Let’s Look at Some Numbers, Part 1?

North Point, the Basement. No pictures or photoshopping, just a simple, straightforward look at Fantasy Football data. Sorry this got longer than intended, though a lot of it is Tables and Tables are awesome….way more awesome than graphs….um, ok, let’s go:

Background: Hammond off-handedly pronounced himself and Enright the best owners. This seems to be loosely based on number of career wins, titles, and maybe career points. Franssen is annoyed by this and writes some stuff.

“He uses statistics as a drunken man uses lamp posts – for support rather than for illumination.” – often attributed to Andrew Lang

Goal: Explore the numbers. Don’t cherry pick to find support for an idea, but rather see if the numbers can give some insight. Hopefully, I figure out why I have been terrible the past several seasons.

I’m not going to get into details, my thought process, or all the rabbit holes I went down (there were many), but I’d be happy to share those in future posts if people are interested. I’d like to avoid rambling on too much, so decided to keep this short and focus on one set of numbers and a bit of my interpretation.

I figured out how many points a league average team would be expected to score (and have scored against them). In 2017, very straightforward. For 2018 and 2019 with variable caps, some extra parenthesis, division, and multiplication are needed in the equation. Here’s the basic equation I used:

PP$ * Util * Cap = EPF (or EPA)

PP$ = League Wide Average Points Per Dollar Spent
Util = League Wide Average of Cap Dollars Used
Cap = Salary Cap Figure
EPF = Expected Points For (EPA = Expected Points Against)

I then took the difference between these expected numbers and the actual numbers to calculate DEPF and DEPA (Difference from Expected Points For or Against). Here are those numbers for 2019 (oh and Diffs is the sum of DEPF and DEPA):

FRANCHISE	W	L	DEPF	DEPA	Diffs
Culver City Krakens	10	3	122.84	-25.26	97.58
Marshfield Pirates	9	4	113.34	-40.50	72.84
Batavia Colliders	6	7	14.92	-93.55	-78.63
Bestine Banthas	6	7	-129.24	-33.26	-162.50
Tooz Town GIMB	10	3	4.68	229.54	234.22
Madtown Marauders	8	5	36.01	-0.13	35.88
Cleveland Battery	2	11	-281.14	80.14	-201.00
North Point Chitons	2	11	-290.13	100.58	-189.55
Holley Hawks	8	5	66.47	11.83	78.30
Fransburg (r)EvoLUtion	7	6	258.46	-75.29	183.17
Oregon Dream Team	5	8	81.04	-92.07	-11.03
Linn County Lumberjacks	5	8	2.76	-62.64	-59.88

Not going to comment much. A few explanations:

A positive DEPF means you did better than league average. A negative means you did worse. To me, DEPF seems to indicate skill of the owner, luck can play a role, but you have control over your own DEPF.

A positive DEPA means you had fewer points scored against you (positive DEPA is good, like positive), a negative DEPA means you had more points scored against you than expected. I think of DEPA as measuring how fortunate or lucky you were–DEPA is completely out of your hands.

Interpret as you will. A few things stand out: Oregon was unfortunate, Tooz Town was really pretty lucky. Fransburg, Marshfield, and Culver City seem to really know what they are doing in picking/acquiring players.

Diffs by the way is a great predictor of wins. Pearson’s correlation 0.88 (p-value 2.06e-12). This is probably not surprising, but I might explore this further another time.

Here are the Top Ten Teams in DEPF over the last three years with their standing after the post-season (we don’t do 3rd or 5th place games, so two 3rd place teams and two 5th every year).

FRANCHISE	YEAR	W	L	Play-offs	DEPF	DEPA
Marshfield Pirates	2017	8	5	2nd	304.03	-105.10
Fransburg (r)EvoLUtion	2019	7	6	Missed	258.46	-75.29
Batavia Colliders	2018	11	2	Champion	210.24	225.11
Madtown Marauders	2018	8	5	2nd	208.81	36.34
Fransburg (r)EvoLUtion	2017	12	1	3rd	199.62	109.77
Culver City Krakens	2019	10	3	2nd	122.84	-25.26
Marshfield Pirates	2019	9	4	Champion	113.34	-40.50
Culver City Krakens	2017	6	7	5th	109.97	-155.07
Holley Hawks	2017	9	4	Champion	104.96	140.00
Oregon Dream Team	2018	7	6	Missed	101.28	-93.57

A few notes: Four 2017 teams are in this list, 3 took the top three spots. In 2018 and 2019 three teams made the list. In each of those years they either missed the play-offs, or took the top two spots. For play-off success you need a DEPF of at least 100 and some luck to get in (the two that missed had high DEPA’s, this doesn’t work if you are coming from a division with only one play-off spot).

Now just for fun, the sums of DEPF and DEPA for 2017-2019 (only teams with 3 seasons shown).

FRANCHISE	DERF		FRANCHISE	DEPA
Fransburg (r)EvoLUtion	538.05		Culver City Krakens	-168.03
Marshfield Pirates	418.35		Oregon Dream Team	-149.28
Batavia Colliders	319.21		Fransburg (r)EvoLUtion	-138.79
Culver City Krakens	271.11		Batavia Colliders	-126.10
Madtown Marauders	138.29		Holley Hawks	5.30
Holley Hawks	95.58		Marshfield Pirates	28.12
Oregon Dream Team	87.76		Madtown Marauders	80.27
Bestine Banthas	-248.86		Bestine Banthas	115.60
Cleveland Battery	-531.53		Cleveland Battery	164.62
North Point Chitons	-564.35		North Point Chitons	249.14

I’m still interpreting DEPF as telling us which owners have been displaying the most or least skill. DEPA lets us know if any team or teams have been particularly unfortunate over these years.

Make what you want of the numbers. They are only there to shed light, not to prop up our hypothesis.

Next time, Regular Season numbers and predicting Play-Off Success, or something else, or maybe there is no Part 2.

© 2020, Tom Rothfus. All rights reserved.