Our Blog


Analyzing FantasyLabs’ Stolen Base Prediction Metric

Just when it seemed like the MLB was ready to make stolen bases great again, they once again disappeared. It has been an up-and-down last 26 years for one of the sport’s most-exciting events:

There was an average of 0.73 stolen bases per game from 1990 to 1999 compared to an average of 0.58 stolen bases per game from 2000 to 2016. The 2011 and 2012 seasons produced an average of at least 0.66 stolen bases per game – the highest mark in 12 years. There hasn’t been a season with an average higher than 0.57 stolen bases per game since then.

DraftKings and FanDuel both make it worth your while to target players with the ability to steal bases. A double is worth five DraftKings points, but a single followed by a stolen base amounts to eight DraftKings points. A triple is worth nine FanDuel points — the same amount as a three-point single followed by a six-point stolen base.

Stolen bases may not happen as much as they used to, but they’re still valuable commodities in DFS and can help provide upside for players without a ton of power. FantasyLabs’ SB Prediction metric is thus quite important for DFS. This prediction is based on factors such as hitter and pitcher stolen base rates, as well as the hitter’s probability of getting on base. Let’s take a look at how correlated the SB Prediction metric has been to DFS production.

SB Prediction Percentile

Using our Trends tool, we can see how hitters with the best and worst SB Predictions have performed over the past three seasons:

A few takeaways:

  • 91.2 percent of all hitters over the past three seasons have had a SB Prediction in the bottom-20th percentile.
  • Hitters with a SB Prediction in the top-80th percentile have performed well, posting a +0.28 Plus/Minus on DraftKings with a 42.5 percent Consistency Rating over the past three seasons.
  • The specific tiers of the SB Prediction metric haven’t been correlated with DFS production. This is at least partially due to the small sample sizes in the top-40th percentile.
  • Ownership has remained fairly low regardless of the percentile and doesn’t appear to be correlated with SB Prediction.

Within our Player Models, we offer a slider for a hitter’s SB Prediction that allows the user to weight the importance of the metric in their customizable model. It’s rare to find a player with a SB Prediction in the top-20th percentile, but simply targeting players who pose a threat to steal a base has proven to be correlated with DFS production.

Now that we’ve established that the SB Prediction metric has historically led to value, let’s take a look at where that value has come into play.

Targeting players with a SB Prediction greater than .20 has historically led to DFS value. The majority of players won’t have a SB Prediction over .20, but this shows that players with even a better-than-average chance at stealing a base have been valuable DFS assets. The list of players who have posted a SB Prediction greater than .61 over the past three seasons is limited to Billy Hamilton, Dee Gordon, Jarrod Dyson, and Rajai Davis. While rare, players with a SB Prediction greater than .40 have consistently produced the most value – and without an ownership premium.

Takeaways

FantasyLabs’ SB Prediction metric has proven to be correlated with DFS production. Here are some specifics:

  • Over 90 percent of hitters over the past three seasons have had a SB Prediction in the bottom-20th percentile. Players with a SB Prediction in the top-80th percentile have historically produced value.
  • Hitters with a SB Prediction over .20 have consistently produced value. SB Prediction has been positively correlated with DFS production, although instances of a hitter with a SB Prediction greater than .60 are very rare.
  • SB Prediction hasn’t been strongly correlated with a hitter’s ownership. There has been higher average ownership on FanDuel for players with a very high SB Prediction, but generally the public on both DraftKings and FanDuel have not rostered base stealers at a high rate.

When you’re constructing rosters with our Lineup Builder, remember to take into account hitter SB Predictions. And, of course, be sure to do your own SB Prediction research with the FantasyLabs Tools.

Just when it seemed like the MLB was ready to make stolen bases great again, they once again disappeared. It has been an up-and-down last 26 years for one of the sport’s most-exciting events:

There was an average of 0.73 stolen bases per game from 1990 to 1999 compared to an average of 0.58 stolen bases per game from 2000 to 2016. The 2011 and 2012 seasons produced an average of at least 0.66 stolen bases per game – the highest mark in 12 years. There hasn’t been a season with an average higher than 0.57 stolen bases per game since then.

DraftKings and FanDuel both make it worth your while to target players with the ability to steal bases. A double is worth five DraftKings points, but a single followed by a stolen base amounts to eight DraftKings points. A triple is worth nine FanDuel points — the same amount as a three-point single followed by a six-point stolen base.

Stolen bases may not happen as much as they used to, but they’re still valuable commodities in DFS and can help provide upside for players without a ton of power. FantasyLabs’ SB Prediction metric is thus quite important for DFS. This prediction is based on factors such as hitter and pitcher stolen base rates, as well as the hitter’s probability of getting on base. Let’s take a look at how correlated the SB Prediction metric has been to DFS production.

SB Prediction Percentile

Using our Trends tool, we can see how hitters with the best and worst SB Predictions have performed over the past three seasons:

A few takeaways:

  • 91.2 percent of all hitters over the past three seasons have had a SB Prediction in the bottom-20th percentile.
  • Hitters with a SB Prediction in the top-80th percentile have performed well, posting a +0.28 Plus/Minus on DraftKings with a 42.5 percent Consistency Rating over the past three seasons.
  • The specific tiers of the SB Prediction metric haven’t been correlated with DFS production. This is at least partially due to the small sample sizes in the top-40th percentile.
  • Ownership has remained fairly low regardless of the percentile and doesn’t appear to be correlated with SB Prediction.

Within our Player Models, we offer a slider for a hitter’s SB Prediction that allows the user to weight the importance of the metric in their customizable model. It’s rare to find a player with a SB Prediction in the top-20th percentile, but simply targeting players who pose a threat to steal a base has proven to be correlated with DFS production.

Now that we’ve established that the SB Prediction metric has historically led to value, let’s take a look at where that value has come into play.

Targeting players with a SB Prediction greater than .20 has historically led to DFS value. The majority of players won’t have a SB Prediction over .20, but this shows that players with even a better-than-average chance at stealing a base have been valuable DFS assets. The list of players who have posted a SB Prediction greater than .61 over the past three seasons is limited to Billy Hamilton, Dee Gordon, Jarrod Dyson, and Rajai Davis. While rare, players with a SB Prediction greater than .40 have consistently produced the most value – and without an ownership premium.

Takeaways

FantasyLabs’ SB Prediction metric has proven to be correlated with DFS production. Here are some specifics:

  • Over 90 percent of hitters over the past three seasons have had a SB Prediction in the bottom-20th percentile. Players with a SB Prediction in the top-80th percentile have historically produced value.
  • Hitters with a SB Prediction over .20 have consistently produced value. SB Prediction has been positively correlated with DFS production, although instances of a hitter with a SB Prediction greater than .60 are very rare.
  • SB Prediction hasn’t been strongly correlated with a hitter’s ownership. There has been higher average ownership on FanDuel for players with a very high SB Prediction, but generally the public on both DraftKings and FanDuel have not rostered base stealers at a high rate.

When you’re constructing rosters with our Lineup Builder, remember to take into account hitter SB Predictions. And, of course, be sure to do your own SB Prediction research with the FantasyLabs Tools.