The system itself is totally complete. However, as with anything else relying on data, the more data you have to draw from, the more accurate the conclusions will be that you come to from using that data. A not so good analogy would be if a new player moved into town and started playing at your pool room. If you only saw that player shoot one shot, you can probably get a very rough idea how they play but it won't be very exact at all. If you see them play a whole game you have a better idea but still not very precise. Seeing them play a whole set or even a whole night gets you an even better idea, but for a variety of reasons they could be really off or on that night and you still might have them pegged a little wrong. Same thing if you watched them play for the week. Maybe they just started playing again that week after years of a layoff, or they have had the flu all week and are not at their best, or maybe they have just been really in stroke playing above their head that week and that weeks play still isn't a perfect representation of their true average skill level. But after you have seen them play hundreds or thousands of games over the course of months or years you have them pegged pretty well because that additional data let you get more and more accurate with your assessment of their skill level. And so it is with FargoRate or anything else, more data always equals more accuracy.
Now limited data doesn't always mean horrible accuracy and FargoRate actually works pretty darn well right from the start even with very limited data because of the way it cross references players against each other. Kind of like how with sex and STDs they say you aren't just sleeping with that one person, but with every person they have slept with in the last ten years, and every person each of those people has slept with in the last ten years, etc. Well to drastically over simplify things FargoRate can in a similar way compare you to everyone else through just the people you have played against by using a very accurate and complicated version of "well if Adam usually beats Bob by this much, and Bob usually beats Carl by this much, then Adam will likely be beating Carl by this much" except that instead of just the three players there are usually hundreds or thousands of players in the equation and it becomes very accurate when you cross reference them in this way. But the more data you have (the more games you have played, and the more people you have played against), the better and more accurate the conclusions you can draw from that data, always, even if it was pretty good already with only limited data.
When FargoRate was made public for worldwide use (it had been used on a local/regional level for years), the initial problem you will have with a system like this is that it doesn't start off with as much data as would be ideal. Many people have played few or sometimes no matches where the results of those matches where give to FargoRate for inclusion in their data set. Mike Page did the best he could to get data from previous tournaments and leagues going back as far as ten years (the system uses play from the last ten years with more recent play carrying more weight and having more effect on one's rating that older play) so that it could start off with as much data as possible, but one of the difficulties was that not many tournaments or leagues keep records of all the individual matches, and especially not with the final scores also which FargoRate also makes use of and requires--not just who won or lost. The BCA national singles events is one where this data has been saved over the years, and they were kind enough to provide that information to FargoRate for inclusion, and Mike was able to scrounge up other results from some other tournaments and tours and league sources as well, but because past results including the final scores weren't saved by most people, especially outside of the pro events, the data that FargoRate started with was not as comprehensive as you would ideally like it to be. Now that FargoRate is public and worldwide however, and tournaments, and tours, and leagues start contributing their match statistics to FargoRate as they occur, more and more people will be included, and the already good accuracy gets better and better exponentially and quickly.
