Baseball Notes

Here on this 31st day of July, in the year 2007, there are two long months left in the season. For some teams (like my favorite), the end can’t come soon enough; for others, today was a day spent bolstering the roster and upgrading to get ready for the playoff push. At this date and time, the only thing we know is exactly how each team has performed through the first four months and with their “old” rosters.

So I thought it might be a good time to look deeper into the records of each MLB team a little bit before they begin play with their new rosters intact and see if we can peer into the future. We can look not only at what teams current records are, but what their expected records should be at this point, as well as the playoff odds for each team, using a couple of handy tools.

Baseball Prospectus uses some well-known ideas and formulas to calculate adjusted records based around a team’s runs scored and runs allowed. The theory that runs scored versus runs allowed works in determining adjusted records was an idea thought up by Bill James to determine what a team’s true winning percentage would be based upon those two totals. His original formula was simply W% = RS^2/(RS^2 + RA^2), or the Pythagorean expectation. Eventually, he changed the exponent to 1.83 and found that the winning percentage more accurately reflected how a team actually performed.

In the following years, Clay Davenport more accurately modified the formula (in what he called Pythagenport), to reflect the run environment per game for every one of a team’s games. Instead of a fixed exponent of 2 or 1.83, he declared the best way to calculate winning percentage is to determine an exponent using data each team’s games with the formula for the exponent being X=.45 + 1.5 * log10 ((RS+RA)/g). So, in calculating W% for each team now looks like this: W%=(RS^x)/(RS^x + RA^x). While others have modified Pythagenport (read here and here for more explanation), Baseball Prospectus uses this theory to determine its adjusted standings for a given year.

And through these adjusted standings, BP has a program that runs every game for the rest of the season one million times, and plots the outcomes and assigns each team odds (or a percentage chance) that they will reach the postseason in that year. For example, the Astros adjusted record is 46-59 (they are where they should be), and they have a 0.09% chance of making the playoffs after running the simulation one million times for 2007 (wild card included).

So, before the last third of the season starts, here are the records, adjusted standings, run differentials, and postseason odds for each MLB team, according to Baseball Prospectus.

All of these stats are as of 3:00 PM on July 31, and the postseason odds includes winning the division or being the wild-card leader.

You start to notice some interesting things heres, particularly relating to teams with run differentials that don’t accurately reflect their current records.

Seattle - despite scoring only nine more runs than their opponents, the Mariners are 12 games over .500, and definitely in the wild card hunt. Pythaganport catches up to them in this exercise and shows that they truthfully only have a 33% chance of making the playoffs.

Arizona - the Diamondbacks own a half game lead in their division, and they are nine games over .500, but they are being outscored by 30 runs this season! This is the reason why they lose eight games in their adjusted standings and have only a 19% chance to make the playoffs.

NY Yankees - by outscoring their opponents by about 120 runs, the Yankees should have a record that is closer to the Red Sox than the Blue Jays. Part of this though, could be attributed to the fact that when the Yankees win, it can often be a blowout, skewing the run differential. Still, they only have a 29% chance of the postseason.

Kansas City -  despite playing in the toughest division in baseball, the Royals actually should be better than they actually are. With a 50-55 adjusted record, I imagine the only reason their postseason odds are so low is because they play in the same division as the Tigers, Indians, and Twins.

And according to the playoff odds report, your four AL representatives would be Boston, Detroit, LA, and Cleveland. The NL checks in with New York, Milwaukee, LA, and Chicago by a hair.

Of course, the reason I do this now is to be able to compare it to Sept. 30 when we can look back and see what kind of impact Teixeira and Dotel have on the Braves, or Castillo has on the Mets, or Gagne has on the Red Sox. We don’t know how the balance of power will shift over the next 60 days (especially in the NL East), but looking at a team’s performance in run differential, all things being equal, should give us a best guess as to how things will shake out with their records over the next two months.