Wednesday, January 12, 2011

A 2011 Brothers Keeper Draft Order Discussion and Proposal

INTRODUCTION
For those looking for an idea on how to come up with an unbiased draft order from a previous fantasy football season's results this post is for you. This assumes you have most if not all of your league participants returning to play year after year in a keeper or dynasty league or the like. The following discussion is based on the desire for parity within the league. This requires a working knowledge of Building a Fantasy Football Draft Value Chart and the results of Do Bookends Rule? A Study on the Value of Draft Slot/Position in a Snake Draft. From those previous posts you should be able to see that the draft order's importance because in ideal settings assuming perfect market efficiency fantasy drafts reward the teams who pick earlier in a snake draft. Therefore determining a keeper or dynasty fantasy football draft order is a very important aspect in keeping league play competitive.

Because of the change in the playoff format from an 8 team playoff to a 6 team playoff, discussion arose on the Brothers Keeper League Message Board via Commissioner's Note to determine the 2011 draft order. So Brothers Keeper needs to come up with a new way to determine the draft order after a season is over. The former order as outlined in II. E. of The Canons are re-posted here:
Section E: The Keeper League Draft Order: In an effort to prevent people from neglecting or throwing their playoff and consolation bracket games the next season’s draft order is subject to the previous season’s playoff results:

1st pick = 10th Place (Worst regular season record)
2nd pick = 9th Place (2d worst regular season record)
3rd pick = 5th Place (Winner of Consolation bracket)
4th pick = 6th Place (Loser of 5th place game)
5th pick = 7th Place (Winner of 7th place game)
6th pick = 8th Place (Loser of 7th place game)
7th pick = 3rd Place (Winner of 3rd place game)
8th pick = 4th Place (Loser of 3rd place game)
9th pick = 2nd Place (Loser of league championship)
10th pick = 1st Place (League Champion)
As a result of the playoff changes from an 8 team playoff to a 6 team playoff this draft order ruling no longer holds true as the last 4 teams in the regular season standings now enter into a 2 week consolation bracket.

The discussions according to the Commissioner's note (12/15/10), B Garner (12/12/10), and P Smith (12/15/10) came up with four or five suggested factors should be considered in determining the league's future draft order in order to keep the league as competitive as possible (a.k.a. parity):

1) League Champion should have last pick (10th) overall.
2) The regular season team total fantasy points scored.
3) The regular season standings (winning percentage).
4) The playoffs.
5) Strength of regular season schedule/total fantasy points against.

SCENARIOS 1-5
In an effort to visualize these potential elements let's utilize this year's results for the discussion of the alternate issues involved one at a time. (1) Obviously, Yer X-MOM is the 2010 Brothers Keeper league champion. However, knowing who should draft last does not resolve the rest of the draft order.

(2) Here's the logic behind considering the team's total fantasy points scored: the higher the total fantasy points scored the later your team should draft. If you utilize this method you need to consider the same time period for all teams. For example, the whole regular season of fantasy play. Yer X-MOM had the highest team total of fantasy points scored through the regular season of fantasy football (NFL weeks 1-13). It would logically follow that Yer X-MOM would pick 10th overall in 2011's fantasy draft because the draft actually rewards teams with earlier draft picks as demonstrated in the post: Do Bookends Rule? A Study on the Value of Draft Slots/Positions in a Snake Draft. If we consider the rest of the league's total team points for the regular season from the lowest to the highest the draft order would be thus:

1. CYBORG PIRATE NINJAS (1180.51 FP)
2. Simba's Buttkicker! (1193.53 FP)
3. Points ON the Bench (1222.65 FP)
4. Kabuki Strikes Back (1380 FP)
5. Ghost (1418.61 FP)
6. SOCIAL ISoTopes (1445.81 FP)
7. TheSloppyButts (1448.64 FP)
8. Redman's CatClippers (1451.1 FP)
9. Yankinfowlsandfeline (1478.1 FP)
10. Yer X-MOM (1633.19 FP)

