Chicks dig the long ball. It’s a matter of fact. Also a matter of fact is that male humans also dig the long ball. Everyone digs the long ball! But the side effect of hitters suddenly knocking balls out of the park and rounding the bases like never before (the 12.8 percent league homers-to-fly balls rate in 2016 was the highest mark since FanGraphs has data, which goes back to 2002) is that now there are fewer opportunities to steal bases. You can’t steal a base if you hit the ball over the fence!
Love it or hate it, stolen bases is either a rotisserie category or an event that earns you points in most fantasy leagues. So with the rise in power and decline in steals, what’s a fantasy owner to do?
Let’s first take a step back and examine the stolen base landscape. We’ll begin with a pretty chart of historical stolen base totals over the last 10 years.
We actually hit a 10-year low in 2015, but barely bounced off it in 2016. So clearly, stolen bases are tougher to come by than they were before 2015. How has this affected fantasy leagues? Luckily, I have been running my local fantasy league since founding it in the early 2000s and I have access to all its historical data. The league is a standard 12-team mixed format, with 23-man rosters, so it is close to a standard league and as such should provide a good example from which to draw conclusions.
Let’s see how many stolen bases my league has accumulated each season, and then what percentage of all major league stolen bases that represented for that year.
|Season||MLB SB||Fantasy League SB||% of MLB SB|
And, in graphical format:
So my local league’s stolen base trend essentially mimics the major league trend. Overall stolen bases have declined, and my league’s stolen base total hit a low in 2015 and rebounded marginally last year.
The interesting column is the last one, which represents the percentage of steals my league had on its active roster earning stats. It’s been incredibly consistent and hasn’t followed the same trend as total steals.
Let’s continue digging by looking at the group of players who stole at least 30 bases each year. It’s an arbitrary number, but we’ll go with it. I’ll also include the stolen base total of the major league leader, as well as the percentage of total steals that leader represented.
|Season||MLB SB||# 30 SB Players||SB Leader||SB Leader % of MLB SB|
Not surprisingly, the number of players reaching the 30-steal threshold has followed a similar pattern to the total major league stolen bases. And gone are the days when the major league leader is in the high 60s or even into the 70s. But if we look to the last column, the percentage of overall steals the leader represents has remained relatively constant, with an odd valley in that 2011-to-2013 period.
What that last column suggests is that the current crop of elite base thieves are no more valuable in fantasy leagues now than they were when steals were more plentiful. So if you happen to hear that because stolen bases are down across the league, it’s even more important to draft a top base stealer like Billy Hamilton, you now know that such advice is not supported by any sort of statistical or mathematical evidence.
Perhaps the thought process behind such advice is that the Billy Hamilton’s and Dee Gordon’s could always go off for 70-80 steals and then they would be more valuable than they had been in the past. Well, of course if they steal more bases than they ever have, they would be more valuable! But also, this would be true in a relative sense as well. If major league-wide steals remained the same, an even swipe-happier Hamilton would represent a slightly higher percent of the league total, making him more valuable than if he had accomplished that in 2014. Even so, it’s poor advice because any player exceeding his historical performance, assuming a constant league environment, is going to be more valuable.
So how does stolen base scarcity impact the value of the top base thieves? The answer is … it doesn’t. As long as the top base thieves are also following the trend and stealing fewer bases, then there’s no effect. Everything is relative in fantasy baseball and the speed trend and its lack of influence on the value of speedsters is a perfect example of this concept.
Stolen Bases and Spring Training
The first day of spring training games is always an exciting time of year. Baseball! Naturally, we want to find some glimmer of meaning in the stats players post. Is that hitter who just swatted nine spring homers about to enjoy a power breakout? Is the pitcher who just struck out 30 percent of the batters he faced on his way to putting it all together en route to a career best ERA?
While the varying levels of competition and small sample size make it difficult to glean anything impactful, perhaps stolen bases does provide us some insight. That’s because stolen base totals aren’t just due to player skill, but also managerial tendency. If a manager decides he wants to run more often, he might let his team loose during spring training to test out his new strategy. We can then look at historical data to determine whether those spring training stolen base spikes or declines versus the previous year carried over to the regular season.
Let’s begin by calculating the correlation between stolen base attempts during spring training and the regular season. I’m not going to use straight stolen base attempts since they are heavily influenced by the number of times a hitter gets on base. Instead, I have created a formula that estimates the number of stolen base attempts a player or team records per the number of opportunities. I call the equation SBA/TOB, or stolen base attempts per times on base. It’s not technically all times on base, but just those times a hitter would typically have the opportunity to swipe a base, assuming another player isn’t standing on the base in front of him (which I have no choice but to ignore).
