Blast from the 2015 PAST. "How do you" ? Series Issue #5 Does more information increase accuracy in fantasy football?
I am reposting my thoughts from 2015
Does more information increase accuracy in fantasy football?The data suggest that it does not.
Hall et al published in this article a discussion on this concept. The illusion of knowledge: When more information reduces accuracy and increases confidence. (2007) Crystal C. Hall, Lynn Ariss, Alexander Todorov, Organizational Behavior and Human Decision Processes 103 : 277–290.
Their abstract lays out using NBA predictions and sports fans. They setup the experiments with Fans and they were given some stats on the teams. One group had no team names and the other group had the team names. The Team names gave a false confidence to winning picks. What happen was the group relied less on the stats because they were “familiar” with the team and “knew” that team. More familiar teams caused a failure in “using stats etc.” objective data to make bets and picks. They group that had the team names had a “false” belief in the strength of their confidence.
So in fantasy football, the issue is because we “know” the “worth” of a player because of whatever (Fill in the reason), do we ignore the objective data?
Drew Brees is a first ballot hall of famer so he is always the best. Losing Jimmy Graham and Kenny Stills does not matter.
Our confidence is high and we focus elsewhere in our draft If we have 20 pieces of information, then I am really knowledgeable about that situation?
Cueing specific knowledge can increase the sense of expertise!
Paul Slovic, "Behavioral Problems of Adhering to a Decision Policy," unpublished manuscript, 1973.
The quotes and figure below illustrates the issue. Quoting “Eight experienced horse-race handicappers were shown a list of 88 variables. Each handicapper was given the data in increments of the 5, 10, 20 and 40 variables he had judged to be most useful. They predicted each race four times--once with each of the four different levels of information. For each prediction, each handicapper assigned a value from 0 to 100 percent to indicate degree of confidence in the accuracy of his prediction.
The average accuracy of predictions remained the same regardless of how much information the handicappers had available.
Three of the handicappers actually showed less accuracy as the amount of information increased, two improved their accuracy, and three were unchanged.
All, however, expressed steadily increasing confidence in their judgments as more information was received. This relationship between amounts of information, accuracy of the handicappers' prediction of the first place winners, and the handicappers' confidence in their predictions is shown in Figure 5. Above”
This discussion was found in Chapter 5 Psychology of Intelligence Analysis (1999) by Richards J. Heuer, Jr. “Do You Really Need More Information? “
Point 1. If you have an artificial sense of confidence then mistakes will happen.
Point 2. More information past a small amount will not add to your accuracy in drafts but will elevate your sense of confidence.
See Point 1.
Point 3. The time spend searching for too much information is wasted except for injuries or some black swan event. Looking at 20 websites for your rankings is wasted.
Point 4. So what is the amount of information you (YOU) need for winning?
I suggest you go back and read the multiple hypothesis blog post for some insight.
Point 4A. Keep records to record your journey. So go back and ask yourself the whys for your mock picks. This is a way to see what information was driving your decisions.
Points 4B. What are successful players doing? In your leagues the top group may be doing things that you do not. Why are they making their decisions?
Point 4C Make a decision flow chart for each of your real drafts and save for next year to look back at. Be reflective in short.
Point 5. There are three types of forecasting techniques:
Point 5A. Subjective or qualitative where you rely on expert judgments
Point 5B. Time-Series Projections where you use quantifiable observations over time. – Record Keeping!
Point 5C. Casual Models where you emphasize causal/correlational relationships. The principal emphasis in forecasting is looking for patterns and fluctuations over time. Little tidbits.
Here is an example. “When the 1QB was hurt and the backup QB in place sometimes the WR 3 4 or 5 is elevated because they have spend time with the backup more than the QB1 in that team. So before the season determine if such relationships exist.” Who is on the practice team?
Point 6. Three things to do.
Point 6A. List all your assumptions about your player, team, defense strength etc.
Point 6B. Reverse each assumption. What is its opposite? Write down
Point 6C. Ask yourself how to deal with each reversal assumption in a draft or waiver wire etc.
If you work on these things you can grade yourself in the off season and do reflections on your journey.
We are use to being reflective given our careers as a University Professors. We are used to delayed gratification for success. So it is maybe easier for us.
If you hate to write, then use a voice recorder and document your drafts if you can. There is software that records your screen in a video/audio. Use that while drafting!
Imagine finding out you are really are sucky at picking late round sleepers but great at WR3 selections. That gives you an area to focus on.