3) If we were to take the regular season standings (winning percentage) into account, we would end up with a few teams tied and need a tie breaker to determine who picks before the other. Screaming "First!" should not be the determining factor. Nevertheless, the draft order would look like this if the lowest regular season winning percentage teams had earlier picks than the highest winning percentage teams.:

1. Simba's Buttkicker! (.154)
2. Points ON the Bench (.231)
3. CYBORG PIRATE NINJAS (.385)
4. Kabuki Strikes Back (.462)
4. Ghost (.462)
6. Redman's CatClippers (.615)
6. Yer X-MOM (.615)
8. SOCIAL ISoTopes (.692)
8. TheSloppyButts (.692)
8. Yankinfowlsandfeline (.692)

4) If we were to look at the playoff results, including the consolation bracket, we'd have the following draft order as the reverse in the Final League Standings. We should probably use the team's place at the end of the playoffs rather than who won each consolation game thus foregoing rewarding consolation winners. The fantasy playoffs for any team in the consolation bracket or after being eliminated from the championship no longer are penalized for starting potential developmental players (players started in the playoffs for championship contenders are penalized starts toward their developmental eligibility status). The ruling that began the rewarding of teams that won consolation games was put in place in order to keep teams from "throwing" games (losing intentionally) in order to get a better draft position.

1. CYBORG PIRATE NINJAS
2. Kabuki Strikes Back
3. Simba's Buttkicker!
4. Points ON the Bench
5. Redman's CatClippers
6. SOCIAL ISoTopes
7. Ghost
8. Yankinfowlsandfeline
9. TheSloppyButts
10. Yer X-MOM

5) If we considered the strength of schedule and awarded teams with a harder regular season schedule the opportunity to have a better draft position for 2011, the draft order would be:

1. Simba's Buttkicker! (1526.57 FPA)
2. Ghost (1460.77 FPA)
3. Yer X-MOM (1430.79 FPA)
4. Kabuki Strikes Back (1420 FPA)
5. Points ON the Bench (1417.66 FPA)
6. Yankinfowlsandfeline (1387.66 FPA)
7. CYBORG PIRATE NINJAS (1372.25 FPA)
8. TheSloppyButts (1325.8 FPA)
9. SOCIAL ISoTopes (1303.08 FPA)
10. Redman's CatClippers (1207.56 FPA)

Not many of these options alone really works out well. Some of those draft orders reward teams with early draft picks that did really well and logically should draft later (e.g., Yer X-MOM would pick 3rd if we used FPA but is the champion who should pick last). What we need is a way to value each issue equally. We need a power ranking for our league's teams. Thankfully, if we wanted to weigh all of these factors equally, we need to use an equation I accessed a few years ago when trying to determine how to rank baseball or basketball players according to the specific statistical categories I wanted to utilize (or were required of my fantasy teams according to the league scoring settings). That meant rewarding players for high percentages in some categories, rewarding high stats in some categories and low stats in other categories. After some conversions here are the results that I have come up with to determine the fantasy draft order for 2011 (see the Draft Order Equation spreadsheet at the end of this post with the equation's explanation):

4 Factor Power Ranking Draft Order
1. Simba's Buttkicker!
2. Points ON the Bench
3. CYBORG PIRATE NINJAS
4. Kabuki Strikes Back
5. Ghost
6. Yankinfowlsandfeline
7. SOCIAL ISoTopes
8. Redman's CatClippers
9. TheSloppyButts
10. Yer X-MOM

While the 4 Factor Power Ranking Draft Order model considers the regular season strength of schedule a 3 Factor Power Ranking Draft Order would not. So for your consideration here is the model without the strength of schedule adjustment:

3 Factor Power Ranking Draft Order
1. Simba's Buttkicker!
2. CYBORG PIRATE NINJAS
3. Points ON the Bench
4. Kabuki Strikes Back
5. Ghost
6. Redman's CatClippers
7. SOCIAL ISoTopes
8. Yankinfowlsandfeline
9. TheSloppyButts
10. Yer X-MOM

