revolutionary tools.  groundbreaking articles.  proven results.

Cracking the Code on MFL10 Optimal Roster Construction

JamaalCharles

If you’re anything like me, the MFL10 bug has bitten you hard and you’re getting a little injection of adrenaline multiples times a day when you see an email come through with those sweet words: “You’re On The Clock!” From what I’ve seen anecdotally, the approach this year for most is a combination of Zero RB and late-round quarterback. Both are staples of the RotoViz drafting diet, so what other approach can we use to differentiate ourselves from the pack and gain a tactical advantage?

With this question in mind, I used the data from Fantasy Douche’s article that aggregated data from a 350-league sample of last year’s MFL10s and found the weekly scoring averages for different positional roster combinations. I filtered the data and isolated each roster position in order to calculate the average scores for all teams with a particular number of roster spots allocated to that position. Here are the results:

QB

RB

WR

TE

DEF

PK

Spots

# of teams

Avg Score

# of teams

Avg Score

# of teams

Avg Score

# of teams

Avg Score

# of teams

Avg Score

# of teams

Avg Score

8

0

35

139.7

104

144.7

0

0

0

7

0

252

143.9

929

146.3

0

0

0

6

0

1439

145.9

2095

146.4

0

0

0

5

0

1816

146.3

677

144.6

0

0

0

4

0

263

147.2

0

0

0

0

3

1269

145.4

0

0

1107

145.6

133

145.7

37

147.1

2

2536

146.3

0

0

2698

146.2

3253

146.4

3201

146.4

1

0

0

0

0

419

143.3

567

143.7

3805

3805

3805

3805

3805

3805

A couple of caveats: 1) These results are based on last year’s format, which had a 20-man roster limit; Rosters have expanded to 22 this year which could affect results so much that my conclusions are misleading, 2) The results do not incorporate the interaction between positions allocations; you get an idea of what happened to results if position allocations are added/subtracted, but you don’t know exactly how the other position allocations were affected. That said, I think the numbers give us a good quantification of positional opportunity cost/benefit, i.e. how much you gain or lose adding to a position.

The results show that 2QB teams outperformed 3QB teams, 4RB substantially outperformed higher allocations, 6WR teams were tops (but 7WR teams were very close), 2TE teams were better than 3TE, 2DEF were best, and 3PK teams were better than 2PK and 1PK squads (the 3PK sample size is pretty small at less than 1% of total teams; adjust your view of the PK results’ significance accordingly).

Results

Your team’s foundation should have 2QB, 4RB, 7WR, 2TE, 2DEF, and 3PK. But, that’s only 19 players. How do we determine how to optimally fill the last three spots? My theory is that we should add the last three spots based on the smallest negative differential between the optimal allocation and the next highest allocation. In other words, add first positions where adding a roster spot hurt the least. Adding QBs and RBs cost teams almost a whole point on average, adding to DEF cost 0.63 points, TE 0.61 points, and WR only 0.04 points. So, we should add to WR first, then TE and DEF to get to a total of 22 roster spots. Assuming this logic works, the final optimal lineup for a 22-roster team is 2QB, 4RB, 7WR, 3TE, 3DEF, and 3PK.

There are quite a few takeaways from the results. Only rostering 4 RBs would certainly fly in the face of a Zero RB approach were you need to sow many late-round RB seeds to reap the benefits of anti-fragility. Perhaps the best-ball format – without the ability to use the waiver wire – makes it more valuable to have solid contributors at RB that you can count on for weekly starts. Plus, there is less value in best-ball for the consistency that higher-round WRs give since boom-or-bust, later-round types (e.g. Kenny Stills) can be key contributors without the headache of having to guess correctly which week they’ll go off.

There are quite a few 3QB devotees that probably would argue with the results. My theory on why 2QB is optimal is that QBs naturally have higher floors and less benching/injury risk, therefore three QBs to fill one roster spot is excessive. You’ll get better value out of a 3rd allocation to TE where scoring is more random. Plus, TE injury risk is much higher and there is less certainty a TE will keep his starting position all year long. Some would also argue that 7 WRs is too much, but it’s likely that your flex spot in PPR generally, and especially in best-ball will be filled by a WR. Therefore, 7 WRs to fill the 3 WR slots and the flex isn’t even a 2-1 ratio.

In the 3,805 sample from last year, only 133 teams used 3DEF, and only 37 used 3PK. From what I’ve seen this year, the majority of teams aren’t using 3DEF or 3PK. Using higher DEF and PK allocations might be the biggest statistical advantage you can have on the field. We all know the scoring for DEFs and PK is random week-to-week, but the total points scored over the season by PKs and DEFs is significant versus other positions. Rostering 3 of each allows you to put up 10-15 point performances from your DEF and PK every week, a huge advantage versus the rest of your league. Plus, late-round DEF and PK additions should be even more valuable this year with an expanded roster from 20 to 22, since the opportunity cost for passing up a position player in rounds 21-22 is lower than doing so earlier in the draft.

I hope you can use this analysis in your drafting, or that it at least makes you think a little about how you’re allocating positions in MFL10s and whether it could be tweaked. Please provide me your feedback in the comments and let’s make some more money while feeding our MFL10 addictions: You’re On The Clock!

 

recent and related...

in case you missed it...

Putting 2019 in Context: Is the Running Back Dead Zone Dead?

Last July, I wrote an article about the running back dead zone — basically, RBs picked in Rounds 3 through 6 have busted at incredible rates since 2015. That article — along with several others which had similar findings — fundamentally changed my strategy in best ball drafts. Oops.1 Leonard

Read More
Connect
Support

rotovizmain@gmail.com

Sign-up today for our free Premium Email subscription!

© 2019 RotoViz. All rights Reserved.