SBA/TOB = (SB + CS) / (1B + 2B + BB + HBP)
It’s pretty simple and straightforward, as it assumes a hitter has the opportunity to steal a base on his singles, doubles walks and hit-by-pitches. Sure, there will be times when he’s blocked from swiping, but since we care more about this rate relative to the team’s rate in the past, the absolute accuracy of the rate isn’t all that important. One note — for all the research I’ll be presenting, I’m using spring training and regular season statistics from 2012 to 2016.
So let’s get back to the correlation between SBA/TOB during spring training and SBA/TOB during the season. Below I have included the correlations between all the components of the metric from spring to the regular season.
|Spring to Season Correlation||0.355||0.216||0.463||0.100||0.243||0.306||0.272|
So we find that the majority of these metrics, aside from HBP, does correlate to a meaningful degree between spring training and the regular season, though the correlation isn’t that strong.
Now let’s look at SBA/TOB during spring and the regular season over the years at the major league level.
|Season||SBA/TOB ST||SBA/TOB RS|
We can make two clear observations here:
- Teams run more frequently during spring training than the regular season.
- The SBA/TOB trend has followed the overall stolen base trend discussed above.
Now let’s get to the meat of the research and determine how teams that ran significantly more or less during spring training versus the previous regular season performed during the season following. Below is a table with group averages based on how the teams in that group performed in spring versus the previous season and then how they ended up performing during that season.
|SBA/TOB Diff Yr 1 to ST Yr 2 Range||# Teams||Avg SBA/TOB Diff Yr 1 to ST Yr 2||Avg SBA/TOB Diff Yr 1 to Yr 2||% Up||% Down|
|>= .02 & < .04||26||0.030||0.002||50.0%||50.0%|
|> -.01 & < .02||43||0.006||-0.006||39.5%||60.5%|
|> -.02 & <= -.01||14||-0.014||-0.010||35.7%||64.3%|
Well gosh darnit, the data tell us exactly what we were hoping. Spring stolen base spikes and declines do mean something, and perhaps a lot!
When looking at the table, just remember that the average team posts a .014 increase in SBA/TOB during spring training versus the regular season, so you see that with that middle group, if they are generally around the same in spring versus the previous year, that’s not a good sign.
You see the second group from the top even increased their SBA/TOB by a whopping 0.03 on average and that still only resulted in an almost identical SBA/TOB as the previous season, with half the teams increasing and half decreasing.
So, really, if you’re looking for players who might benefit from a managerial tendency to run more often, stick with the top group only, otherwise it’s far less of a guarantee. Similarly, be cautious with players on teams that posted an SBA/TOB at least .02 worse in spring than the previous season, as all 19 of those teams ran less frequently!
Once again, this kind of analysis looks at the averages of a group, which means that not every team within the group is going to follow. Let’s look at some of the teams you might be curious about that either increased their SBA/TOB or decreased it significantly in 2016 and whether their spring performance was a sign of things to come.
|Team||SBA/TOB Diff 2015 to 2016||SBA/TOB Diff 2015 to ST 2016|
The Brewers ran wild in 2016 and not only led baseball in SBA/TOB, but did so by a crazy 0.027. Their .138 mark was the highest in the five-year period by .022. Of course, not all of that should be explained by managerial philosophy, as speedsters Jonathan Villar, Hernan Perez and Keon Broxton all had histories of serious thievery.
Surprise, surprise, the Padres’ stolen base outburst could not have been predicted by spring performance! The biggest regular season shocker was Wil Myers, who nearly doubled his SBA/TOB from a then-career high .092 in 2015 to 0.174 in 2016. And guess what … he attempted just one steal during the 2016 spring!
Now let’s look at the teams that declined the most from 2015 to 2016.
|Team||SBA/TOB Diff 2015 to 2016||SBA/TOB Diff 2015 to ST 2016|
Wowzers, only three of the seven teams actually posted a decline in SBA/TOB during spring training, and the four that posted increases enjoyed dramatic jumps. In fact, the four spring surgers made up half of the top eight in spring spikes, and the Cardinals actually topped the list! It goes to show you that just because group averages from a data set suggest something doesn’t necessarily mean it’s going to happen every time.
Obviously, the Marlins’ loss of Dee Gordon for half the season had a lot to do with their decline. The Cardinals’ problem was that only one of the six players who stole more than one base during spring training actually earned any sort of significant regular season playing time. The one was Stephen Piscotty, who attempted six steals in just 57 spring plate appearances and posted a 0.286 SBA/TOB, then was back to normal during the season, posting just a 0.061 SBA/TOB. It’s certainly not the first time Cardinals manager Mike Matheny has defied logic.
So while nothing is ever a guarantee when dealing with humans and baseball, when looking to mine for stolen base upside, target players on teams that posted a spike in SBA/TOB during spring versus the previous season. Equally, be wary of the potentially hidden downside for players whose teams posted an SBA/TOB significantly lower than their previous season mark.