The Drake equation (named for Frank Drake of UCSC) is just a way to organize our thoughts about the likelihood of hearing radio signals from extraterrestrial civilizations (or just thinking about life in our Galaxy). It calulates this number as the product of the rate of production of ETs and the lifetime of each ET civilization (actually the period during which they send signals). You can write:
N_ET (number of them) = R_ET * L_ET. Since we have no clue what L_ET might be, we concentrate on R_ET and write the answer in terms of L_ET.
To produce ETs, we need a star with habitable planets around it (we think). Then a planet has to produce life, the life has to become complex, and finally has to have the will and wherewithal to send out radio signals we can get and recognize as such. So R_ET = R_star * P_planet * P_habitable * P_life * P_complex * P_radio emitting, where each P is the fractional probability that, given everything up to that point, the next step can be completed. R_star is an observed quantity in our Galaxy (admittedly not very precise); let's say that is about 2 per year. P_planet is observationally informed, in that our very limited search techniques already find planets around several percent of the stars tested. We can't be wrong too far if we set P_planet=0.5. The question of habitability used to involve wondering how many planets similar to Earth there are; now we are willing to admit that there might be habitable zones inside rocky planets or icy moons (if tidally stressed like Europa) - at least for simple life. Furthermore, we know that life arose on Earth almost as soon as it could, so the odds seem favorable based on that one example. Obviously we are already guessing, but it is not unreasonble to set P_habitable * P_life as high as 0.1. Once the Earth had simple life, it was content to stay with that over most of its existence, only during the last 20% of time has there been any complex evolution (and of course humans appeared only at the last figurative second). So even optimistically we probably should not set P_complex higher than perhaps 0.01 (but this is pure guesswork and a somewhat personal choice). The likelihood that complex, even intelligent, life will choose to (or end up) sending radio signals recognizable as artificial we have no clue about - other than that we are doing it. Let's be wildly optimistic and say that P_radio emitting is 0.1. Then we find at last that N_ET is 0.0001 * L_ET (in years). This means that any radio-emitting civilizations need to do it for at least 100,000 years on average in order that there be a few of them doing it right now in the Galaxy. The Galaxy is a big place, so in order that the nearest such civilization be within 1000 light years of us pushes the required average lifetime into millions of years. And of course this is our optimistic case; it is not hard to choose the numbers above so that it is likely we are the only radio-emitting civilization currently in the Galaxy. On the other hand, the same numbers suggest that life is abundant thoughout the Galaxy. Anyone who does not like these conclusions can take comfort in the fact that it should also be obvious we don't really know what we are talking about. Yet.