ROTISSERIE: Where have all the steALs gone?

Recent MLB trends regarding the decline of stolen bases have been analyzed in a number of outlets, including the 2017 Baseball Forecaster ("Devaluation" and "2016 Trends, 2017 Responses" essays), as well as the 9/23/2016 BaseballHQ Radio podcast with Patrick Davitt and Joe Sheehan. A deeper dive into the data raises some particular findings for those in AL or NL "only" leagues, which will be explored here. 

Macro Level Analysis

Yes, as hinted in this article's title, there has been a major stolen base adjustment in the American League.

AL   2013  2014  2015  2016
===  ====  ====  ====  ====
SB   1428  1394  1185  1149
CS    513   492   533   448
SB%   74%   74%   69%   72%
SBO  8.9%  8.8%  8.2%  7.5%
                
NL   2013  2014  2015  2016
===  ====  ====  ====  ====
SB   1265  1370  1320  1388
CS    494   543   531   553
SB%   72%   72%   71%   72%
SBO  8.3%  9.1%  8.8%  9.1% 

Steals dropped by over 200 in the AL during 2015 and continued their decline in 2016, while the NL has remained relatively stable during the period. Starting in 2014, teams ran less frequently in the AL (SBO), and their poor success rate in 2015 probably contributed to further clamping on the brakes in 2016. At the same time, home runs are up in each league (+382 in the AL in 2016 and +319 in the NL) as are strikeouts (about +700 in each league), making it less likely for managers to risk outs on the basepaths. 

Team-level data for stolen base attempts in 2016 shows a great disparity between the runners (MIL, 237) and the non-runners (BAL, 32).  The Orioles' 19 steals in 2016 were the lowest MLB team total since 1972 and the eighth lowest since 1913. 

2016
SB Att    AL                NL
======    ===============   ==========
   <40    BAL
 40-80    OAK,TOR           NYM,STL,LA
 81-100   SEA,DET,NYY,TAM   MIA,CHC
101-120   BOS,LAA,CHW       COL,ATL,SF   
121-140   MIN,TEX
141-160   HOU,KC            PHI,PIT,WAS
161-200   CLE               ARI,SD,CIN
200+                        MIL
---------------------------------------
Lg Avg    107               130

Note that the distribution is slightly tighter for the AL (BAL 32 to CLE 165) than the NL (NYM 60 to MIL 237), which has many more teams at the upper end. Another phenomenon, discussed during the above-referenced podcast, is that the incremental improvements in pitchers' times to the plate and catcher pop times have shaved the margins of success for all but the best baserunners as they go from first to second base.

Since fantasy games are based on individual players and their steals, rather than whole teams' attempts, how does this translate down to the player population? Building off of Patrick Davitt''s chart in the 2017 Baseball Forecaster, here is the distribution of actual 2016 and projected (February 2, 2017 projections) basestealers in the various cohorts:

Players by Stolen Base Tier
     2016 Actual     2017 Projected
        AL  NL       AL  NL
================     ================================================
50+      0   2        0   3  B Hamilton, J Villar, D Gordon
40-49    1   1        1   1  J Dyson / T Turner
30-39    3   6        2   4  R Davis, T Anderson / S Marte, J Peraza
                             T Jankowski, K Broxton
20-29    8   7       14  14
10-19   28  24       35  30

At this time of the off-season, it is not surprising to see more players in the lower tiers of the projections, since playing time is still in flux while some free agents are still unsigned and a few jobs will be won in spring training. Expect the projected tier cohorts to approach last year's actual counts as we get closer to Opening Day. 

Gaming Implications

The loss of 200 SB in the AL tightens the category, making every steal more important in the standings. Current Standings Gain Points (SGP) Denominators in RotoLab highlight that, with a 7.5 SGP Denominator for 12-team, AL-Only, 5x5 leagues (each 7.5 steals are worth, on average, one point in the Stolen Base category standings) with a higher 8.1 SGP Denominator for Stolen Bases in NL leagues of the same format. 

