[MonkeyF0cker]

I think the issue is that the vast majority of beginning bettors just aren't on a path towards principled advantage play.

As such the advice Crazy gave has some very important advantages for these sorts of rec. bettors. To wit:

1. Control your betting -- If a rec. bettor follows this rule he'll keep from chasing and won't bet outrageous amounts.
2. Never bet openers -- As noted this probably doesn't much matter way or the other, except I suppose, to potentially keep a bettor from betting too many games (see below).
3. Always watch the sport you're wagering on -- if the bettor only bets on games he plans to watch he'll bet on fewer games, which given a fixed unit size would imply he'd expect to lose less money.
4. Never lay more than -800 on an individual event -- The greater the extent to which a rec. bettor focuses on a faves, the more inexorable the decline of his bankroll. This is a bad thing insofar as most of the derived utility for a rec. bettor comes from reasonable outcome volatility.

While I can see how these items would create a sort of stop loss for recreational bettors, I still stick by my assertion that it will most likely result in a limit down scenario in the end. That said, I suppose they seem like reasonable bankroll management methodologies for purely recreational bettors.

[MonkeyF0cker]

I propose an SBR Ultimate Handicapper's Contest.

Let's say you have 500 units to bet within a given calendar year. Let's also suppose you may bet as little as a half-unit to 10 units per wager. Let's include all lines offered at the most recently voted Top 10 sportbooks in an SBR poll, or something along those lines, and let's appoint a committee that consists of one SBR owner/boss/admin/mod, one poster in the 5K+ range and one poster in the 1K-4.999K range to work out the fine print. Obviously, SBR is left to determine prizes.

Let's dub it the 2009 SBR Handicapper's Challenge, and invite pros, sharps, squares, rec players, your Wives/girlfriends, their lovers and others to join in.

I would absolutely be interested in such a contest, assuming that the 500 unit size is a rolling bankroll and that the lines are live and not stale (posted daily). This would be a far more comprehensive and accurate handicapping competition than anyone else currently offers.

[Ganchrow]

While I can see how these items would create a sort of stop loss for recreational bettors, I still stick by my assertion that it will most likely result in a limit down scenario in the end. That said, I suppose they seem like reasonable bankroll management methodologies for purely recreational bettors.

There are really two separate utility maximizations, one for each general type of market operator:

Advantage:

Maximize U(EV, risk), where U'_EV > 0, U'_risk < 0, subject to EV ≥ 0

Recreational:

Maximize U(EV, fun, risk), where U'_fun > 0, U'_risk < 0, fun'_risk > 0, given essentially constant EV as % of total wager < 0

[Ganchrow]

I would absolutely be interested in such a contest, assuming that the 500 unit size is a rolling bankroll and that the lines are live and not stale (posted daily). This would be a far more comprehensive and accurate handicapping competition than anyone else currently offers.

Almost paradoxically, to win a large contest such as this an advantage player will often do better in ignoring (or at least highly modifying) many sound principles of bankroll maximization. The idea is that with no downside to losing (other than some fixed cost), there's inherent optionality to the contest implying that U'_volatility will very frequently be > 0.

[MonkeyF0cker]

Heh. I've never found any utility in losing money. I suppose that's why I generally consider the advantage utility function to be the only option.

[MonkeyF0cker]

Almost paradoxically, to win a large contest such as this an advantage player will often do better in ignoring (or at least highly modifying) many sound principles of bankroll maximization. The idea is that with no downside to losing (other than some fixed cost), there's inherent optionality to the contest implying that U'_volatility will very frequently be > 0.

Very true.

[Willie Bee]

Almost paradoxically, to win a large contest such as this an advantage player will often do better in ignoring (or at least highly modifying) many sound principles of bankroll maximization. The idea is that with no downside to losing (other than some fixed cost), there's inherent optionality to the contest implying that U'_volatility will very frequently be > 0.

Is the sample size of one calendar year too short or is the risk of a contest such as this too small? If it was an 'invitational' of some sort, would that make it more valid or are you suggesting a buy-in?

[pavyracer]

Without having inside information regarding the outcome of a game the statistics and math will not allow you to win in betting. If winning in betting sports was easy all the fantasy league winners would be millionaires. Follow your guts as a recreational bettor is the best advice to win long term. Never look at the spread before you pick a team. If it's +EV after you make your pick play the ML if it is -EV pay the juice.

[Ganchrow]

Is the sample size of one calendar year too short or is the risk of a contest such as this too small? If it was an 'invitational' of some sort, would that make it more valid or are you suggesting a buy-in?

The point is that a contest doesn't test general handicapping ability so much as it tests a bettor's ability to succeed at that one particular contest.

Because losing by a tiny margin is no better than losing by a huge margin, optimal play will tend diverge between "real" handicapping and contest handicapping. This will only be further accentuated when the contest has many participants (increasing the importance of needing to get "lucky" in order to win).

Consider this ... against a typical advantage handicapper playing standard spreads, most H2H contests over sufficiently small samples (let's say around 500 or so) will allow a player only picking by looking at the MLs and completely ignoring both the sport and the participants, a considerable advantage in terms of contest win expectancy.

In other words, with many contest structures handicapping is frequently less important than bet price selection.

This is why many contests limit a player to a fixed range of prices at a fixed range of bet sizes.

[MonkeyF0cker]

Is the sample size of one calendar year too short or is the risk of a contest such as this too small? If it was an 'invitational' of some sort, would that make it more valid or are you suggesting a buy-in?

He's basically saying that two variables would essentially cause one to play outside of normal bankroll rules in the contest: field size and risk. The effect could be reduced with a smaller (invitational) field and a sizeable buyin.

[Ganchrow]

This is why many contests limit a player to a fixed range of prices at a fixed range of bet sizes.

Of course even this is a bit problematic.

If the range of acceptable bet sizes were too small, than the the contest would unfairly limit the advantage of proper bet size selection.

OTOH, if the range of acceptable bet sizes were too large, than a player who were losing could up his bet size and thereby his concomitant volatility. This would increase his chances of a "lucky victory", tending to partially negate skill.

And of course limiting a player to a fixed price range would tend to be an unfair detriment in determining the overall skill of a player who would otherwise tend to bet mainly MLs and exotics.

[Ganchrow]

The effect could be reduced with a smaller (invitational) field and a sizeable buyin.

If you think about it, the buy-in is just a constant modifier to contest valuation, and as such the actual contest "vega" would be unaltered.

So while a buy-in might limit the field size, it wouldn't alter optimal strategy beyond that.

Anyway, I think we've clearly ventured beyond the intended scope of this thread ...

[keystonekid]

Keep your number of bets to a minimum.

[MonkeyF0cker]

True, Ganch. Perhaps, a per bet commission (both winners and losers) would be the way to do it...

[Ganchrow]

True, Ganch. Perhaps, a per bet commission (both winners and losers) would be the way to do it...

Yeah exactly, the more the contest came to resemble a swap as opposed to an option, the more it would tend to resemble "true" handicapping (but then the less "use" there would be for such a contest other than possibly vig capture).

[EaglesPhan36]

Dude there is no way you can only bet on games you watch. If that was the case, I'd probably bet twice a month.

[Kingctb27]

tail pags and justin7 in college football and arena for justin. A recipe for success my friends.

[MonkeyF0cker]

Yeah exactly, the more the contest came to resemble a swap as opposed to an option, the more it would tend to resemble "true" handicapping (but then the less "use" there would be for such a contest other than possibly vig capture).

I believe this would be a rare occurance where the recreational utility function would come into play for me.