I tend to hear mental alarm-bells ringing when I read of staking plans that involve increasing the stakes (even gently at first) during a losing run. Many people say that if you can't make a profit at level stakes, you can't make a profit at all. I'm not quite as pessimistic as this, but I believe it's certainly true for systems with low strike-rates. The point is that with any system you _will_ eventually have a longer losing run than the system has previously had, and working out the probabilities of such long losing runs realistically is essential before you start. (This subject is well covered in Mason Malmuth's "Gambling Theory And Other Topics" and Van K Tharp's "Trade Your Way To Financial Freedom".) I believe that Isiris have had a strike-rate in excess of 40% (and at pretty good average odds) for each of the last 12 or more years, and that this has been independently monitored by the Racing Information Database. I'm not sure if backing all Isiris's "account bet" selections at £10 will produce enough to cover the cost of the service, though. At the moment I think the subscription's £1,950 for the rest of 2004 and all of 2005.