Tracking Your Results on DraftKings
We’re all in the business of making predictions. The more accurate the predictions, the more money we make.
I’m of the belief that the single-best way to improve at daily fantasy sports—the only way, really—is to treat it as a scientific endeavor, testing different theories and strategies and making small but meaningful alterations. Science is marked by making predictions that are not only testable, but also falsifiable. That’s the difference between scientists and astrologists; scientific theories can and do get falsified all the time, with the best left standing.
Science is always evolving, as are the best DFS players, but that’s not possible without understanding where we went wrong in the past. We hear about ‘good process’ a lot, but how is it possible to know which processes are good and bad without evaluating past decisions?
As a side note, this is why I tend to analyze solely evergreen daily fantasy topics as opposed to time-sensitive ones. There’s no substitute for weekly analysis and certainly your choices should be made based on all of the evidence available at the time, but it’s really difficult to look at an individual decision and know if it was good or bad.
Should I have started Sammy Watkins last week? Well, looking at how well Watkins performed is going to tell us next to nothing about the quality of that choice; there’s just too much variance. But it wouldn’t be worthless to analyze rookie wide receivers of similar quality to Watkins and see how they performed against comparable defenses. That has value and allows us to more quickly identify ‘good process.’
With all of that in mind, the value of assessing your playing history is obvious. Here’s how to do it on DraftKings.
Exporting Results
You can export your entire league history on DraftKings very easily. Go to the ‘My Contests’ page and click the ‘History’ button. You’ll see a ‘Download Entry History’ link.
If you click that, you’ll have access to an Excel spreadsheet with all of your past leagues, sorted by sport, data, place finished, points, winnings, and entry fee.
Average Scores
The first thing that users tend to want to uncover is their average scores. I think it’s fun to analyze average scores, but it’s also pretty irrelevant. The scores needed to win a league change all the time (based on starters, salary cap flexibility, actual game results, and so on). Think about how much DraftKings’ scores have changed with the additional wide receiver slot this year, for example, or the loosening of the salary cap.
If you want to analyze how your average scores change over the course of an entire NBA or MLB season, I think that has some merit. But there are better ways to assess your progress as a daily fantasy player.
Win Rates/Profit
Even though you obviously need to score a lot of points to win over the long run, the goal of daily fantasy sports is not to score as many points as possible. It’s to win as much money as you can, which is not synonymous. I’ve discussed this issue a bunch in the past and won’t get into it in great detail here, but there’s a major difference between maximizing projected points and maximizing your win probability, particularly in tournaments.
Evaluating win rates/profit is easiest in cash games. You can quickly build up a significant game log by playing head-to-head and 50/50 games every day in pretty much every sport except football (football is a totally different beast and one that’s really difficult to analyze because of the small sample of games).
I particularly love 50/50s as a gauge of player quality because you face a massive number of opponents. If you play NBA or MLB, enter some 50/50 leagues (even if they’re just $1) solely as a way to evaluate your lineup strength. After a couple months, you’ll have exposure to an immense number of lineups. You can then assess not only how often you finish in the top half of entrants, but also your typical finish overall; if you regularly finish in the bottom 40 percent of lineups, that’s a sign you need to change some things around.
Distribution of Scores
I mentioned that assessing win rates and even profit is difficult in GPPs. That’s because, due to the top-heavy payout structure of all tournaments, GPP profits tend to be characterized by a punctuated equilibrium—long periods of no growth or a small decline interrupted by huge cashes.
That’s the main reason that I think you need to analyze your GPP results differently than your cash game history. One way to do that is to evaluate the distribution of scores in each league type. Since there’s no value in finishing in, say, the 70th percentile in a GPP (whereas that would be profitable in a 50/50), your goal should be hitting a home run.
In a previous article, I posted a sample probability of reaching certain scores on DraftKings with different strategies. If you’re approaching tournaments and cash games correctly, your scores should be distributed in a manner similar to this…
A safe, low-variance approach is going to limit your upside and your downside, giving you more exposure to moderate scores—a narrow range of potential outcomes. A risk-seeking, high-variance approach opens up the range of potential scores—more access to elite performances, but also more downside.
If you’re properly structuring your lineups according to the league type, you should see more top-tier lineups and more horrible lineups in your tournaments. If you see that your distribution of scores is similar regardless of the leagues entered, that’s a sign that you’re approaching either cash games or GPPs (or both) inefficiently.
The goal in tournaments is to maximize your access to a high ceiling, which necessarily increases your downside. While bottom-of-the-barrel scores clearly suck, they could actually be a positive sign of your tournament prowess if they’re accompanied by an increase in elite scores, too.