Looking at some of the issues that arise with projections.
My columns at Rototimes (sadly) are for subscribers only. My first column is on the theory of projections. It is my contention that slavish adherence to projections is not only flawed when dealing with the better players, but that the entire concept of “value” is meaningless. Here are the main points from the article:
There are various projection systems available. However, there is very little difference between the good ones. The only way to profit from a better projection is to profit from the marginal difference between systems A and B. If system A has a player at $15 and system B has him at $17 the accuracy of the projection is irrelevant and has virtually zero profit potential.
Here are some theoretical principles to keep in mind when dealing with projections:
1. Projected values are merely probabilities: When we say that a player is projected to be worth $20, what are we really saying? We are saying that 65% of the time he will be worth $17-23, 25% of the time he will be worth $14-26 and so on. Sadly, what we are trying to do is to reduce the irreducible. Do not fall into the trap of desiring a quick and easy way to say a player is worth “x” without understanding what that really means.
2. Projected values are merely symbolic: The philosopher Ludwig Wittgenstein posited that words are nothing more than symbols. The same is true of projected values. A projected value is nothing more than a shorthand symbolic way to refer to a player’s “worth” even though it may have little correlation to the actual worth of the player’s season. We are not trying to maximize “value” (defined as the excess between what we paid for a player and what he cost) but to maximize our team’s production. The implication of this idea is that…
3. The projected value of a player is a mirage: Once we shed the idea that a projected value has some necessary and sufficient relation to the actual value of a player, we shed ourselves of the biggest yoke that fantasy owners must carry. Projected values are riddled with grays and uncertainties that bewitch their users. As a result, slavish adherence to a plan that emphasizes value above all else is a faulty plan.
Often the fantasy owner will hear plans such as the 75% plan, whereby the owner buys players at 75% of their value. This may be laudable but simply will not work against tough competition if you are shooting for first place. Value is a necessary condition to victory, but not a sufficient one.
4. Often it is better to target specific players: Very often auction advice eschews the targeting of specific players as if all players are fungible. This is simply false. I can’t count how many times I have heard this hoary advice and yet it is fundamentally flawed. You should not target specific players when you are dealing with fungible commodities. However, the top of the auction pyramid contains non-fungible players. In these cases the best strategic plans consider that some players are worth acquiring regardless of “value.”
7 responses so far ↓
1 Chad // Feb 8, 2008 at 4:41 pm
Interesting points, but the problem here is that you don’t really pose a viable alternative. Lacking projections, how can you begin to (1) rank players, and (2) compare players across positions?
2 Patrick DiCaprio // Feb 8, 2008 at 4:45 pm
Chad thanks for reading. This is far too complex to discuss in a comment here. But the main point is that if you like a player by whatever criteria you use the fact that he is auctioned above your projection should be no impediment to acquiring that player. If you have Jake Peavy at $27 and he is at $29 it would be foolish to drop out MERELY because he is above your projected value.
3 Chad // Feb 8, 2008 at 4:53 pm
I agree with you there. Projections should only be estimates, particularly in an auction format where your projected values constantly fluctuate based on your roster and dollars remaining, your opponents’ rosters and dollars remaining, the remaining player pool, and so on.
I enjoy the site - keep up the good work.
4 Bob Moore // Feb 9, 2008 at 1:45 am
Patrick,
I am new to your site, but I have really enjoyed your thinking having gone back to read your articles. Thank you for sharing your thoughts and insights. As a 59 year old investment advisor, I never really considered myself to be a geek, but if that is what it means to match wits with other people in any endeavor, then a geek I am. Keep it up and thanks again for your entertaining and interesting perspectives!
Bob Moore
Phoenix, AZ
5 Patrick DiCaprio // Feb 9, 2008 at 1:59 am
Thanks for the kind words Bob!
6 Cobb // Feb 12, 2008 at 9:22 pm
I had read your thoughts on “value” before and had somewhat disagreed, but I think I just misunderstood.
Your example about Peavy makes sense. Drafting simply on the basis of value probably won’t result in the most productive team. Sometimes you have to pay a little more to get what you want/need.
Good stuff.
7 digglahhh // Feb 21, 2008 at 11:16 pm
Great post.
By all means, value doesn’t apply to the uberstuds. Since, we’re talking about dollars and auctions, it’s fitting to make a consumerist analogy. A-Rod is like a luxury item, “value” doesn’t apply to him like “value” doesn’t apply to a Dolce & Gabana evening gown.
There are points in the middle of a draft where you are dealing largely with interchangeable parts. There’ll be a set of players that fill a position you need, and you’ll be completely willing to say, I’ll take whichever one I can get for X dollars.
That’s like your mind set when you’re going shopping because you need some new sneakers to mow the lawn in. Shopping for A-Rod on the other hand is like a woman shopping for her wedding dress.
There’s certainly a point when overbidding becomes egregious, but the underlying point remains.
Leave a Comment