Which MLB teams will improve in 2015, and which will decline?

In his annual Baseball Abstracts during the 1980s Bill James often used to predict which teams would improve or decline relative to their win total of the previous year based on six "indicators":

1.  Pythagorean Record. If a team exceeds its Pythagorean expectation in one season then James expected them to decline the following year (and vice-versa). This is basically the "Johnson effect" and was discussed in my last sabermetrics article. For 2015, the biggest bounceback candidates using this indicator are the Athletics and Rockies, whereas the Cardinals and Yankees would be most expected to decline.

2. The "Plexiglass Principle". According to Bill James "when a team improves in one season, they tend to decline in the next season; when they decline in one season, they tend to improve in the next season" (This Time Let's Not Eat The Bones, p.480). The biggest improvers in 2014 were the Angels and Astros, whereas the Red Sox and Rangers were the largest fallers.

3.  The "Law of Competitive Balance" (also known as the "Whirlpool Principle"). This is simply "regression towards the mean". James stated that "all things in baseball are drawn powerfully toward the center" — teams which exceed 81 wins are expected to decline and those that finish below .500 are expected to improve. Under the Law of CB, the teams with the best record in 2014 (the Angels, Orioles and Nationals) would be expected to decline in 2015, whereas the Diamondbacks, Rockies and Rangers would be predicted as improvers.

4. Age. James expected that younger teams would improve and older teams would decline. Based on predicted hitting rosters, the Rays and Padres have got much younger and the Astros and Mariners have got older, whereas for pitchers the Blue Jays and Yankees have got younger and the Astros (again) older.

5. AAA performance. Teams with a strong AAA team are expected to improve and vice-versa. The Nationals had the best record in the International League in 2014 whereas the White Sox were the worst, while in the Pacific Coast League the Diamondbacks and Mets had the best record, whereas the Rockies were the low men on the totem pole.

6. Second-half performance. In a study in the 1986 Baseball Abstract, James found that "teams which finished the season playing well did, in fact, do substantially better in following seasons". The Orioles, Rays and Nationals improved significantly in the second half, while the Brewers, Athletics and Reds declined.

Using the indicators to make predictions

In 2006 Craig Burley revived discussion about the Bill James indicators in an excellent article in the Hardball Times. He used results from the 2005 season to generate indicators for the 2006 season, which subsequently proved very accurate in predicting which teams would get better or worse. In this article we shall revisit the 2006 season, while also looking at the predictions/results for 2014 and the predictions for 2015.

Of course, using these indicators in their raw form is a rather crude means of making predictions, since they take no account whatsoever of off-season free agency moves or trades, nor indeed of any injuries. Furthermore, there is no consideration of the strength of any single indicator — for example, for the law of competitive balance in this system, a team that wins 100 games counts the same as a teams that wins 82.

Also, as Craig Burley states, "the system tends to award points to bad teams. Not only is one point explicitly awarded to bad teams (thanks to the Law of Competitive Balance) but also bad teams tend to do well in the other categories as well." However, as he points out, this simply reflects the reality that bad teams do tend to improve. And, as the tables for 2006 and 2014 (see below) show, when there are many indicators that point either in one direction or the other, these indicators appear to have been pretty successful in predicting which teams will improve or decline.

Now let's move on to the predictions themselves, where we show each indicator as either "up" (suggesting improvement) or "down" (suggesting decline). If a team has a significant number of "up" indicators (say 4.5 or more) then we would expect improvement, whereas 4.5 or more "down" indicators would indicate decline.

Note that the age category has been split into two (hitters and pitchers) — half a point is awarded for each category. I used a combination of Baseball Reference and Roster Resource (for current predicted rosters) to generate the information in the 2014 and 2015 tables. The number in brackets in the prediction column indicates the strength of that prediction so "UP (4)" means that four of the six indicators are "up",

2006

Let's start with the 2006 table, which is based on results from 2005. This table has been updated (to include outcomes) from Craig Burley's original HBT article, and is reproduced here with the kind permission of the author. (I've allocated the Astros to their current division so that all the tables are consistent.)

The 2006 indicators came up with the following "strong" predictions:

• 6 down: White Sox
• 5.5 up: Royals
• 5 up: Pirates, Rockies
• 5 down: Angels, Cardinals, Padres
• 4.75 up: Tigers, Athletics, Reds
• 4.75 down: Nationals
• 4.5 up: Rangers, Brewers
• 4.5 down: Diamondbacks

Ten of those predictions were correct, with the Pirates matching their 2005 win total and incorrect predictions for just the Brewers and Padres.

2014

Now let's take a look at the 2014 indicators (based on 2013 results).

In 2014 the system correctly predicted the outcomes for all eight of the teams with 4.5 or more indicators pointing in one direction or the other:

• 5.5 up: Angels
• 5 up: Marlins, Mets, Brewers
• 5 down: Diamondbacks
• 4.5 up: Mariners, Astros
• 4.5 down: Indians

2015

Finally, let's take a look at the indicators for 2015:

The Bill James Indicators for 2015 show unanimous support for the Rays to better their 77 wins in 2014, while also supporting the Red Sox and Padres (+4.5). Conversely, the White Sox, Brewers, Reds, and Giants are all strongly predicted (-5) to decline from their win totals last season, with the Angels (-4.75), Orioles (-4.75) and Tigers (-4.5) also predicted to get worse. It will be interesting to see how these predictions turn out.

As you can see, the indicators don't take a strong view on the 2015 Indians, but this may be just as well since in 2014 they correctly predicted (-4.5) that the Tribe would decline.