Using Closers in Daily Fantasy Contests

On Mondays I write about daily fantasy strategy. On Wednesdays I’m supposed to write about closers. To be frank, I enjoy the topic I have been assigned for Mondays much, much more. I despise chasing saves (as I discussed a couple of weeks ago), and the guys doing the Bullpen Report really have relievers covered. But a commenter on one of my recent daily fantasy strategy posts asked me a question that involved both daily fantasy strategy and closers. The commenter wanted to know how viable it is to use closers in daily contests.

Closers are usually the cheapest players available in daily contests. In today’s contests on Draftstreet, the 34 cheapest players are relievers and only 2 relievers have a price tag above $3,000 when the average salary of all non-relievers is just over $6,000. The obvious reason is that relievers are the only players that we don’t know whether they’ll play before their game starts. That uncertainty drives the price down and is why I’ve ignored using closers to this point. But I may have been doing it wrong.

When I initially started digging into some closer usage research, my plan was to see what the probability of a closer pitching on a given day was based on long it had been since he last appeared in a game. This proved to be too difficult, but I was able to determine how often closers appear in a game and, from there, what kind of value they have in daily contests.

The closers in my sample (guys who led their team in saves last year and have a closer’s job right now) appeared in 36.8% of their team’s games last year. Assuming we knew nothing else about the likelihood of a closer pitching on a given day, I believe this one piece of information gives us enough to believe that closers are worth using.

I took stats for all relievers last year and ran them through the Draftstreet scoring system to determine how many points each closer scored for the year. When I broke that down to a per appearance basis, closers were contributing anywhere from around two fantasy points per appearance on the low end to over four fantasy points for the elite closers.

Let’s use Craig Kimbrel as an example. Kimbrel scored 291 fantasy points on Draftstreet last year, which breaks down to 4.28 points per appearance. When you take the 4.28 and cut it down by the 36.8% chance that a closer will pitch on a given day, you have an expectancy that Kimbrel will score 1.57 fantasy points on that day. If you knew nothing else and were looking at Kimbrel’s cost of $3,037 for tomorrow, his cost per projected point would be $1,928. That’s excellent value.

I run the Steamer RoS projections for starting pitchers through the Draftstreet scoring system each day and break them down on a per game basis to give me a baseline projection for what the pitcher will do that day. I adjust the projection for a few variables and then compare it to cost to get an assessment of the pitcher’s value on that day. The lower the cost per projected point the better. There are very few starting pitchers each day who have a cost per projected point under $2,000. Yesterday there were only six starters below that threshhold on a day with a full slate of games. All of those starters were worth considering because of their value, and it doesn’t make sense for me to ignore closers who have similar value.

The nice thing is that the 36.8% chance of a closer pitching isn’t all we know. We can adjust that number based on the circumstances of the day. If the closer’s team is playing a team they are favored to beat, the odds the closer gets a save that day go up a bit. If his team is likely to lose, that percentage will tick down a bit. If the closer has not pitched in several days, the probability he pitches that day increases. The question is how much does it increase, and I wish I was capable of answering that question. Likewise, if he pitched the day before, he’s less likely to pitch that day. We can improve on that 36.8%, and thus we can determine when the cost per projected point for a closer is even lower.

The uncertainty is not a prohibitive factor. But because it is a factor, the use of closers is probably best left to high upside contests. By that I mean large contests where you have the chance to increase your entry fee by large multipliers and not double up or head to head contests.

The research I did for this piece is somewhat flawed, and there are several ways this topic could be taken further. If you have any suggestions or thoughts on how to improve upon this idea, please let me know.