4 Factors or 3 Factors?
Before entering into the debate as to whether or not the 4 Factor Power Ranking Draft Order is better than the 3 Factor Power Ranking Draft Order, it should be observed what these equations do. First, it levels the playing field for each stat involved by calculating the league averages (AVG) and the standard deviation (STDEV). Because there is more than one stat involved it demonstrates with a broader view how good teams really are. Third, the process circumvents potential negative outcomes like "throwing" games (intentionally losing). It is a pretty objective view of the health of each team after the fantasy season is over. It really does consider the important issues and does not focus on anything too small. Plus, either scenario, has the current league champion drafting last, though all that team's elements were considered along with the rest of the league's data.

The nice thing about seeing these two models is that you can see a team's merit regardless of their opponents (3 Factor Power Ranking Draft Order). Like I have stated before, "sometimes that one dude picks your man right before you were going to!" So the difference between these two models, while small is noticeable to 4 teams in particular (CYBORG PIRATE NINJAS, Points ON the Bench, Redman's CatClippers, and Yankinfowlsandfeline). The other 6 teams it will not effect their draft position if we utilize these two models.

In my humble opinion, the 3 Factor Power Rankinig Draft Order formula is more accurate in placing Yankinfowlsandfeline later in the draft than Redman's CatClippers. Redman's CatClippers gets bumped lower in the draft order of the 4 Factor Power Ranking Draft Order simply because he had such a ridiculously easy schedule (see the Strength of Schedule order above). Yankinfowlsandfeline was a much stronger team as indicated by a 1st round bye (regular season 1st place), 3rd place playoff finish, and 2d most fantasy points scored. Yankinfowlsandfeline had a very average strength of schedule and should not be rewarded because his strength of schedule happened to be more difficult than Redman's CatClippers's strength of schedule.

Now how should the league handle the other much more subtle difference in the draft order between Points ON the Bench and CYBORG PIRATE NINJAS? One draft pick could make all the difference for both of those teams in next year's draft. I'd be interested in the league's responses to this question. I would assume that CYBORG PIRATE NINJAS's team needs more help (if you look at the total fantasy points his team scored he ranked the lowest) but B Garner might not want to reward a team who failed so poorly in the consolation bracket. Points ON the Bench did perform better in the consolation bracket and scored more fantasy points in the regular season but had a poorer regular season winning percentage and had a more difficult strength of schedule, though not by much. Both teams were straddling the strength of schedule league average. CYBORG PIRATE NINJAS's team needs more help regardless of strength of schedule considerations and therefore the 3 Factor Power Ranking Draft Order makes more sense.

TEST CASES
The only way I can think about testing the 3 Factor Power Ranking Draft Order or 4 Factor Power Ranking Draft Order for accuracy is to utilize previous seasons of Brothers Keeper fantasy football statistics to determine if the draft orders in the years that followed were in any way similar or even at least reasonable.

Actual 2008 Draft Order
1. PAIN
2. Simba's Buttkicker!
3. Cleaning Lady
4. Redman
5. Squash
6. Team Cake Farts
7. Yer X-MOM
8. Ghost
9. Kabuki
10. Yank

2008 Test of 3 Factor Power Ranking Draft Order
1. PAIN
2. Simba
3. Cleaning Lady
4. Squash
5. Ghost
6. Kabuki
7. Team Cake Farts
8. Redman
9. Yer X-MOM
10. Yank

Actual 2009 Draft Order
1. Simba's Buttkicker!
2. Yankinfowlsandfeline
3. Kabuki Strikes Back
4. Redman's CatPunters
5. Sea Buzzards
6. CYBORG PIRATE NINJAS
7. CustodianPandemonium
8. Yer X-MOM
9. Ghost
10. Favre$Footlong

2009 Test of 3 Factor Power Ranking Draft Order
1. Simba
2. Team Cake Farts
3. Yank
4. Squash
5. Kabuki
6. Redman
7. Cleaning Lady
8. PAIN
9. Ghost
10. Yer X-MOM

