Tom:
Park factors are computed by looking at the offensive levels put up in those parks (by the home team and the visitors), and comparing them to the offensive levels put up by the same team on the road (and the teams that are hosting them). For RFK, that means that they look at all the runs scored in games played at RFK, and compare that to the total number of runs scored in Nats road games.
Basically, the end result of doing that is that it doesn't matter how strong or weak the offense is, or how strong or weak the pitching is - since it's the same offense and pitching in both the home games and the road games.
So, a quick and dirty park factor for Washington can be calculated by looking at the total runs scored per game in RFK (264 RS by the Nats and 274 RS by their opponents in 72 games, for an average of 7.47 R/g) and the total runs scored per game in nats road games (301 runs scored bt the Nats and 326 by Nats opponents in 73 games, for an average of 8.59), and taking the ratio: 7.47 divided by 8.59 = 0.87, which means that RFK strongly deflates scoring. Since a team only plays half its games at home, you'll usually see park factors reported as half of the deviation from neutral (since half the games are on the road, which presumably averages out to a neutral park factor of 1. The average of 1 and 0.87 is 0.93, which is the value you'd normally see for the RFK park factor. Baseball Prospectus uses a more complicated method to determine park factor, and has RFK listed as 940, or 0.940.
If you've got the time and inclination and data, you can do more specific park factors - instead of just an overall park factor for runs, you can do park factors for hits, doubles, triples, HR, etc. But the calculation is basically the same.
Scot.
