Which team is more likely to have a good season is getting into projections rather than actual records. At this point in your scenario, both teams are actually 1-2. Obviously using the pythagorean method, the first team is more likely to have a better season. While you're playing fast and loose with statistics, small sample sizes, and projections, do you really think that team 2 will only win 9% of the games (their pythagorean win ratio) during the whole season? Of course not. Obviously in that tiny of a sample size, a million variables come into play. The margin of error is so high. Even after 162, there is still margin of error on both Pythagorean records and actual records. Neither records a team's talent level accurately, although obviously the margin is much smaller than the microscopic 3 game sample. I personally don't agree with throwing out either number when coming up with projections. The Pythagorean record is a very good allocation of the things that translate into wins (runs scored/runs allowed), but at the same time we don't know exactly why a team overperforms/underperforms it. A projection that ends up somewhere between the two records (before taking into account changes in the roster) would probably be the most reasonable approach IMO. Using actual records is a good starting point for a projection, but I don't know why that wouldn't be amended using the data we have that might suggest a team is better or worse than their record. To get back to the beginning of the discussion, I'm not sure why you would want to throw out the Cubs Pythagorean record when making a projection for next year. It's an important data point even though it definitely still has error in it. While it's fun and interesting to project and predict what might happen, the bottom line is that the actual W-L record is all that counts. If you're a 70 win team that projects to be a 78 win team, you still have to gain 11 actual wins, not 3, to be .500. With the CBA, money being tight, and more teams locking up young stars, it's making it much harder to gain 8-10 wins in a season.