When trying to assess a team’s overall talent in reference to the rest of the league, several things come up. How good would the team be if they were in division ‘X’? How good would they be if there was a balanced schedule? What if there were no divisions?
As Jays’ fans, we ponder these questions often in reference to our own team’s plight in the daunting and hellacious AL East. We find ourselves wondering if our lovable but maddening purveyors of mediocrity would see the playoffs if only the gods of geography would allow them a move to either Central division. How much of Toronto’s failure (or success) is due to its division and how much of it is due to actual team talent?
I have been arguing all season that the Jays are at least as good, if not better than, the Tigers and Angels. Obviously Detroit’s late season surge ended the debate for now, but in my mind the Jays are a better team than the Angels and have been at least as good as the Tigers up until the last month.
I started wondering if there was a way to prove this. Is there a way to estimate a team’s record if both luck and division placement were non-factors?
Part of that question has already been answered by Bill James. James developed Pythagorean Winning Percentage (PW%) which is a team’s estimated winning percentage based entirely on run differential. The record of most teams closely resembles their PW%, but in some cases a team either drastically over- or under-performs their PW%. There are a number of theories as to why, but because it doesn’t seem to be a repeatable team skill, most chalk it up to luck.
So that takes care of the luck factor (at least as much as possible), but what about the other portion of my original question? What would a team’s record look like if every team were to play a balanced schedule? In other words, how much better would the Blue Jays be if they didn’t have to face the Red Sox, Yankees and Rays 19 times each in a given season? What if they only had to play them as much as everyone else? And what if Interleague Play was banished so teams wouldn’t have to play in games against other teams with different rules?
Enter: Weighted Pythagorean Winning Percentage (wPW%).
This stat is entirely too simple in concept to be A) original or B) accurate, but it’s interesting nonetheless. How is it calculated?
First let’s look at PW%:
James’ original PW%, modified by Baseball Reference for accuracy:
= [RS^1.83]/[(RS^1.83) + (RA^1.83)]
‘RS’ is runs scored and ‘RA’ is runs against.
wPW% is calculated simply by taking each team’s PW% against each division in their league and finding the average. This allows us to estimate how the team would have performed if they played an equal number of games against each division. Smaller and larger divisions like the AL West and NL Central are weighted accordingly as well. Again, Interleague Play was removed from the equation partly because of its ridiculousness and partly because teams do not play against every team in the opposite league in a given season, so estimating it here would be misleading.
So, to start off, let’s have a look at the actual 2011 final standings in both leagues; Pythagorean records are also included in these standings:
As you can see, removing luck as best we can would have put the Red Sox into the playoffs as the AL Wildcard, three games over the Rays. The Indians would see a five game deduction, while the Royals underperformed their P-Record by seven games. In the NL, St. Louis would have won the Wildcard by three games over Atlanta and four games over the Dodgers, while only finishing two games back of the Central division winning Brewers.
Now, here is an estimate of how the standings would look in each league if divisions were removed and everyone played a balanced schedule against only their league. First the AL:
The ‘/Record’ column shows how many games better or worse a given team would be against their actual record, while the ‘/P-Record’ column does the same against their Pythagorean record. The far-right column shows the team’s actual ranking within their league.
As you can see, the Blue Jays would benefit most in the AL by playing a balanced schedule, while the Clevelands would suffer the most. The Jays wPW% would have them as an 84-win team, outperforming their actual record by three games and their Pythagorean record by five. The Clevelands, meanwhile, would drop to a 69-93 record (fourth in their division!) according to wPW%. The Tigers would finish with 88 wins and the Angels with 82, meaning my original thought about their skill compared to Toronto’s wasn’t too far off. The Jays would actually have been the sixth-best team in the AL, only three games worse than the Rays and only four worse than the Tigers.
Also of note in the AL, both the Rays and the Red Sox would actually be worse (in terms of their Pythagorean records) if a balanced schedule were to be introduced.
In the NL, there are a lot of large discrepancies. The Phillies would go from being the best team in the NL, to being the only truly elite team as the Brewers and Diamondbacks wPW% have them at just 89 wins each, 14 games worse than Philly. The Braves meanwhile would improve their standing by three wins over their Pythagorean record, moving them past the Cards for the NL Wildcard.
The two craziest results in the NL are found in the NL West where the Padres, helped by a solid run differential overall, would improve their record to 84-78 with wPW%, buoying them into marginal playoff contention. The Giants meanwhile were so terrible against both the NL East and NL Central this season that their wPW% dropped them to a terrible 73-89 record.
Lastly, the Pirates’ wPW% puts them two-tenths of a percent behind the Astros for the worst mark in the NL.
You could make the case using this stat, that the AL Central far outperformed their collective talent, and for the most part, the AL East was brought down slightly by their division mates, but no concrete conclusions can be made about how good or bad any of the other four divisions are. We can still assume, because of Interleague Play and other factors, that the AL is a much better league overall than the NL, but wPW% does not seek to answer that question.
In sum, the Jays would have been better (even marginally contending) if there was a balanced schedule and they would probably be better than the Angels and only slightly worse than the Tigers and the Rays. When I first calculated this stat back in late August, the Jays were about two games better in wPW% than the Tigers; it wasn’t until Detroit’s late-season hot streak that they truly surpassed them.