I use the Trust Tracker to publicly track the usefulness – or lack thereof – of the recommendations I made in my most recent Daily Games Cheat Sheet article.

For a more detailed introduction, check out my first and second posts here and here.

Players are measured according to their p/K (points per thousand dollars spent). Depending on their p/K and the approximate number of points needed to win a 50-50 league that night, players are rated as either "helpful", "competitive", or "harmful".

On a normal night with a full slate of games (roughly 10 or more), you will need to score between 240 - 250 points to win a 50-50 league. On those nights, a p/K of 4.8 or higher is helpful, a p/K between 4.0-4.8 will keep a roster competitive, and a p/K below 3.5 is harmful.

While there is always variation from contest to contest, the scores required to win each one on any given night are usually very closely grouped, regardless of entry cost. Last time I wrote the Daily Games Cheat Sheet, Thursday, January 1, the scores required to win most contests were within the benchmark 240-250 range.

In my New Year's day article, I recommended 9 players for use:

• Three players (33% of those recommended) recorded p/K values above 4.8.
• All three helpful players recorded p/K values that placed them among the top the at the position, and among the top 12 overall.
• Two players (22%) recorded p/K values between 4-4.8.
• Two players (22%) recorded p/K values below 3.5.
• The best recommendation was Jimmy Butler, who recorded a p/K of 6.0. He registered the 6th highest p/K in the league Thursday night, and the highest p/K among shooting guards.
• The worst recommendation was Ray McCallum. McCallum had played 17 minutes or more in five games straight, a streak that seemed likely to continue given the absence of Ramon Sessions and the quality of the opponent – the Timberwolves. The game ended up being a three-point squeaker, and McCollum scored only two points and 4.75 fantasy points in just 10 minutes, for a p/K of 1.6.

Overall, 56 percent of my recommendations were of competitive value or better, 33 percent were actively helpful, and 22 percent were damaging.

Since only two of the players I recommended cost less than the per-roster spot average, I needed to include at two players not mentioned on the team that I entered Thursday on DraftKings. The rest of the roster was built using only players highlighted in my Thursday article. I tallied 261.75 points for an overall p/K of 5.2.

Season total

Through 17 articles this season, I have recommended 205 players for use.

• 112 have been helpful (55%)
• 148 have been competitive or better (72%)
• 43 have been harmful (21%)