As you are probably aware, Fantasy Alarm has unveiled a new tool for the daily fantasy baseball player. It converts daily player projections into fantasy points customized per site then divides the player salary by those points to get a bang-for-the-buck calculation. The smaller the number, the better the value. There’s a bit more to it than that but we’ll save that for the DFS Strategy columns. Today we’ll focus on the projection element. Why? Because they’re mine!

Everything begins with a neutral-park projection. For the daily game, the projection is normalized to per plate appearance or per inning pitched. The baseline projections will be updated over the course of the season. The best way to think about this is as a weighted average between expectations and current level of performance with the weight leaning history early in the season then sliding more current as the season progresses. The extent of the slide is dictated by when the different skills stabilize (which will be discussed in future columns).

HOME vs AWAY

Strictly from a skills basis, player’s performance is better at home so the skills based stats are adjusted accordingly – better at home, worse on the road. In addition, a park factor is applied as appropriate. This is done for both hitters and pitchers.

HANDEDNESS

This will be a topic for an upcoming DFS Strategy column, but handedness is an integral consideration for lineup construction. For hitters with an established baseline, their career splits versus left-handed pitchers and right-handed pitchers are applied. For everyone else, the league-average is used. In general, the greatest advantage is enjoyed by a lefty swinger facing a righty thrower followed by a right-handed batter against a southpaw. The worst scenario for a hitter is lefty on lefty.

No handedness adjustments are made for pitchers.

QUALITY OF OPPOSITION

This is the tough one. With respect to gauging the quality of the team a pitcher is facing, there isn’t enough of a sample to consider a current performance to be real yet it’s a different squad than last season so using that as a guide isn’t perfect. Until the sample is such that the current team’s skills are representative of their true level, the individual player projections themselves are used to measure the quality of opposition. While it may seem intuitive, in general, pitcher performance is elevated against weaker teams and vice versa. Expected strikeouts and homers allowed are also adjusted based on the opponent’s propensity to whiff and go yard. The probability of a win (and save for a closer) is also influenced by the opponent.

With regards to hitters, the skills of the batter are tweaked based on the pitcher. For example, if a batter is facing a pitcher with a strikeout rate better than league average, the projected strikeouts are increased accordingly. The influence of the pitcher is prorated to match the number of innings he’s expected to throw.

PLAYING TIME

For hitters, the number of plate appearances is projected using their expected lineup spot that evening. The per plate appearance stats are multiplied by the projected number of plate appearance to render the daily projection. Unfortunately, this needs to be posted before the actual lineups are announced so there’s a chance the spot in the order is different. The lineups are adjusted in the template on a daily basis so hopefully the instances this occurs are minimal.

For pitchers, an innings expectation based on history and opponent is used for starters while closers are all projected for an inning. Like hitters, the per innings projection is multiplied by the projected innings to yield the daily projection.

THE MAGIC OF EXCEL

All of the above is programmed into a spreadsheet. All I have to do is type in the names of the starting pitchers and everything flows from that. Well, the park neutral projections need to be kept up to date as do the teams’ batting order versus lefties and righties, but the automation to generate the projection avails ample time to knock off those tasks.

CONVERTING TO FANTASY POINTS and DOLLAR per POINT

Each daily site has its own scoring. Our tool converts the daily projections to projected points based on the scoring of the most popular sites. The tool allows sorting by points.

The tool also pulls the player’s salary for that day from the various sites. It calculates the dollar per point for each site. The smaller the number, the better. You can also sort by this. Future columns will detail the best use of these values.

The tool allows filtering by position as well as by daily site.

SUMMARY

Here’s a bullet point step-by-step review of the entire process

  • A current park neutral projection for hitters and pitchers is generated
  • The projection is normalized to per PA or per IP
  • Hitters are adjusted for home/away, park and opposing pitcher. Pitchers are adjusted for home/away, park and opposing lineup.
  • Projected PA and IP are determined and used to generate a daily projection
  • This daily projection is converted to expected fantasy points per site
  • Based on the salary for that day, the expected dollar/point is calculated
  • You dominate

As mentioned, I’ll be dedicating a lot of future bandwidth explaining the most efficient means of using the new Fantasy Alarm tool. Obviously, the lower the dollar per point the better, but setting an optimal lineup goes beyond simply choosing the best value at each position.

I’m happy to address any inquiries regarding my projection process and the subsequent conversion to daily fantasy points. Good luck and I hope to see you in the money.