Experts Poll: alternate scoring systems
Posted Feb. 04 at 02:25 AM
I want to make sure I am fair to our cast of experts. I appreciate the time they take to submit a top 20 at each position – I think it’s something the readers of the magazine enjoy.
And as I pointed out yesterday, luck (particularly with injuries) plays a big role in this kind of format. If Clay Matthews hadn’t taken out Kevin Kolb in Week 1, Lenny Pappano probably would have been our overall winner. Instead, Kolb lost his job, causing Pappano (the highest of everyone on Kolb) to finish next-to-last in quarterbacks.
We had the same deal with Alan Satterlee of Dynasty Rogues, a contender going into that final position. He was the highest on Tony Romo, so maybe if Romo doesn’t miss the second half of the season, Satterlee is your overall winner. Instead, he finished last in quarterbacks.
Smart, dumb or unlucky? I don’t know. I’m just tabulating numbers at the end of the deal and seeing who comes out on top. I’m not saying that it’s a perfect system.
Out of respect for the time these experts put in, however, I have looked at some alternate scoring systems, to see how much it affects what we call the overall results.
As we’ve explained a few times, in our system scores are tabulated using a system of 20 points for a first-place vote, 19 points for a second-place vote and so on. So if a player scores 100 fantasy points, an expert ranking him No. 1 gets 2000 points (100 x 20). Rank him 2nd and that’s 1900 points (100 x 19) and so on, right down to somebody ranking him 20th getting 100 points (100 x 1).
Those rankings were published yesterday. They are as follows:
EXPERTS POLL: Official scoring system
199,321 Pretzel May
199,093 Lance Ralls
199,038 Scott Pianowski
198,943 Nathan Zegura
198,794 Christopher Harris
198,566 Chris Liss
198,527 James Serra
198,389 Gregg Rosenthal
198,299 Jeffrey Kamys
198,021 Lenny Pappano
197,546 David Dorey
197,503 Scott Sachs
197,325 Bob Henry
197,151 Alan Satterlee
197,110 Michael Nazarek
197,029 Sam Hendricks
197,012 Rick Hawes
196,964 Jon Millman
196,939 Cory Bonini
195,630 Louis Tranquilli
198,097 Fantasy Football Index (not officially part of poll)
As I mull it over, however, is that system fair? Maybe it’s not placing enough value on the bottom of those top 20s. A guy ranked first, after all, is 20 times more valuable than a player ranked 20th. That might be out of whack, given that those top-20 running backs and wide receivers are guys who’ll be selected in the fifth and sixth rounds.
Maybe a fairer system, therefore, would be give 30 points for a first-place vote, 29 for a second, and so on, down to the last guy in top 20 getting 11 points (10 + 1), making him worth slightly more than a third of the value of the guys at the top of the draft. This system seems as fair – maybe even better.
So I re-ran the numbers using this system. Pretzel May would still be the winner, but there are a number of big movers on the overall board. Chris Liss moves up from 6th to 2nd. Pappano climbs from 10th to 3rd. And two of our bottom 3 (Jon Millman and Cory Bonini) move up into our top 6 overall.
The losers using that scoring system are Lance Ralls (drops 2nd to 9th), Nathan Zegura (drops 4th to 11th), Christopher Harris (drops 5th to 12th) and James Serra (7th to 17th).
The complete rankings for that system are as follows (with ranking in orginal scoring system in parentheses).
EXPERTS POLL: Alternate scoring system #1
With the last picks (20th) worth slightly more than a third as much as the top picks.
Points (official rank)
377,925 Pretzel May (1)
376,474 Chris Liss (6)
376,183 Lenny Pappano (10)
376,182 Scott Pianowski (3)
375,939 Jon Millman (18)
374,618 Cory Bonini (19)
374,362 David Dorey (11)
374,055 Alan Satterlee (14)
373,578 Lance Ralls (2)
373,419 Bob Henry (13)
373,353 Nathan Zegura (4)
373,339 Christopher Harris (5)
373,323 Scott Sachs (12)
373,063 Jeffrey Kamys (9)
373,054 Gregg Rosenthal (8)
372,854 Michael Nazarek (15)
372,344 James Serra (7)
372,281 Louis Tranquilli (20)
371,299 Sam Hendricks (16)
371,031 Rick Hawes (17)
374,416 Fantasy Football Index (not officially part of poll)
Taking steps even a step further, I ran the numbers with each player getting a 20-point bonus. Using this format, the multiplier number for a player ranked No. 1 is 40, and a player ranked 20th is worth 21 (20 + 1). So the guys ranked last are now worth a little bit more than half as much as the players at the top of the board. This system, I think, places a greater reward on finding value players in the middle round (or probably more accurately – at least avoiding dud picks there).
