--- In vpFREE@yahoogroups.com, "rob.singer1111" <rob.singer1111@...> wrote:
>
> I'll input on this Frank since I agree with Mike. I'll start off by explaining why having others "watch" trainees is inconclusive to the point of being irrelevant.
> The irrelevancy of having others "watch for accuracy and errors" is simply adding another imperfect parameter to the equation. As such, multiple assumptions and conclusions are arrived at that really have no basis in science.
This whole point you are making is GIGO since what matters in evaluating the efficacy of a video poker player is not error rate per se, but the effect of what errors the player DOES make on EV. For instance: a player who makes an incorrect hold in FPDW because he failed to perceive a penalty card may cost himself 0.02 bets by doing so. Compare this with something basic, like holding KK instead of KQJs. Also, the frequency with which a situation comes up matters just as much. Again defaulting to FPDW, an error like holding all five cards from 22266 costs (if memory serves) 0.1 bets, but the frequency of dealt 5OAK with three deuces is so small, the impact of that error on overall EV is minuscule.
All that said, the technique of "qualifying" potential prog team members by watching them predates computer video poker practice programs, so it may be more of an artifact than anything else. Certainly, you can tell whether a player is skilled or not simply by watching him play for a few minutes. Quantifying his skill (i.e., how close to optimal he plays) is considerably more difficult, and consequently inadequate.