Note in the table above that eight of the top 11 projected base stealers are in the Senior Circuit. NL-only league owners could choose to grab two or three of the top players listed in an attempt to win the category, or roster at least one for a good chance to be in the middle of the stolen base pack. In the AL, the dearth of elite basestealers makes the incremental gains possible from judiciously rostering players in the lower tiers a viable approach. Chasing the few top basestealers in the AL is both risky (more about that below) and likely expensive if at least one other owner in your league has the same strategy. A better approach in the AL may be to bump up each player in the two lower projection tiers a dollar (or round) or two, and grab several of them to ensure landing in the middle of the category at the end of the season. In 2016, those two tiers combined for 36 players between 10-29 steals, or about three per team in a 12-team league, so collecting four (or more) of them may be needed to stay out of the bottom of the category.

Also note that the two top projected AL basestealers have changed addresses in the 2016-17 offseason. Jarrod Dyson (OF, SEA) moved from KC to SEA, while Rajai Davis (OF, OAK) moved from CLE to OAK. In both cases, those moves are from teams in the upper echelon of stolen base attempts to those in the bottom tier. Are they likely to run as frequently in their new homes? Managerial tendencies provide some interesting insights. 

For Davis, the answer appears to be clearer. In OAK, manager Bob Melvin's teams have averaged 108 SB attempts since the start of the 2011 season. That 2011 season did feature 154 SB attempts, including 58 by a then age-31 and healthier Coco Crisp (OF, FAA), which shows that Melvin is willing to run when he has an elite basestealer. Davis generally led off in CLE against LH pitching and batted lower in the lineup against RH pitching during 2016, but his SB attempts occurred at about the same rate regardless of where he batted in the lineup. The one question with Davis is how long he can continue to amass stolen base totals that exceed his age (turning 36 in 2017). 

Of course, Dyson and everyone else ran in KC (175 SB attempts per year under Ned Yost 2011-2016), but Yost is the only MLB manager Dyson has ever played for. The track record for SEA manager Scott Servais is much smaller (2016 only), but his 84 SB attempts coupled with only 79 runners in motion in 2016 (AL-average was 124) could have reflected less willingness to risk those baserunning outs despite Dyson's 82% success rate. However, recent quotes by Servais and General Manager Jerry Dipoto indicate that the Mariners plan to take full advantage of the speed skills belonging to Dyson, trade acquisition Jean Segura (SS, SEA) and returnee Leonys Martin (OF, SEA) during 2017.

One place to look for confirmation of these strategy-change clues as well as other team stolen base tendencies, is spring training baserunning, as suggested in this article in the BaseballHQ Strategy Library. Monitoring SBO during spring training, and comparing that to the past three years of regular season SBO activity, can signal shifts in team baserunning philosophies. Servais' one year at the helm featured a definite SBO slow down for Martin, Ketel Marte (SS, ARI) and Kyle Seager (3B, SEA), but a rebound from the returnees and early frequent base-stealing from the new arrivals during the spring training sample would lend credence to the above-mentioned quotes and signal the managerial shift solidifying Dyson's current league-leading projection. As a recent example, Trea Turner (SS, WAS) ran amok in 2016 spring training (58% SBO), presaging his 33-steal half-season in the majors.

Conclusion

The recent history suggests stolen base stability in the NL with a general decline in the AL. Fantasy owners in NL-only leagues should feel confident in applying a variety of strategies for the stolen base category, exploiting the top tier of basestealing teams and the broad distribution of the basestealing players. While mixed leagues may notice a slight decline in stolen bases (about 9% compared to 2014) thanks to the decline in running tendencies in the AL, MLB totals have varied by up to 10% annually since the year 2000, so no significant change in strategy is needed. However, the current stolen base scarcity in AL-only leagues is significant, requiring savvy owners to adjust pricing models accordingly. Yes, it is only one category, but the total quantity decline must be addressed within the context of an overall draft strategy. The risk-averse owners in AL-only leagues will focus on collecting multiple players from the lower tiers of the base-stealing strata for a mid-category finish, and leave the few top AL base stealers for the risk-takers.   


Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.