UFC 274 DFS Mass Multi-Entry Review With Backtesting

UFC DFS Mass Multi-Entry Backtesting

UFC 274 DFS Mass Multi-Entry Review With Backtesting

Unfortunately, this week was not profitable for DraftKings DFS Mass Multi-Entry. However, money was won back, so it was not a complete loss. Although money was not won, there is still a lesson to be learned. Therefore, this brings us to the purpose of this article which is to review this week’s Backtesting approaches to inform better our UFC DFS practices moving forward.

What is Backtesting?

Go to the Domination Station; you will see three tabs; a tab should be titled Backtest. This is where you backtest, as it will allow you to collect data on various outcomes regarding the slate. Furthermore, one can manipulate the Domination Station to find systems that work in lineup construction.

For example, say you wanted to see the baseline average projection for those lower-priced players who end up in the nuts. You could backtest multiple slates for Nuts scores and then find the average score of those lower-priced players. This is just one example of using Backtesting to prepare yourself for success in Mass Multi-Entry tournaments. 

For more information on Backtesting, see the “NASCAR, PGA, UFC DFS: MME Optimizer Setup” article, as it will provide greater context to the work of Backtesting.

Starting with 300 Lineups

I mentioned in previous articles that I am a coach and educator as my primary career; therefore, like any passionate teacher, I believe in progressions as a key to building skills regardless of what you are learning. Likewise, 300 lineups were generated in the first Backtesting approach as this would have the possible outcomes, making it easier to find hit the nuts.

Nuts Score: 619.65

300 Lineups Trial #1

The DS was set up to include:

• 100% Max Exposure
• Lock Oliveira
• 25% Randomization
• two unique
• 0 fades

In the first run, I attempted to do as little as possible to see how well embracing variance and seeing where locking a stud (Oliveira), then shaking up the rest would get us in terms of payout. The results were surprisingly strong for such little manipulation done to the Domination Station.

A base score of 569.86 created a clearer picture of what the next steps would be to get closer to the nuts score of 619.65.

300 Lineups Trial #2

In addition to a decent base score, we also were able to have a framework to work off of for trial 2. Again, I was trying to do as little as possible. Therefore we adjusted by cutting out four fade-able fighters.

4 Fighters Faded:

1. Roberts 8000
2. Trinaldo 8200
3. Vancamp 6700
4. Rua 7000

By making this minor adjustment, the top score went from 569.96 to 582.87. While this is promising, we are still building 300 lineups. Therefore the goal is to hit that nuts score of 619.5. With that being said, we had to make one more set of adjustments to get to that nuts score of 300.

300 Lineups Trial #3

Finally, we hit the nuts on the third trial (being sarcastic, this is pretty quick for Backtesting)!!

What did I change, you ask??

The only thing I changed was I added one top price fighter I wanted to fade, and I followed through with it, bye-bye Dumont!

By fading one more fighter and being intentional with taking advantage of the pricing, room to generate different outcomes with other top-priced fighters was able to happen.

Why Fade Dumont?

Because the research pointed to Dumont being the possible fade with her low output compared to her price tag, we were able to make an educated decision leading to a nuts score. Specifically, we saw that Dumont was priced at 9100 w/ only a projection of 68 Ftpts as the favorite. Additionally, it was found that Dumont had not scored over 100 Ftpts in the past three years and did not hold credible wins as of late.

In conclusion, because of her data surrounding her and looking at her competition, Dumont was a clear-cut fade in terms of finding that high-priced stinker, so we did it!

Final protocol used for building 300 lineups and hitting the nuts:

• 300 Lineups
• 25% Randomization 
• Ownership Cap 100%
• Five fades
  • One 9k Fighter 
  • Two 8k fighters
  • One 7k fighter
  • One 6k fighter 
• Lock the stud (Oliveira) 
• Two unique Players
• Optimal Mode 

The nuts score of 619.65 was achieved using this approach without changing exposures, randomization, or uniques.

Transferring to 150 Lineups

We don’t want to go straight from 300 to 60. This is a drastic difference in the number of lineups created, causing a theoretical difference in the number of players in the pool for the domination station. As a result of this theoretical perspective regarding the number of lineups being used and progressions, our next step in the study is going down to 150 lineups.

150 Lineups Trial #1

For the first run, the same outline was kept for the 300 approaches to Backtesting as there was interest in seeing the transfer from 300 to 150. This outcome was a score of 583.5, which is not bad, but left room for improvement, which leads us to trial #2.

Check out this crazy knockout from UFC274!

150 Lineups Trial #2

At this time, we have faded five fighters:

1. Roberts 8000
2. Trinaldo 8200
3. Vancamp 6700
4. Rua 7000
5. Dumont 9100

To better narrow the focus, we faded two more fighters, and this time we looked at our lock to make the first fade. Locking Olivereira tells the confidence in Justin Gathje to put up an intense fight and fantasy performance. Therefore, our first additional fade was Justin Gaethje, and we hedged it off with a Garcia fade, bringing us to a total of 7 fades.

Fade List

1. Roberts 8000
2. Trinaldo 8200
3. Vancamp 6700
4. Rua 7000
5. Dumont 9100
6. Gaethje 7700
7. Garcia 8600

Outcome

By adding 2 to the fade list and not changing anything else, we hit the nuts score of 619.65 when generating 150 lineups. Below is the complete outline:

• 150 Lineups
• 25% Randomization 
• Ownership Cap 100%
• Seven fades
  • Roberts 8000
  • Trinaldo 8200
  • Vancamp 6700
  • Rua 7000
  • Dumont 9100
  • Gaethje 7700
  • Garcia 8600
• Lock the stud (Oliveira) 
• Two unique Players
• Optimal Mode 

60 Lineups Trial #1

Following the 150 lineups, the final progression was ready to be attacked in preparing 60 lineups. Everything was kept the same as above, and the score was 616.26. That is less than three points from the nuts! The excitement was so real that we went right into trial two and started the tinkering.

60 Lineups Trial #2

After testing the slate repeatedly to get those three points, it was found that manipulating the ownership max can lead to a nuts score. Therefore, in conclusion, we added a rule that all fighters that have low upside according to the research would be capped at 20%, which ended up being six fighters total. Additionally, everyone else left in the pool was set to 40% while keeping Oliveira as the lock.

By making these changes, the nuts score was achieved with 60 lineups. Below the entire outline is provided:

• 150 Lineups
• 25% Randomization 
• Seven fades
  • Roberts 8000
  • Trinaldo 8200
  • Vancamp 6700
  • Rua 7000
  • Dumont 9100
  • Gaethje 7700
  • Garcia 8600
• Lock the stud (Oliveira) 
• Two unique Players
• Optimal Mode 
• Low Confidence fighters capped at 20% (6 soldiers total) 
• All players outside of the Lock and Low confidence capped at 40%

Conclusion

This article hopes to guide readers through the process of reviewing previous performances. This is not very sexy or fun sounding, but it is the meat and potatoes of hitting the nuts and winning money in Multi-Entry contests. As you work through reviewing, please feel free to manipulate my templates how they work for you, as this is simply a tool to help get you started or provide you with a changeup if you are already doing the work of Backtesting.

Happy hunting, everyone, and thank you for reading!!!

Thanks for hanging in, and feel free to reach out with any questions!

Talk to me! I am in the discord 24/7 and value connecting with people.

Twitter: @therealgc9

Pricing for DFS Army VIP
Coupon Code for 10% off DFLGENO

UFC DFS Mass Multi-Entry Backtesting