PACKHAM'S PUNCH: STARS/SCRUBS VS BALANCED - RESULTS ARE IN!!!

 

 

Hello Jurors.  So many of you will remember that I wrote an article nearly 5 weeks ago regarding performing a scientific experiment to determine whether the best type of lineup was going stars and scrubs or balanced.  I will somewhat recap how I performed this experiment and then spend the majority of this article giving you the results and how we can interpret that into our daily DFS lineup decisions.

 

 

 

 “The Process”

 

When conducting an experiment, you need to take out variance as much as possible. 

So… for each slate, I chose 5 “core” players priced between $4500-$7500 dk dollars to leave around $20K DK dollars for the last 3 spots.  These 5 players were used in both lineups and they were the controls. (or control group per se) The remainder of each lineup provided the variables which formed the experimental groups from the scientific method.

Examples from day of the experiment… Thursday November 30th Lineups:

 

Core Players (31.3K)

1.  L Williams 7.3K

2.  R Rubio 5.7K

3.  D Saric 4.9K (can you believe he was once this cheap?)

4.  A Horford 6.8K

5.  D Favors 6.6K (can you believe he was this expensive?)

 

Studs/Duds (18.3)                             Balanced (18.6)

6. Giannis 11.1K                                 6. G Harris 5.6K

7. A Johnson 3.8K                              7. Bledsoe 6.8K

8. Bayless 3.4K                                   8. R Covington 6.2K

 

I used players that I personally felt were the best plays of the day from a “cash” perspective.  Even though I also entered a single-entry tournament as well for data collection purposes.

So… I set my plan into action and recorded data everyday for a month.  What were my results?

 

 

 

The Results

 

The data that I tracked was points per dollar using 5x to equal 5 points per every $1000 spent.  Here were the results.

Over 30 slates:

·         Core: 5.38 pts/$1000 (I will drop the 1000 from here out)

·         Studs: 5.30 pts/$

·         Balanced 5.44 pts/$

The average Cash line for a 50K lineup on dk

·         229-entry double up: 5.37 pts/$ or 268.5 pts

·         Over 1000-entry $1Single Entry gpp cash line:  5.49 pts/$ or 274.5 pts.

Commentary:  Obviously this is an average and relative small sample size, but both the core and the balanced lineups reached the cash line on average and the Star/scrub did not.  None of them averaged reaching the cash line for gpp’s.

I next looked at Small vs Large vs Middle-sized slates

 

On smaller slates

·         Core: 5.49

·         Stars/Scrubs: 5.49

·         Balanced: 5.23

 

On Mid-sized Slates

·         Core: 5.25

·         Stars: 5.33

·         Balanced:  5.88

 

On Large Slates

·         Core: 5.38

·         Stars: 5.32

·         Balanced 5.16

 

Discussion:  The only statistically significant findings on this part of the study was that a balanced lineup seems to really work out well on Mid-sized slates.  Yet seems to really struggle on either larger or smaller slates.  Otherwise there did not seem to be much statistical significance in the study of slate size.

 

Overall Discussion:  I’m not sure that my study came to any significant Earth-Shattering conclusions.  Obviously, as a DFS player you need to pick the correct Star, and the right general core, etc…  However, I felt that more often then not on a slate from 5-9 games it seems best to go with a Balanced lineup.  Furthermore, on any slate with 4 or less games… or on any slate over 9+ games, this study suggests you would be better off putting a Stars/Scrubs lineup into a cash game contest.

Also, previously, I have been looking at 5x as my magical “value” number… however this study shows that a corrected value of somewhere between 5.4-5.6 is the metric we need to shoot for to achieve both double up and Single Entry cash lines. 

I would be interested in studying other aspects of DFS.  Please reach out to me in slack and we can come up with some other ideas to help us each achieve glory.

 

 

You can reach Craig on twitter @cpackham007 or on the slack chat.  You can also email him at cpackham007@yahoo.com.  He’d love to hear your takes as well.