Man + Machine: FanDuel Disconnect

This article is part of our Man + Machine series.

For the most part, I've stuck with one double up or 50/50 entry and one tournament entry with the same lineup. The thought process being that if the 50/50 entries can mitigate the losses of the tournament entries then eventually a lineup will hit and provide extra wiggle room. Or better.

This is has been the case at DraftKings, but not at FanDuel.

Also, there is the reality that lineup assembly should likely be tilted towards the contests you're entering. As in, tournament entries likely need to feature more high-risk, high-reward plays. This is not something I've tried to accomplish.

Additionally, I've essentially abandoned the strict lineup assembly process I outlined last week. Specifically, I didn't use rigid locks and then build around them with the optimizer, which had shown the best results.

Finally, you'll notice a lot of early- and late-only entries, which is the preference for my personal time management. There are nights and afternoons where I'm not available for the due diligence required to build competitive lineups. This reality is also why lineups weren't tailored to both cash and tournament contests.

Before showcasing the embarrassing results, it is worth noting over the same five-day period I'm plus-5.5 units at Pinnacle Sports, where I also have a hobby account. In less than a month, my bankroll has grown just over 600 percent. So there is a huge disconnect somewhere between my handicapping and player selection, especially at FanDuel.

Results

There is one major difference and simple explanation for why the FanDuel results have been less successful. Each day, I've started with FanDuel, and I've spent more time creating the FanDuel lineups, sometimes it has been significantly more time. DraftKings always came second.

Additionally, with the previously mentioned time constraints often an issue, the DraftKings lineups have been more "thrown together," for lack of a better term.

For example, the optimizer loved Carlos Gomez on Wednesday, but as I built my FanDuel lineup, I switched to Adam Jones in that price point because he had the handedness advantage. I also had Manny Machado locked in. However, when building my DraftKings lineup, the Gomez fondness stuck out, and I made him a building block and left him without exploring other options or lineup combinations.

Continuing, part of the reason I locked in Gomez was that there were approximately 20 minutes left before the contests locked. I had to search for the Tigers' lineup and potentially make switches to both the DraftKings and FanDuel rosters, if needed. This happened to be the case, too, as Ian Kinsler was out.

Gomez went off, as the optimizer expected, and I had plenty of exposure to the Brewers in both lineups. It was just that the FanDuel lineup utilized Jean Segura, Scooter Gennett and Adam Lind, whereas the DraftKings lineup had Gerardo Parra, Khris Davis, Jonathan Lucroy, Gomez and Lind.

Reflections

While the results have been extremely lackluster, and somewhat embarrassing, there have been a few telling realizations to better the results going forward:

1. Looking back over the FanDuel lineups, in the four losers, there have been three home runs, four other extra-base hits and zero pitching wins. There is a degree of bad luck there, especially knowing how well I've handicapped the field of late. Sometimes it is just the wrong players from the right teams on the right day.

2. Pitching has been a horrible weakness and having two or three duds in the field won't kill a cash lineup. Having a dud on the mound will do it every night. This is nothing new or earth shattering. Being cute with pitcher selections should be reserved for lottery lineups. This is a human error, too.

3. At DraftKings, I'm buoying because the lineups have been assembled quicker and in a fashion closer to Step 3 outlined last week. My leans combined with players the optimizer liked. A hat tip to the optimizer for showing Joey Votto love against Max Scherzer on Tuesday, too.

4. A better lineup assembly strategy and process needs to be in place while using the optimizer. For example, I have an extended desktop, and usually I'm building the lineups half in the optimizer and half in FanDuel itself while going back and forth. It doesn't take long before there have been multiple tweaks and the entire lineup has undergone a facelift. And a lot of times, the sacrifices were the lineup building-blocks, as noted in the Gomez example.

I'm going to revert to beginning with my own lineup locks while simplifying the approach to creating lineups to avoid too much tinkering. The optimizer is a valuable tool, but all tools have rules and techniques for use. Their value is diminished when they're not used appropriately.

I suspect reverting back to being more rigid with how I incorporate the optimizer's selections will help. Over the past few days, there has been more randomness and looseness to the hindrance of effectiveness.

Continuing to enter double-ups and tournaments is the plan because eventually the disconnect will fade. The cash lineups should become more stable with the odd lineup returning a few extra dollars in tournaments.

Under the microscope of daily games, horrible runs are commonplace, albeit there are no guaranteed quick fixes, either. I noted the recent handicapping success being at odds with this daily games streak because both regressing towards one another is likely. Hopefully, though, the daily games just catch up.

Sometimes you don't have the right players on your roster when you hit submit, and you have to be okay with it. An error in the process of player selection is a different thing entirely, though.