Hi All
Just a post on probability.
If a system is closed and fully understood - such as a lottery - then the probability of an event within that system can be given with certainty. So Binadub is right that if a lottery of 6 from 59 balls (odds of about 17 million to 1 of winning) was won by the same person even twice in a row you have evidence that the lottery is rigged or there is something special about that person. In effect this means you may not fully understand the system and/or it is not closed. So you need to investigate. Of course, repeat enough lotteries enough times and the same person will win any number of times in a row - however unlikely.
Where systems are not closed and are not fully understood then the best we can do is give our estimate of a probability of a specific event given our current information set.
In the case of biological processes for some things we know the system reasonably well enough to be able to estimate with accuracy an event happening provided the environment is closed. For example, biologists can estimate the probability of a new protein arising in an organism over a specific time period that has a useful function if they know the number of genes required to mutate to generate that protein coupled with the rate of mutation. They can then conduct an experiment where they can control the environment to see if they are right.
For biological processes where the system is only partially understood or the system is open to interference from a range of exogenous factors it becomes far more difficult to estimate the probability of an event with any degree of accuracy. What biologists can try and do to estimate the probability of an event in this instance is make observations. This is very difficult though in the case of chemicals coming together to start life. To begin we still do not understand the mixture of chemicals, how it gets sparked and what the first step looks like. Biologists speculate and from those speculations conduct experiments to see what happens. From those experiments they may ascribe a probability of a specific event happening based on observations under a closed system. But that event may not be how life started and the conditions when the event happened may be markedly different then the closed system in the experiment.
So the upshot of all this is that we should not confuse and use in an argument the known probability of an event in a fully understood and closed system such as a lottery with the estimated probability of an even in a system we do not fully understand and is open. In any case, as has been pointed out before it is not just the probability that matters in isolation but the number of opportunities the event has to happen.
As a final aside I always find arguments about probability from creationists a bit odd. If the likelihood of life originating by 'chance' is really low non-creationists only need that unlikely instance to happen once. So do creationists, just a step back.