You are in your draft and are choosing between two players for your third pick, with the following profiles:
–Player A will generate exactly third round value 25% of the time and the remainder will be below third round value;
–Player B will generate above-third round value 25% of the time, will never be worth exactly third-round value and will be below third-round value the remainder of the time.
Who do you pick?
If you consider this example to be too abstract then lets put some names to the example. Using MockDraftCentral.com and looking at the third round, we can roughly call Freddie Freeman a Player A “type” and Matt Kemp, drafted one spot below him, a Player B “type.” I am calling them “types” with the hope of avoiding a specific debate on Freeman and Kemp.
How about this one: you are looking at the fifth overall pick. You are deciding between Chris Davis and Clayton Kershaw. In this example, Davis is a Player A type and Kershaw is a Player B type, as compared to the value of the fifth overall pick.
Who do you pick?
In both cases there is an obvious answer, but most fantasy players will make the wrong selection.
One of the basic ideas generated from game theory is that of a “dominant strategy.” When a game player is faced with two competing strategies, one of which does no worse than the other but sometimes does better, then the game player always should choose the former. It is called a “dominant” strategy as compared to the second.
Virtually everyone in fantasy baseball, from neophyte to “expert,” routinely and continuously makes the mistake of violating this rule. Every person who either picks Chris Davis or recommends that you pick Chris Davis in the above example violates this rule. And it is a mathematical mistake 100% of the time and can easily be proven to be so; I will spare you the math but if you want a better explanation feel free to tweet me @pdicaprioFP911 and I will be happy to oblige.
Almost everything you read on drafts uses the ridiculous idea that draft value exists a priori and that you should rate your draft selection on comparing ADP to when you make your individual selection. Though it has such gaping, obvious flaws that show how silly this notion is, fantasy players follow it unquestioningly, mostly because it came from the Fantasy Ten Commandments back when we knew nothing of ideas of game theory and behavioral economics, and no one even considered ideas like basic statistical probability.
The mistake is comparing ADP to the draft pick. What you should be comparing is the value of the draft pick spot to the player’s value to see if passing on him is a dominant strategy.
Going back to the Chris Davis/Clayton Kershaw example, Clayton Kershaw has been worth $36, $35 and $41 in the last three years. Chris Davis was a $36 player last year and was never close in any other season. So, unless you think Chris Davis will be worth more this year then last year the “picking Clayton Kershaw” is a dominant strategy over “picking Chris Davis” and it is a mistake 100% of the time that you choose a dominated strategy over a dominant one.
Even a quick scan of the ADP reports shows many similar examples. How about Adam Jones at 9 overall? Jones has been worth $32 at best and never worth more than $30 other than last year. unless you think Adam Jones may be worth More than $35 this year (and maybe you do) then “passing on Adam Jones” dominates “picking Adam Jones.”
Take any list of translations of round values converted to dollars. Then take your dollar value projection. If the player you are considering at a certain pick has never exceeded that level, then unless you project improvement passing on that player is a dominant strategy over picking that player. And it is an error every single time you pick the player unless you project improvement.
Of course, you may not want to do this type of math. And you needn’t do it just to take advantage of it. Just get as close as you can; the easy way to do this is to refuse to select a player who only has downside and no upside at the point where you are picking him. Mostly this will negate your early round possibilities. But every time you don’t draft a player with only downside you are virtually guaranteeing that despite not putting in the full analytical work you will not be committing to a dominated strategy.
If a player, like Mike Trout for example, has at least once generated a dollar value that justifies his selection at a given pick, then your selection is not a dominated strategy. But if your selection must improve off his career best to be worth the selection you are using a dominated strategy. That is not to say he will not improve; but that if you have to project improvement to be worth the pick then you are making a mistake picking the guy, and are making the mistake no matter what tortured mental gymnastics you think of to justify the selection.
Another example of a dominant strategy: “using the idea of dominant strategies” dominates “using the traditional ADP-based notion of value.”
Latest posts by Patrick DiCaprio (Posts)
- Closer Identifier Algorithm Results June 23, 2014 - Jun 23, 2014
- 2014 Fantasy Baseball Pitch or Ditch for Sunday June 22, 2014 - Jun 21, 2014
- Closer Identifier Algorithm Results June 16, 2014 - Jun 16, 2014