Do fades exist?

[Ganchrow]

Kodiak is another great one. His dubious performance is well documented (not by him). He is one of the forum touts that tries to compensate for his losses by chasing and does nothing but dig himself a bigger hole long-term. His martingale attempts to get back to the + side are hilarious. Even when he wins the 10x normal bet to get back on the + side he crows like he has conquered the world.

All you're doing is explaining why you believe Kodiak to be a poor gambler. The same general description might well be applied to a craps player who behaves in such a manner.

This doesn't imply that fading either one (as a player) is likely to be profitable.

[Arilou]

TLD's post is excellent. The reason for assuming that on average a picker is under 50% is that on average we can assume that the average DOLLAR bet on a pick is a 50% pick pre-vig. But that pool includes syndicate money and otherwise very sharp money that doesn't talk, and that money is large and overall it is profitable so it is well over 50%. That leaves the remaining dollars, which overall are under 50%. Every time a sharp moves a line, the side he bet gets picked less (more expensive) and the side he didn't bet gets picked more (less expensive). All pickers are, to some extent, price sensitive. Therefore my instinct is that most posters would be under 50%. The way to avoid this would be if the true public was well under 50% and that could balance the scales.

The efficient market hypothesis is, in my practical experience, completely wrong in the gambling world.

The thing about a fade, like all other methods of picking, is that it's very hard to get something that beats a -105 vig (let alone a -110 vig) on its own. However, that doesn't mean you can't use them for fun and/or profit. You can use a combination of line shopping and timing, reduced juice, other picking methods and such to improve your game. If nothing else, simply avoiding betting on any picks of someone with a 49.5% prediction record would be a reasonably good idea and improves your chances to be a winner.

[Ganchrow]

What it seems to me you're saying, TLD, is that insofar as a touts (and I use the term "tout" to refer to both a professional tout as well as a forum tout) makes picks that are in line with some set of preexisting conditions that tend to presage losers, he would make a good fade.

I suppose that's almost self-evidently true, and there's not really much to say in response other than it probably represents a "weak form" of a fade. Indeed you acknowledge this when you write, "I would think it would be a more efficient use of your time and effort to identify those factors and use them in your handicapping directly, rather than to infer that a given tout seems to be influenced wrongly by them to some small degree in his picks and so should be faded."

Essentially you're saying that for exogenous reasons it's in the best interest of certain touts to make picks that correspond to the more heavily juiced side of bets. I guess the problem is that this simply represents a rehash of "known influences" that could just as easily be gleaned from elsewhere. To me, saying that touts who tend to pick MLB favorites and public money-heavy teams, while tending not to pick home dogs generally tend to make good fades is rather unsatisfying.

I'll acknowledge that I've been tweaking my target from post to post within this thread, and I suppose that's testimony to some of the good responses here, but at this point what I'm looking for is evidence or rationale for why certain touts might provide new contrarian information that's not readily available elsewhere. In other words are certain touts simply that much better able to process information and turn them into losing picks? Possibly. But again, that's a far cry from the claim that fading these "public touts" is likely to be profitable.

[Ganchrow]

TLD's post is excellent. The reason for assuming that on average a picker is under 50% is that on average we can assume that the average DOLLAR bet on a pick is a 50% pick pre-vig.

Based on my experience with markets in general, I don't think we can necessarily assume that this is correct. I think that we could even say that the average dollar bet is better than 50/50. Reason for this could be that the market is a better judge of probabilities than any one bookmaker. As such until the market pushes prices into line there's a surplus of money coming in on the +EV side. Is there a reason why public money should be biased toward opening lines? All we know from efficient markets is that the public shouldn't be able to make a profit -- not that they shouldn't be able to out pick 50/50 (throwing out efficient markets we could say that the public might even be better than breakeven).

What's the reality? I'm not really quite sure, but I think it's safe that the more the market moves away from an order-driven sportsbook model towards a price-driven exchange model, we'll see the public opinionated pick percentage (that is ex pure liquidity providers) start exceeding 50%, if it isn't doing so already. Any way you look at it, however, I don't think any of us can argue that we actually expect fades to bne profitable.

This is exactly how it works on financial exchanges as well (and from my experience the gambling markets are at their core just like the financial markets or indeed any other market). Market makers, be they bookmakers or exchange traders, tend to make on average a little bit less than half the spread per trade (the equivalent in sports betting being the theoretical hold). The reason is that because by putting oneself out there, quoting an even partially static price on both sides, a market maker is giving up a little bit of optionality. Of course they're being compensated for it by way of the spread they're charging, but there still is that little extra piece of information that can be utilized by price-takers. (To the extent this is not the case I'd suggest it's due to structural inefficiencies resulting from books not opening up their markets to the public. I believe the impact of this will continue to diminish with the continued rise of the sports exchange.)

So what you then have existing along a continuum are a few market "sharps" picking at better than breakeven, more market "semi-squares" picking at slightly better than 50/50 on average (but still unprofitably), many market "squares" picking at right about 50/50 on average (or maybe even a bit less due to the impact of public money), and a few market makers getting picked off at worse than 50/50. But this is exactly what the market makers are being paid for. The idea that there's some invisible force which tends to dupe the entirety of the non-sharp public into making picks at a rate lower than 50/50 just doesn't pass the smell test. Maybe it's true, but it would just be so incredibly different from what exists in other markets. Sure, some recreational-style books tend to get a certain type of non-professional action and hence skew their lines, but there's no reason to apply that to the market as a whole. The market as a whole is pretty darn smart.

I know I'm running the risk of getting a little too carried away here. I'm really just trying to demonstrate that there are many reasons to doubt the existence of profitable fades. Certainly there are some touts who systematically pick worse than expected by chance alone. Nevertheless, if the claim is that fades exists in the personae of those touts who make "public plays", that at best implies only the existence of a "weak form" of a fade.

I think the important question is whether or not knowledge of these fadeable tout's picks provide new information or whether they're simply illuminating already "known" biases. Furthermore, do these touts provide profitable opportunities for those shorting their picks. It seems like we most all agree that the answer the to this latter question is in general, no. So what this seems to reduce to is:

1. There exists some subset of touts who tend to make "public plays".
2. Following these touts can give an indication of where this "public money" is going.
3. Public money tends to under perform a coin flip.
4. Insofar as 1-3 are likely to be true for a particular tout/bet pair then a fade is expected to be profitable.