Dynamic Pricing Expectations


You may have noticed in our datasheet we have now use a dynamic system when calculating projected +-. This is a higher level strategy that was brought to my attention by Facebook u/JackSavage. Shout out to him for much of the following insight.

The rationale behind dynamic the approach is that, simply put, a 5X system expects too little from your lower salaried players and too much from your high salaried players. Thinking about it another way, for the very few times a season a 10K+ player hits 70 points (thus 7Xing value) there are dozens of times cheaper players manage to 7X, or even 10X value.

What is really comes down to is this, if your goal is to win a GPP contest, a 3K player 6Xing value isn't going to cut it, whereas a 10K player 6Xing could very well be a key piece in a GPP winning lineup. A highly priced player returning 5X on his value is MUCH more valuable to you than a 3K player returning 5X on his value. But of course, this doesn't change the way an optimizer creates an 'optimal' lineup, regardless of our pricing expectations, our optimizer is going to find the maximum point value.

So, how should you change your lineup construction. 

1) Avoid cramming cheap players into your lineup just because you think they might 5X value. If you don't see a legitimate chance for a cheaper player to 6-7X value, avoid them entirely. 

2) Pick stars you feel confident can return 5-6X. Think about it this way, if you find 3 stars who return value at the top end, you're already halfway to cashing with 5 roster spots left to go. 

3) Use our dynamic projected +- values from the datasheet. These give you a perfect sense of who is really creating value based on exactly what we discussed above. If you can create a lineup of players with high projected +- values, you should be in a strong position to succeed.

Our current system uses a 4.6X + 6.25 system to evaluate DraftKings player point expectations, and a 4.4X + 6 system to evaluate Fanduel player point expectations.