the opponent might over-perform.

I prefer the guideline approach, because it is a more accurate read. One week you might be on a roll and consider a golf play a 2 unit play, where as 3 months later, when you are on a slump, you may rate an identical
type play a 1 unit play or even a 3 unit play, depending on your personality.



[This message has been edited by SLAM DUNK (edited 07-26-2000).]

[buckeye]

SLAM,
You make some valid points. In college hoops I only use 1 or 2 unit bets and do have more objective guidelines for what qualifies for each. I know that some golf cappers set their own $line and vary their units based on how off the books lines are. I don't use objective measures like that, mostly observations on "how" the players are playing and many other techniques that aren't easily quantifiable, but they WORK. So I am susceptible to weighting a similar play differently from one week to the next - that's true. I try to be consistent and not vary it on how hot or cold I am at a given moment, but it often does affect my plays anyway.

In golf, I have found that if I get too objective, or try to apply hard and fast numbers and formulas, my capping suffers from overanalysis and my results really decline. I can find an angle to HATE every play that my initial capping rated highly - by overanalyzing. I don't advocate this approach for pros, but I only do it for fun anyway and it has worked for me over the past 4 years ( I've had weeks of slumps but never a - year ).

What are your thoughts on the following phenomenon?
I am sure that this will sound ridiculous to many, but I have found that on many days if my picks start out RED HOT ( like I win the first 4-5 plays ) they tend to reverse later. This only happens for sports that I don't cap well, or have a mediocre long term record in, like football. The converse tends to happen if I start off COLD. It's as if the law of averages is always trying to bring the day back to 50% in the end. I am sure this sounds falacious, but I have been reasonably successful by laying off later plays when I start off real hot, and parlaying late plays when I start ice cold, and it has usually worked. I just find that I never have any MONSTER 10-0 days or HORRID 0-10 days, I am much more likely to end up 5-5, 6-4, or 4-6, no matter how I started off. This doesn't happen in golf and other sports where I feel I cap well, as I do have MONSTER days/weeks in them on occasion. I don't always "second guess" my capping, but it seems to be an interesting phenomenon. I notice it a lot in the NFL, if the early NFL games are kind to me the late games tend to bring me back to even, and verce-visa. I first noticed it in playing "pools" where you have to pick all of the games, but it seems to happen for me even when I am selective. Any thoughts about this?

[SLAM DUNK]

Back to Binomial distribution. Make 10 50% plays and your chances of winning

0) .1%
1) 1%
2) 4.4%
3) 11.7%
4) 20.5%
5) 24.6%
6) 20.5%
7) 11.7%
8) 4.4%
9) 1%
10) .1%

note a 65.6% chance that you will win 4 to 6 games and a 90% chance you will win 3 to 7
games. Does you 50% game have an 80% chance of winning because you are 0-8 already for the day, absolutely not. Does your 50 game
have a 20% chance of winning because your 6-0 for the day, once again absolutely not. A 50% game, is a 50% game no matter what happens in the prior games, a 60% game is a 60% game no matter what happens in the prior games, but by looking at the chart above, the phenomenon of your days usually fitting in the 4 to 6 of 10 range makes a lot of sense. Whatever success you have had with your method is either short term or perceived. It is like going to the race track and watching the first 8 favorites lose, does that mean that in race 9 the favorite will win, of course not. Streaks come in all sizes, shapes and forms and trying to handicap a winning or losing streak will leave you very frustrated, because it CANNOT BE DONE.

[PerpetualCzech]

Exactly right, Slam Dunk. It's the classic gambler's fallacy, that a streak of independent events that all go one way will somehow have an effect on the outcome of the next one.

buckeye, for me to believe you, you somehow have to convince me that how the Yankees are playing in a Sunday night game is influenced somehow by your betting patterns on the day games. You have to show some kind of connection between the two.

[buckeye]

I understand the fallacy of a single game being dependent on a streak. I agree that a coin flip is not dependent on the prior ones, and the dice have no memory, etc. But the streaky nature of some trends and capping methods may have some intelligence behind it. If all dogs cover one week I would be willing to play all favs the next in "reaction" to what I percieve will be an over-reaction to that short term trend by books. I know one successful gambler who bets against teams that just blew someone out and with those that just got blown out because the lines will overadjust to players "overvaluing" the last game played. It isn't the only factor he uses, but it has some merit. I know another gambler who bets against trends, instead of with them, because he feels that linesmakers overadjust to counteract the ones that are well known and have been hot. These examples don't say anything about "predicting late games based on early games" but they could contribute to streaks.

I don't claim that 8 losses in a row makes the next game X% more likely to W, I just think the up and down nature of "poor" capping tends to make things even out and short term winning is more often followed by short term losing, or mediocre results, than continued winning. This may sound illogical, and fallacious, but it has held true in my poor capping sports. It could well be just luck, but it may be my capping techniques have something to do with it.

GL2U

[SLAM DUNK]

I think you grasp the point, so I will not go any farther. I do want to address a couple of the points you brought up. "You know a successful gambler bets against teams who just blew out an opponent because they will be overbet". Be careful with your interpretation of data. I am of the theory, that the team will not be overbet, but might come up a little flat. So your friend may actually be on to a good angle but for the WRONG reason. In general, lines DO NOT react from one game. If the Lakers are going to be home to Boston next week and the line should be Lakers by 11 and 2 nights before the Lakers are going into Portland, and happen to win by 30, I feel that the line will still be 11. Now admittedly the "squares" might be chomping on the bit to bet the Lakers, but many sharp players are seeing letdown. You know how the books are, they could care less who the square is betting, they just do not want to get popped by the sharp ones. Also tend to disagree about the trend theory. Most lines are fairly consistent with my power ratings(which do not incorporate trends), so I do not think that an overwhelming amount of weight is put into trends at all. Also it is the "sharps" that ultimately make the line anyhow, as when the oddsmaker throws up a bad overnight line, it gets hammerred very quickly. A further point, there is no reason to incorporate a trend into the line unless the trend is well know. Let's say you are an oddsmaker and you see that team x is 3-12 at home laying over 4 points, but the public is not reacting to that trend. The Milwaukee Bucks had a similar situation early this season. There would be no reason for you to bring a line down from 8 to say 7, because the public will just bet it back up to 8 again. Since the objective of the linesmaker is to open a game at the number it closes at,
per Reality, there is no reason for a linesmaker to incorporate a trend UNTIL the public starts reacting TO THE TREND.

[PerpetualCzech]

OK buckeye, I see what you're saying now but I'm afraid I don't have enough experience to be able to offer an opinion. Slam Dunk's comments seem to make a lot of sense.

[buckeye]

SLAM,
I have asked the "anti-blowout" bettor the same questions. Doesn't the blowout have more motivational impact than line impact, since the blown out often want some redemption and may prepare harder for the next game, whereas the easy winners may become lazy and complacent. He told me that he feels that may have some bearing, but that the "square" publics overreaction to the last game played by a team is more of a factor. This is for NFL-only, BTW, he doesn't bet NBA and I think he is basing it on the action that NFL betting gets. You may be correct, that it works for a different reason than he thinks, but he has been successful with it.

GL

[Railbird]

Buckeye,

looks like your .5 and 1 unit plays are made for the action,try being an aligator and wait for your 3 to 5 unit plays.

I liked the team with Cornelius Green and Neal Colzie.