The Gut Check No. 414: Crank Scores II

Photo: USA Today Sports

This is my seventh season with Footballguys. One of the things I wrote about early on that I brought with me to the site was player consistency analysis. A popular feature that I included with my consistency analysis was the Crank Score. 

During those early years of my fantasy writing, I was looking for a way to use player consistency as a predictive tool. I put Crank Scores on ice after 2009 when tape analysis became a much larger realm of interest. However, I've often thought about reprising the feature because the data still offers a lot of worthwhile jumping off points for further study. 

This article will give a brief explanation about two types of Crank Scores I've developed and I will share by 2016 data for each of them in PPR formats with these starting lineup types: 

• 1 QB, 2 RB, 3 WR, 1 TE (Go here)
• 1 QB, 2 RB, 3 WR, 1 TE, 1 RB-WR Flex (This article)

Each table will include 240 players—20 rounds of data. If you like to do a lot of your own work, these tables should offer a worthwhile starting point. If you're new to Crank Scores, you'll get to see the data points that create the formulas. 

Those of you lacking the time or the desire to wade in the waters of consistency data, don't worry, my next post will be focused on tiered cheat sheets that use these Crank Scores as a starting point for rankings (early next week). In that article, I'll also explain the factors I included in my analysis to adjust my ranking of players. 

ABOUT CONSISTENCY AND CRANK SCORES

Consistency scores are a lot like "quality games." It's essentially the same thing, but with a finer grain of detail. I create performance tiers based on 12-team leagues and measure of the percentage of games over a given period of time (a series of weeks, a season, multiple seasons, etc.) that the player reaches these tiers.

Unlike some consistency analysis that is available on fantasy football sites, I do not use standard deviation because that data unnecessarily punishes players who perform well above the baseline figure. We don't want to punish players with huge weeks well above the baseline if that fluctuation is mostly on the high side. My performance tiers:

• Elite = The top two players (the best "No.1" starters at their position). 
• #1 = The top 12 players at their position (all "No.1" starters).   
• #2 = The 13th-24th players at their position (all "No.2" starters).
• #3 = The 25th-36th players at their position (all "No.3" starters).
• #4 = The 37th-48th players at their position (all "No.4" starters). 
• Sub-Par = This varies based on how many players are allowed in your lineups. If you start 1 QB, 2 RBs, 3 WRs, and 1 TE, then the tier baseline is after the average points of the 12th QB, 24th RB, 36th WR, and 12th TE. 

I created Crank Scores to combine this data into one number. The original Crank Score takes the player's points per game average and runs it through a formula that includes each of these applicable data points above and gives added weight to the Elite, #1, and Sub-Par values. This version of the score doesn't account for injury and it's heavily weighted by scoring average, which filters the data through a perspective that gives more upside to players who are productive on the field regardless of missed time. 

The second version of the Crank Score (I'll call it Crank II) doesn't give nearly as much weight to average fantasy points, doesn't give added weight to Elite, #1, or Sub-Par value, and factors games played. Crank II is a good look for those with a more conservative outlook about injury and players delivering quality games without the risk of overvaluing the upside of big weeks. 

Formulas for Crank and Crank II:

• Crank:  [FPG]+([FPG]*[Elite])+(([FPG]*[#1])*2)+([FPG]*[#2])+([FPG]*[#3])-(([FPG]*[Sub Par])*2)
• Crank II:  [(Sum of Elite Gs )+ (Sum of No. 1 Gs) + (Sum of No. 2 Gs) - (Sum of Subpar Gs)]

Note: FPG = Fantasy Points Per Game. Gs = Games

Original Crank is probably ideal for best-ball formats as a starting point for future projections (I'll share how I'm exploring this idea in the next article). While original Crank has also worked well as a starting point for head-to-head drafts, Crank II's injury weighting and slightly more tempered valuation of big weeks are best suited for head-to-head. 

I'm sorting these long lists as a combined list rather than split by position. Here are some thoughts from this dump truck of data for those of you seeking some basic insights:

• When the flex option is added into the mix, running backs see their value rise due to the viability of the No.3 role as a starting position rather than a sub par tier. 
• A volume producer like Jarvis Landry was a first-tier option for Crank II, but a low-end second-tier option with the original recipe. 
• C.J. Anderson is an early-third guy of note last year with the original recipe, but 88th on the board with Crank II. This disparity and others like it are worth knowing because it tells you something about that player's potential value if given an opportunity. Anderson could be mired in mediocrity due to the crowded depth chart in Denver. However, he'll get the first shot to be the man while Devonte Book is rehabbing. If he can start strong, Booker may never see time. 
• Flex leagues further devalue quarterbacks.compared to non-flex formats. Aaron Rodgers is player 59 in Crank II and player 65 with the original recipe. If you draft in these leagues, the only compelling reason to draft a quarterback before the fifth round is if you truly believe that player will have a record-breaking season and support at least four fantasy starters in the passing game (which only two have done) or your league drafts quarterbacks so early that if you wait, you won't get worhthile value. 
• Tight ends also have far less value in flex leagues. Kyle Rudolph and Jordan Reed, who were top consistency options last year, were even lower on these boards than the non-flex formats. Sure, Rob Gronkowski or Jimmy Graham have shown that they can blow the curve. However, we've been waiting a while to do them again. At some point, we have to give up on the idea that they will be anything more than good TE1s near the top of the rankings at the position, but not No.1 WRs in TE slots.  

 Remember, I'll provide cheat sheets for 2017 based on this info next week. 

Original Crank SCORES FOR 1 QB, 2 RB, 3 WR, 1 TE, 1 RB-WR FLEX LINEUPS (PPR) 

Crank II Scores For 1 QB, 2 RB, 3 WR, 1 TE, 1 RB-WR Flex Lineups (PPR)