FRESH TAKE WITH MG FRESH
Avoiding Value Traps
As the season gets into gear and team rotations are set, player mispricing, common at the start of a season, should begin to correct. Guys like Tim Hardway Jr. and what’s-his-name Sabonis won’t be lock and load players every night now that their price reflects their new role and production.
DFS players will have to look more and more for one-off value opportunities on a nightly basis.
Some nights, there will be “obvious” plays that every DFS provider, writer, and commentator highlight due to injuries, news, trades, etc.
I am not a fan of fading (not rostering) these types of chalk (expected high ownership) players just to differentiate lineups. Most of the time, there is a reason a player becomes chalk on a slate. If a player is likely to massively exceed value, being “different” may just mean losing. In those instances I tend to think of lineup building as having one fewer slot than normal. In other words, if I have a player as a lock to exceed value, even if they are chalk, I plug them in and focus on figuring out the other 8 roster spots.
However, sometimes a player becomes chalk for nonsensical reasons and there may be an opportunity to make money by thinking independently and ignoring the crowd.
A few examples of Value Trap scenarios include:
1) The High Usage Bench Player That Gets a Spot Start
We hear it all the time.
‘If player X just got more minutes, he’d be a DFS monster. Look how he is dominating off the bench.’
Most of the time this is reasonable logic. A ton of DFS production comes from (minutes played) * (usage). However, there is a certain kind of player whose value may not change as much as advertised due to a promotion to the starting lineup.
A. Are regularly playing heavy minutes off the bench (6th or 7th man types)
B. Have heavy usage/create their own shot
C. Handle the ball for the second unit
For these players the increase of minutes may be offset by a diminished role when playing with and against players of similar caliber.
Some examples from last year include:
Each of these guys spent at least part of the year in the starting lineup and part of it coming off the bench. The average production as measured by FanDuel points scored per contest did not change dramatically for these guys whether they were introduced before tip or not.
2) The Revenge Game
One of the most talked about DFS narratives is the “revenge game” when a player goes up against a former team, teammate, or coach. FanDuel and DraftKings pricing typically does not change significantly for “revenge games”, so if an increase in production can be expected, it would be a value opportunity.
In reality, it is hard to tell whether the extra motivation actually results in additional production. Below is a sample of these types of narratives from 2016, comparing a few high profile players’ performances in “revenge games” against their average production for the season.
As you can see, results are mixed in this extremely small sample. (I would like to do a more complete analysis at some point this year if I can find a dataset that captures each “revenge game”)
While effect on production seems to be mixed, ownership levels on DFS sites are almost always higher than average on high profile matchups like these. (Again, I am looking for a more complete dataset to show full effect.)
My personal preference is to ignore these types of narratives. If a matchup is a good one for a player in a “revenge game” I won’t fade them just to be different. However, I do not typically go out of my way to get a player in a lineup solely because it is a revenge game.
Obviously my limited analysis does not prove that revenge game narratives never work.
I am sure there are numerous counter examples of players having big nights against former teams which I am not capturing in this tiny sample. DFS players who believe this is an opportunity to get better than average production should by all means take full advantage of these opportunities.
However, if you believe revenge games drive higher ownership and no significant increase in production, then the natural conclusion would be to fade these types of players in most instances.
3) Misleading Defense vs. Position Stats
Defense vs. position can be misleading for a number of reasons. Being a fan of the NBA outside of DFS will help you avoid falling into these traps. The Judge does a great job on his podcast breaking down why DvP may overstate the favorability of matchups.
This is not to say DvP is not a useful stat. It is one of the key factors in evaluating matchups. Just make sure you think through why a team gives up a ton of fantasy points to a certain position before locking in opposing players.
For example, a team may be terrible at guarding traditional centers, but stretch 5s may not benefit to the same extent. If an opposing team is getting destroyed by centers that crash the offensive boards or play bully ball do not blindly assume the stretch 5 they play tonight is going to have a huge game.
A team that struggles to guard Andre Drummond may do just fine against Miles Turner or Kevin Love and vice versa. A good test would be to look at box scores of similar players to the guy in the ‘favorable’ matchup to see how they did. If Dwight Howard recently went for 20 and 20 against Boston that is a data point more relevant to Andre Drummond’s expected production against the Celtics than Kevin Love in the same matchup.
Also, keep in mind that teams with especially bad defenders will shift their better defenders to guard the better opposing players. Last year Boston routinely hid Isaiah Thomas on the worst offensive guard/wing on the other team and put Avery Bradley or Marcus Smart on the best offensive players. Detroit did the same with Reggie Jackson and KCP.
Evaluate the reason for DvP attractiveness and whether it will actually apply in a given matchup. Projection systems can routinely get this wrong and over emphasize matchup effects which may not be relevant in all situations.
Last week I wrote about the importance of capturing critical data points in your routine. Injury news, narrative evaluation, and DvP are absolutely things you should consider in lineup construction. Getting value plays right can be a huge component of winning consistently in DFS.
Just do not blindly trust the crowd when it comes to where the value lies.
Opportunities for outsized profits occur when you recognize times that conventional wisdom may be oversimplifying or emphasizing the wrong data points. By thinking for yourself and not falling for value traps you can avoid mistakes caused by group think and win at DFS.
Any other value trap scenarios that I missed? Leave a note in the comments.
You can follow Mike on Twitter @mgfresh