Actual 2010 Draft Order (or here)
1. Kabuki Strikes Back
2. TheSloppyButts
3. Yankinfowlsandfeline
4. Redman's CatClippers
5. Ghost
6. Favre$Footlong
7. CYBORG PIRATE NINJAS
8. Simba's Buttkicker!
9. Sea Buzzards
10. Yer X-MOM

2010 Test of 3 Factor Power Ranking Draft Order
1. Kabuki Strikes Back
2. CustodianPandemonium
3. Favre$Footlong
4. Ghost
5. Redmans CatPunters
6. Sea Buzzards
7. Simba's Buttkicker!
8. CYBORG PIRATE NINJAS
9. Yer X-MOM
10. Yankinfowlsandfeline

If these tests demonstrate anything, it is that the 3 Factor Power Ranking Draft Order is not too far off of the normal draft experience we have already set up. While I haven't calculated the variance of these results (a low variance would be a good thing) a quick visual should show you that in general those who belong at the beginning of the drafts have actually been in the early picks in the draft. Those who belong in the middle of the draft are in the middle of the actual drafts and the those at the end of the draft order in the 3 Factor Power Ranking Draft Order are later in the Actual Draft results.  Since the the draft rewards the teams with the earlier picks with the opportunity to gain more fantasy points through the draft (as indicated by my study on Do Bookends Rule?) the 3 Factor Power Ranking Draft Order structure makes the most sense.

CONCLUSION
In conclusion, I propose the 3 Factor Power Ranking Draft Order as the means by which Brothers Keeper's 2011 fantasy football draft order should be determined.



*The far left column is the Team names
*The next column is Final Standings Rank (after the Playoffs)
*The next column is Final Standings Rank Conversion: It should reward the teams with the lower numbers and not higher numbers = Rank*-1
*The next column is the total regular season fantasy points scored (Pts For)
*The next column is the Winning Percentage (Win Pct) = # of team's wins/# regular season games (13)
*The next column is the Winning Percentage Conversion (Win Pct Conversion) = Win Pct*100
*The Strength of Schedule/FP Against (FPA) column is the total regular season fantasy points scored against
*FPA Conversion column converts the Strength of Schedule into negative numbers in order to reward the teams with harder schedules and penalize the teams with easier schedules = FPA*-1

**Final Rank! = (team's Rank Conversion - League AVG Rank Conversion)/STDEV of Rank Conversion
**Pts For! = (team's Pts For - League AVG Pts For)/STDEV of Pts For
**Win Pct! = (team's Win Pct Conversion - League AVG Win Pct Conversion)/STDEV of Win Pct Conversion
**Strength of Schedule! = (team's FPA Conversion - League AVG FPA Conversion)/STDEV of FPA Conversion

***3 Factor Power Ranking Draft Order Without Strength of Schedule = Final Rank! + Pts For! + Win Pct! This then is ranked from the lowest to the highest total (ascending order). The lower Final Rank! the higher the draft pick should be. The current spreadsheet is ranked with the 3 Factor Power Ranking Draft Order.
***4 Factor Power Ranking Draft Order = Final Rank! + Pts For! + Win Pct! + Strength of Schedule! This then is ranked from the lowest to the highest total (ascending order). The lower the Final Rank! the higher the draft pick should be. Note that this spreadsheet is not ranked by the 4 Factor Power Ranking Draft Order.

****Difference Between 4 Factor Power Ranking and 3 Factor Power Ranking Draft Order = 4 Factor Power Ranking Draft Order - 3 Factor Power Ranking Draft Order
Below the team totals (*) I have two lines that were utilized to calculate the League AVG and STDEV for each category that was utilized in the equations (**!).

*****Below the left/right navigation/scroll bar you will see multiple sheets/tabs with the previous years totals calculating the next year's drafts. The results were posted earlier in the blog but you can now see the numbers in detail. I did not use categories that were not part of the equation as was indicated in the 2011 Draft Equation.

No comments:

Post a Comment