The Captain's Blog

Break up the Toronto Blue Jays! Entering play today, Jose Bautista and company have been tearing up the Grapefruit League, compiling an impressive 22-4 record, which represents the highest spring winning percentage since 1984 (excludes the abbreviated 1990 exhibition schedule). If the Blue Jays maintain their above .800 winning percentage, they’ll join the 1997 Marlins as the only team to end the spring above that level. That year, Florida went 26-5 during the exhibition season and then followed up on its  success by winning the World Series. Talk about a good omen.

Top-10 Spring Training Records Since 1984

Note: Excludes the 1990 and 1995 exhibition seasons, which was shortened by work stoppages. Tie games excluded from winning percentage calculations.
Source: mlb.com and springtrainingmagazine.com

What do Spring Training standings really mean? Teams that struggle in the spring usually dismiss the importance of a poor record, while those who tear it up in March like to believe it’s a sign of good things to come. Perspective also plays a role in the interpretation. Younger teams will often talk about “learning how to win”, while established ball clubs are just “trying to get their work in”. In other words, the importance of spring performance is really in the eye of the beholder.

Putting the clichés aside, what does history tell us about at the relationship between success in the spring and summer? The chart below presents the correlation between each franchise’s exhibition and regular season records since 1984.  Based on the data, the Tampa Rays (a relatively young ball club that until recently was “learning how to win”) are the only team that has had a strong relationship between spring training and regular season results on a year-by-year basis. A few other teams have exhibited a moderate correlation, but almost 75% have shown a relatively insignificant relationship (both positive and negative). So, at least in this regard, it appears as if Spring Training standards don’t count for much after all.

Correlation Between Spring Training and Regular Season Records, Since 1984

Note: Tie games excluded from winning percentage calculations.
Source: mlb.com and springtrainingmagazine.com

Although a correlation analysis suggests that March performance has little bearing thereafter, is that really the best approach to take? Using such a broad method with a relatively small sample size (28 seasons) can leave the outcome vulnerable to outliers, so a more targeted analysis seems in order.  Besides, we really don’t want to know if a change in spring record from one year to the next leads to a corresponding regular season outcome.  Instead, the objective is to figure out if a link exists within a confined year.

Distribution of ST and Regular Season Winning Percentage Divergence, Since 1984

Note: Tie games excluded from winning percentage calculations.
Source: mlb.com and springtrainingmagazine.com

As illustrated by the chart above, approximately 37% of all teams since 1984 had an extreme absolute divergence between spring and regular season records (defined as a differential of 10%-plus, or 16 or more wins), while 34% varied by a modest total (5%-10%, or 8-16 wins), and the remaining 29% came relatively close (0%-5%, or within no more than 8 wins). In other words, there was a pretty even distribution across the broad spectrum, suggesting again that exhibition performance does not have predictive value.

So, is that final word? Not really. After all, what we really care about is whether a strong spring has a tendency to translate into a playoff berth. In order to examine that question, we can look at the composite spring training record of playoff teams, which happens to be .530 since 1984 (.527 for division winners and .546 for wild cards). Once again, however, an aggregate analysis can be skewed by outliers (which is particularly relevant if we assume established playoff teams do not take spring training seriously). So, in order to overcome that bias, the next step is to look at the spring training winning percentage distribution for all playoff teams since 1984.

ST Winning Percentage Distribution of Playoff Teams, Since 1984

Note: Tie games excluded from winning percentage calculations. Pie chart compares spring records among playoff teams only. Bar chart compares playoff and non-playoff teams within defined winning percentage range.
Source: mlb.com and springtrainingmagazine.com

According to the pie chart above, two-thirds of all playoff teams over the last 28 years have at least played .500 in the spring, and only 13% have reached the post season after playing sub.-400 baseball. However, we can refine this snapshot a little further by determining the percentage of teams within a given spring training record range that wound up making the playoffs (the bar chart). Using this approach, it appears as if the chances of making the post season gradually decrease as spring training records decline. So, there may be at least a casual link between success in March and October.

Where does that leave us? Basically, if you are trying to determine a team’s improvement based on a comparison of spring records, it will be a lost cause. Similarly, a particular winning percentage in March doesn’t necessarily mean a corresponding outcome in the regular season. However, if a team’s spring record is on either extreme, it could indicate the likelihood of a post season berth, which brings us back to the Blue Jays. Of the 13 teams since 1984 that have played over .700 ball in March, seven have made the playoffs. What’s more, Jays’ fans can also take comfort in the spring struggles of the Rays and Rangers, who each have won just one-third of their exhibition schedule. That doesn’t guarantee the Blue Jays their first playoff berth in 20 years, but it does give Toronto fans every right to be optimistic heading into the season.