Interesting Data on Stat Corrections

[Ryan G.]

I've been wondering for awhile now about the impact to our league of the final stat corrections that happen each week. It had seemed to me like it was potentially more relevant than most leagues because most of the final stat corrections apply to defensive players and we start 11 each week.

I compiled the data on the change between the initial scores and the final scores for the 13 weeks of the regular season. The initial scores are pulled from the Results email alert that MFL sends out late on Monday night after the MNF game. These scores are what sets record for each team for the rest of the week (until Thursday) and the records that are used for waiver ordering.

I then compared that to the final scores that are processed around 10am EST on Thursday mornings. When MFL sends out that email alert is doesn't identify in any way that an individual score has changed or how that change may have impacted the outcome of matchups or overall records.

I've attached the spreadsheet I came up with so you can look at it if you are interested. Additionally, I will post some further thoughts after work, but there are 3 data points that I find fascinating and that I'm guessing most people in our league don't realize.

1. There were ten (10!!!) matchups in the 2019 regular season where the outcome was changed after the final stat corrections. That's roughly 3.2% of all matchups for the regular season.

2. On average, everyone's score changes by just under a point (.971pts) every single week.

3. On average, fewer than a third of teams (7.7) have no change to their score in a given week.

More to come on this.

TW Stat Correction Analysis..xlsx (18.77 KB)TW Stat Correction Analysis..xlsx (18.77 KB)Stat Correction Snapshot.pdf (100.23 KB)

rg

[Ryan G.]

The highest change is any one week was 8.7pts. Swanny (Lebowski) had 8.7 points added to his week 10 score. The top 5 for the season is below.

1. Lebowski, wk10 (+8.7)
2. SUCK, wk5 (+7.1)
3. Cortez, wk12 (-6.7)
4. Longpoots, wk3 (-5.6)
5. Def, wk3 (+5.1)

There were an additional 18 score changes, besides the 5 above, that were 3 or more points.

[Ryan G.]

On the spreadsheet attached to the initial post I highlighted where the stat corrections changed the result of a matchup. Across the 10 matchups having a different result after stat corrections, there were 12 teams with an impact to their final record. That's half the teams with a different regular season record which is a stunning result in my opinion.



Joe Cooper had 4 matchups changed via stat correction (~22%) including both matchups in week 12. He we went 1-3 on those stat corrections. Rastafarian had 3 stat corrections victories (3-0!!). SUCK and Steelz stat corrected into a d**n tie in week 3 which has to be the black unicorn of outcomes considering we start 19 players with a complex scoring system.

Here is how the records would've changed for the impacted teams this year without the final stat corrections.

Team (flip record)    Intitial Record    Current Record
Joe Cooper(1-3):         16-10                      13-13 (current)
Rastafarian(3-0):         12-14                      15-11
Sharper(0-2):              9-17                        7-19
Def(2-0):                    14-12                      16-10
SUCK(1-0-1):              15-11                      16-9-1
ICP(1-0):                    12-14                      13-13
4 Horsemen(1-0):        16-10                      17-9 
DBFC(0-1):                  8-18                        7-19 
Kraftwerk(0-1):            11-15                      10-16 
Longpoots(0-1):           20-6                        19-7
#96 Cortez(0-1):         15-11                       14-12
Steelz (0-1):                15-11                      15-10-1

[Ryan G.]

Maybe the most important point that jumps out to me is that any matchup that is within about 6 or 7 points at the end of the Monday night game should be paid attention to regarding the final stat corrections on Thursday. Any matchup within (2 or 3 points) has roughly a 25% shot at flipping.

rg

[Ryan G.]

Here is where MFL posts the actual players that were stat corrected between Monday and Thursday every week.

Incidentally, this is also where they post updates on changes to player position designations.

www54.myfantasyleague.com/2019/site_news

You can save that bookmark and just change the year in the URL as the league site rolls over from one year to the next.

rg