One Factor Blindness: How getting too excited about a factor can cause you to make a bad pick
I read a lot of DFS expert picks everyday I do fantasy, and sometimes you see experts have an argument they feel is overwhelming.
“Lebron is playing against Durant on national TV, HE IS GONNA GO OFF!”
“The Philadelphia matchup is just too good to not take Carmelo”
“Detroit is terrible against Centers, go with Al Jefferson”
There’s nothing wrong with this logic as an aspect of your decision making. But if you ignore the other factors that influence fantasy performance, chances are you will miss who the truly great picks of the night are. And you will ignore those other factors if you get too excited about matchup, DVP, or player motivation. I call this One Factor Blindness.
This week I decided to make a spreadsheet of projecting player performances off of DVP. After finishing the spreadsheet, I looked through the best projection of that night, hoping to find some really great picks. What I saw was quite surprising to me. Even in matchups against Philadelphia and LA, at the very most players were only projected to score 3-4 points more than their season average, even in their perfect matchup, while in good matchups players only had 1 point differences.
Using Fanduel salary, and assuming that 4.5 points per $1000 is hitting value, that means that a player will be undervalued at most $1000 according to DVP. That is a lot, but much less than I had imagined. When people think of matchups, they often think there will be no way in hell a player will do badly. But the reality is a good matchup will only cause a player to do slightly better than you expect, and a really great matchup could only reasonably project 5-6 fantasy points higher than their expected fantasy points in an average matchup. That’s quite good, but certainly not the sort of remarkable value you think you’re getting in a great matchup.
The truth is that sort of limit on how much DVP can influence your prediction on a players performance should be applied to all factors you are considering. If Anderson Varejao is out, Tristan Thompson probably becomes a better play. But thinking that you can project him to score more than a couple more points than if he was playing with Varejao would be overvaluing Varejao’s injury. Lebron playing against Kevin Durant on national TV should probably make us expect a better performance. But Kevin Durant is also the best defender at SF in the league, and that should not be ignored in our decision to play Lebron. Carmelo may be matched up against Philadelphia, but Carmelo’s fantasy performance is not correlated to DVP, which makes the matchup less of a slam dunk.
The real best picks tend to be the ones that have multiple, non overlapping factors in their favor. A few nights ago, Nick Calathes had a matchup against the Wizards. Calathes was starting in place of an injured Mike Conley, which caused him to have a drastic minutes increase which his salary had not caught up with. Calathes also had a good matchup against the Wizards who give up a lot of fantasy points to point guards. The icing on the cake was Calathes was also playing at home, where assists tend to be scored more liberally by stat keepers and where we can expect every player to do a little better than on the road. Calathes had tons of non-overlapping factors in his favor, and ended up being one of the best starts of the night.
In a game against Charlotte, Ty Lawson was out and Randy Foye got the start at point guard for Denver. Charlotte is not historically a good matchup for point guards, but with Kemba Walker out, Ramon Sessions, who is one of the worst defenders on the team, started at PG. Denver was also at home. Again, we saw a positive injury impact, matchup, and we had a player playing at home. All these factors pointed to a good game for Foye, who was the start of the night, breaking 50 fantasy points.
The moral of the story is there are many other factors to determine how well a player should perform in relation to his salary: Pricing algorithms, team injuries, emergence, minutes expectations, and many other factors, probably a lot I don’t yet know or understand. And they should all be consider to make a truly good decision regarding who the best picks are. When you limit your scope to one factor, and become blind to the others, you often make bad picks that could have been avoided if you weren’t too charmed by that single factor.View all posts by Daniel Steinberg