Again, there are plenty of big movers. Jon Millman moves up from 3rd-worst to 2nd-best. Bob Henry and Scott Sachs become above-average. ANd a host of guys in the top 10 move down into the bottom 8 – Ralls, Harris, Zegura, Serra, Gregg Rosenthal and Jeffrey Kamys.
Note that May again wins in this category, so there’s little arguing that he’s the top guy this year.
Those rankings are as follows …
EXPERTS POLL: Alternate scoring system #2
With the last player in each category (20th) valued at a little more than half as much as the players ranked first in each category.
Points (official rank)
556,529 Pretzel May (1)
554,914 Jon Millman (18)
554,381 Chris Liss (6)
554,344 Lenny Pappano (10)
553,327 Scott Pianowski (3)
552,297 Cory Bonini (19)
551,179 David Dorey (11)
550,958 Alan Satterlee (14)
549,514 Bob Henry (13)
549,143 Scott Sachs (12)
548,931 Louis Tranquilli (20)
548,597 Michael Nazarek (15)
548,062 Lance Ralls (2)
547,884 Christopher Harris (5)
547,827 Jeffrey Kamys (9)
547,763 Nathan Zegura (4)
547,718 Gregg Rosenthal (8)
546,161 James Serra (7)
545,569 Sam Hendricks (16)
545,050 Rick Hawes (17)
550,734 Fantasy Football Index (not officially part of poll)
Bottom line: Please keep in mind that the scoring system I use, while involving huge numbers of calculations, isn’t intended to be the cure-all or perfect scientific formula to crown a fantasy champion. You can see how much things change when the grading system is slightly altered.
Ultimately, when grading these experts, you’ll probably be best served by simply asking yourself whether you like the quality of their analysis – do you like reading the stuff they’re putting together.
I haven’t looked at different grading systems in the past. For the 2011 season, I am unsure if I will go with system 1, 2 or 3. I find the idea of increasing the value of the picks in the 15th to 20th range compelling, so my leaning is that the middle system is probably the best. Maybe I make like boxing and crown multiple champions – WBC, WBA and IBF was what boxing was using in the early ‘80s when I started to lose interest in that sport.
Your feedback on the different scoring systems is welcome.
Posted by ANDY RICHARDSON | Feb. 04 at 03:11 AM
Just a couple of thoughts about different ways to score things, to cut down on the high cost of players getting hurt. One idea would be to score the players on per-game rankings, with some sort of reasonable minimum, either 6 games or 8 games. It wouldn't have helped somebody who ranked Kevin Kolb, but it might have benefited those who were high on Tony Romo. Injuries are part of the game, so maybe that wouldn't work. Another thought could be to grade each expert's picks only on the top 15 (or 16, or whatever) performers of their top 20. The idea is that the worst performers -- the Kolbs and Romos -- would drop off the board entirely. So if I put Kolb at No. 4 in my top 20, and you only ranked my top 15 performers, I wouldn't have his pitiful number counting heavily against my rankings; instead my No. 16 guy would be there. Of your choices, I think I like alternate scoring system 1 the most. I think moving the 20th guy closer to the No. 1 guy makes sense in categories like RB and WR, where 20-30 of those guys will be drafted in the first 5 rounds, but not as much at QB or especially TE, where there might not even be 20 players at the position drafted.
Posted by LANCE RALLS | Feb. 04 at 12:39 PM
First, the current format is asking the experts to take a flier higher because putting him 20th doesn't add much value, so you make him 10th to earn some value if he pans out. If you want more of a "safe top 20", then you would want to increase the values at the bottom. It all comes down to understanding the scoring system just as one must do in any individual fantasy league. So, it is always hard to adjust a scoring system afterwards as people would have ranked differently. For a suggestion on the scoring, QBs score the most fantasy points, but RBs go first in a draft. So, I think it would make sense to tweak the multiplier per position. Also, most experts will agree that Kickers and D's are more random and more difficult to rank and less important to rank as there is always a solid Kicker or D on the wire at some point. So, reduce down the multiplier on those. So, I'd decrease the multiplier a good amount on D's and Kickers, a little bit on QBs and then increase the multiplier on the important position of RB and maybe a little on WR and TE that score a good amount less than QBs, but you will always see 20 WRs go before 20 QBs.
Add a Comment
Already a registered user? Please sign in to add comments.
To add comments, you must become a registered user of our site. To register, please click here.