"I bet short price favs"

Fair enough actually. I was assuming you were just betting on player priced $1.04 or something. The tennis package has plenty of bets on big favourities, but rarely below $1.10

"I agree, my point is then when we are starting out, we usually start with a small bankroll, by necessity. As we progress in confidence, we start to have bigger stakes and it only takes a small loss on bigger stakes to wipe out years of wins on smaller stakes. For most people starting out on small bankrolls this is a reality."

If you had resized you unit size reguarly, this wouldn't happen (unless you were overbetting). Most punters do overbet though, it's really only the serious ones (looking to profit) who take risk of ruin seriously. If someone is betting for fun, they probally expect to lose so risk isn't a problem

"I know that you suscribe to some selection providers. They do not win every week."

Yes I do. No they don't (unfortunately). By 5-8% profit each year, I assume you mean profit on turnover (return on investment). If you mean profit on turnover, then I can understand why you cut back if he runs hot. You expect that there will be a 'correction', and you want to minimize your losses. It doesn't work like that though. All bets are independent of each other, and are random. Since every bet has a perceived edge, you actually are missing out on money as there will be times there isn't a correction. Your way isn't optimal, but if you are happy locking in your profits then that's fine.

"If you look at the graphs of the MLB product offered by this web site you will see a similiar thing. For example in their 2004 season they made 46 units. I don't know how many bets they had but in the first 2 months they made 39 of those units and then dropped off in June and July before doing well in August."

Sure. However, You CAN'T predict how they will do going forward though. You CAN'T predict when the capper will make his profits. Trying to play trends will cost you money longterm.

"I believe bankroll management should incorporate adjustments for long term trends and it is my experience that all selection methods go through peaks and dips in a season"

Sure there are peaks and dips, but you can't predict when they will happen. That is my point. If you have a 60% capper and he goes 75-25, that doesn't mean he's 'due' for a bad run to knock him back to 60%. The most likely result is that he will hit 60% going forward, so by lowering your bet size or not betting at all, you are missing out on profits.

"However, if your service can provide over 60% winners every week, please let us know who they are."

I can't comment on that. If I had access to 60% winners, why would I share them