Shooter Versus Defender: The Batter Versus Pitcher of Daily Fantasy Basketball
In daily fantasy baseball, there is an everlasting debate about whether BvP data is relevant, or more importantly, predictive. If you donâ€™t know, BvP means batter versus pitcher, and itâ€™s just a stat that tells you how a specific batter has performed against a specific pitcher.
Some analysts argue that just splits are important â€“- how a batter has hit against lefties versus righties. Others will argue that you can go even deeper and look at single-pitcher splits.
Iâ€™ve always thought it interesting that we rely so much on DvP in basketball â€“- defense versus position, or how well a team defends that specific position â€“- but not individual matchups, like some in baseball use BvP. For example, the Orlando Magic have given up the third-most fantasy points to opposing point guards this year. But their point guard, Elfrid Payton, is actually a good defender, ranking eighth-best in the entire league among point guards in terms of DRPM, ahead of guys like Kyle Lowry, Jrue Holiday, Russell Westbrook, Avery Bradley, and Chris Paul.
If Payton is a good individual defender, what gives with the DvP data? Why the discrepancy?
Well, Payton is playing 28 minutes per game this season, which means there are 20 minutes that are going to other players who may not be as good defensively. Backup point guard Luke Ridnour is averaging 14.3 minutes a game this year and is posting a -1.72 DRPM. When judging point guards against the Magic, we should look at the individual parts, not just solely DvP.
Perhaps this means that starting point guards going against the Magic are a little overvalued by DvP since theyâ€™re likely getting most of their minutes against the defensively superior Payton. Portland guard Damian Lillard is only 2-8 this year, for instance, when defended by Payton. In his one game against Orlando, he had 27.5 FD points and 30.75 DK points -â€“ well below what youâ€™d want from his $9,000-plus price tag on either of the big DFS sites.
I say all this because I want to start looking at what Iâ€™m terming SvD â€“- shooter versus defender â€“- as kind of a BvP idea for daily fantasy hoops. Iâ€™ve collected a lot of data recently about how each player performs when theyâ€™re guarded by a specific defender. As youâ€™d expect, itâ€™s a lot of data, and I havenâ€™t even started to come up with a good idea about how to make it available in a searchable database for those interested*.
So this is more of an introduction article, as well as a I-could-use-your-helpful-ideas type of article. Letâ€™s look at some examples from this season and then see what observations we can make. Then you can email/tweet me and we can try to become better NBA DFS players together.
|LeBron James vs Defender (min 5 shots)||Shots||3PM||Points Per Shot|
|vs Matt Barnes||2-8||1||0.71|
|vs Tarik Black||4-6||0||1.33|
|vs Caron Butler||1-5||0||0.40|
|vs Jimmy Butler||4-8||0||1.00|
|vs Robert Covington||2-6||1||0.83|
|vs Luol Deng||2-8||0||0.50|
|vs Andre Drummond||3-5||0||1.20|
|vs Pau Gasol||4-5||0||1.60|
|vs Taj Gibson||2-6||0||0.67|
|vs Solomon Hill||2-9||0||0.44|
|vs Kyle Singler||1-5||0||0.40|
|vs Tony Snell||4-12||2||0.83|
|vs PJ Tucker||5-9||2||1.33|
|vs Andrew Wiggins||11-18||2||1.33|
|Stephen Curry vs Defender (min 5 shots)||Shots||3PM||Points Per Shot|
|vs Trey Burke||7-11||4||1.64|
|vs Darren Collison||6-12||2||1.17|
|vs Dante Exum||3-7||2||1.14|
|vs Jordan Farmar||0-5||0||0.00|
|vs James Harden||3-5||0||1.20|
|vs Jordan Hill||4-6||0||1.33|
|vs Kyrie Irving||4-6||2||1.67|
|vs Zach LaVine||4-8||1||1.13|
|vs Ben McLemore||2-5||0||0.80|
|vs Paul Millsap||4-5||1||1.80|
|vs Derrick Rose||3-7||0||0.86|
|vs Ricky Rubio||1-5||0||0.40|
|vs JaKarr Sampson||1-5||1||0.60|
|vs Thaddeus Young||2-7||0||0.57|
You can see the difference between teammates â€“- look at Stephen Curry versus Trey Burke and against Dante Exum. I mean, he dominated them both because of his elite outside shooting, but it was much worse against Burke. The same thing with Minnesota â€“- you like a matchup against LaVine (1.13 pps), but not so much against Rubio (0.40 pps). DvP is going to give you how he does against both. But what if one of them is out or is playing more? Thatâ€™s where looking at SvD, or shooter versus defender, can help.
The biggest problem with SvD, BvP or any statistic that measures one individual against another individual is sample size. LeBron James has taken a lot of shots this year â€“- while he has had bad splits against Caron Butler this season, perhaps there's context thatâ€™s missing on why LeBron struggled in that particular game.
But much like BvP, all data is useful, if not (though Iâ€™m sure people will argue it is) predictive. Is there a LeBron stopper? There are certainly players who have contained him this year. Itâ€™s a small sample, sure, but a lot of people put a lot of money on the line based on small samples of BvP in baseball.
Look at LeBronâ€™s splits against the different Bulls players. Heâ€™s been average against wings Butler and Snell, and great against big man Pau Gasol. Looking at DvP as a whole is a good place to start to see how the Bulls defend against the small forward position. However, basketball isnâ€™t always as linear. What if Kevin Love is out for the Cavaliers and LeBron has to move over and play a lot of small-ball four? Iâ€™m sure the Bulls would probably still try to get Butler on him as much as possible, but it would increase the chances that a big man like Gasol would end up on him. If a player is multi-positional by nature -â€“ and honestly, most players in the NBA are nowadays â€“- DvP can be a little misleading, just as SvD can be small-samply.
The No-Position-Movement (I just made that up, but Iâ€™m definitely an advocate of it) is an interesting topic for NBA DFS. Thereâ€™s no other DFS sport that has players so loosely defined in real life, yet so tightly defined by DFS sites. Tom Brady is a quarterback, and only a quarterback, in real football. The same is true in fantasy football. LeBron James is sometimes a point guard, sometimes a small forward, or sometimes a power forward. However, on DFS sites, heâ€™s only a small forward.
DvP definitely has value because of this â€“- if youâ€™re going to tightly define a playerâ€™s position, then you might as well have the statistic (DvP) measure defense versus position. But itâ€™s not the whole puzzle piece. Knowing how the Milwaukee Bucks defend the shooting guard position is helpful, and it brings value to judge James Harden in that light. However, James Harden and Kyle Korver are very different players. The Bucks will likely defend them differently with different, unique defenders. Looking at those specific defenders might give us a better idea at how both Harden and Korver will perform.
You may be reading this article and think, â€œWow, this is a crazy dumb idea.â€ Well, you very well might be right. So let me know! Iâ€™m always looking to think about things in new ways and learn more about both basketball and DFS. This is merely an introduction to an idea. Hit me up and tell me your thoughts.
*Note: The wonderful nbasavant.com has this information via play-by-play data. However, it's not sortable by defender on the site, so you'll have to manually go through and search and organize by defender. Hence my wish to make a database!