Playing Kicker Matchups?

[TheDiplomats]

I played around with a formula to statistically look at the best kickers and matchups to play every week last year and it actually worked pretty well. There normally isn't a huge difference score wise in kickers and there are always kickers that randomly do very well or very poorly but for the most part it's pretty accurate. I'll post the rankings I came up with and also the 10 best and worst matchups I've calculated for kickers to be playing against.

These rankings are based on many different aspects. I Incorporated individual stats (including giving more points for longer FGs and accuracy), team offense stats including opportunities and redzone efficiency, and team defense including opportunities and redzone efficiency.

Here they are for week 11 if anyone is interested:

1 Phil Dawson 92.452
2 Matt Stover 90.501
3 Jeff Wilkins 86.361
4 David Akers 86.325
5 Olindo Mare 85.714
6 Adam Vinatieri 83.618
7 Neil Rackers 82.273
8 Josh Brown 81.733
9 Kris Brown 81.178
10 Joe Nedney 80.927
11 Rob Bironas 80.661
12 Jeff Reed 80.263
13 Jay Feely 78.508
14 Mason Crosby 77.576
15 Shayne Graham 77.512
16 Sebastian Janikowski 76.108
17 Matt Bryant 74.092
18 Ryan Longwell 65.112
19 Shaun Suisham 64.471
20 Robbie Gould 62.793
21 Morten Andersen 61.310
22 John Kasay 60.728
23 Nick Folk 59.818
24 Nate Kaeding 58.221
25 Stephen Gostkowski 57.147
26 Jason Elam 56.333
27 Jason Hanson 54.720
28 Lawrence Tynes 48.668
29 Rian Lindell 47.617
30 John Carney 44.082
31 Dave Rayner 42.890
32 Mike Nugent 41.873

And here are the best 10 and worst 10 matchups:

Best 10 matchups for kickers:
1. Baltimore
2. Houston
3. Cleveland
4. Miami
5. San Francisco
6. Kansas City
7. St. Louis
8. Cincinnati
9. New York Jets
10. Philadelphia

Worst 10 matchups for kickers:
1. Indianapolis
2. Pittsburgh
3. New England
4. Detroit
5. San Diego
6. New York Giants
7. Tennessee
8. Washington
9. Tampa Bay
10. Green Bay

Good stuff.

[DraftDodger]

Sorry that looks really screwed up to me. Too many good kickers at the bottom and too many poor ones higher up. I applaud the effort, but just go with the better kickers on the better teams, where possible. Even though the Patriots have a "must score TD's" mantra going these days, Gostkowski is still the #6 kicker in my league and is unlikely to be anywhere near as low as projected this week.

The Cafe Sit/Start list is a good place to look for a plug-in kicker (or to fill any other position, for that matter).

[mattb47]

Sorry that looks really screwed up to me. Too many good kickers at the bottom and too many poor ones higher up. I applaud the effort, but just go with the better kickers on the better teams, where possible. Even though the Patriots have a "must score TD's" mantra going these days, Gostkowski is still the #6 kicker in my league and is unlikely to be anywhere near as low as projected this week.

The Cafe Sit/Start list is a good place to look for a plug-in kicker (or to fill any other position, for that matter).

I've found that BY FAR matchups have the most influence on a kicker's performance. I think a team like New England is probably the exception to this because they seem to put up a lot of points against most teams so obviously some things need to be taken into account, but all in all it ends up being pretty good from week to week. You can rip it now, but it makes no sense to do so before you see the results as I have in the past.

Outside circumstances must always be taken into account (such as weather, indoor vs outdoor, home vs away, injuries to key offensive players, etc) but you're going to have that with ANY statistical approach (take Plindsey's defense formula as an example, generally on target but things do happen randomly sometimes).

[milo]

Dawson did alright Whats the coming weeks projections look like?

[mattb47]

While not everyone in my top 10 had great days, the top 2 guys on my list in standard scoring had the 2nd and 3rd highest points respectively and half of the players in the top 10 reached double digits.

Here's how the average scores worked out for the formula:
Top 10: 9.90 pts
Bottom 10: 7.40 pts

While it may not look like 2.5 pts is a huge difference for an average with these numbers, when you're only looking at this few points total, it does end up making a pretty good size difference overall.

NFL.com hasn't updated all of their stats yet, hopefully they do it soon so i can get an updated list out by tomorrow before I leave for thanksgiving.

If not, here is at least the list of the top 10 matchups for this week:

1. Nate Kaeding vs BAL
2. Phil Dawson vs HOU
3. Kris Brown vs CLE
4. Jeff Reed vs MIA
5. Neil Rackers vs SF
6. Sebastian Janikowski vs KC
7. Josh Brown vs STL
8. Rob Bironas vs CIN
9. Nick Folk vs NYJ
10. Stephen Gostkowski vs PHI

[waswik]

I'd be interested in more details of how you come up with your numbers. Also, if you have numbers on the historical accuracy of your method (ie from last year), that would be really interesting as well. As of now, we only have 1 weeks worth of data to go on. But, you might be onto something here so keep at it!

[mattb47]

I actually haven't kept any logs of how it worked last year. I just got working on this for this season in the past couple weeks so this past week is the first week of true "data" that I've really logged. I am planning on logging it on a regular basis now though. Last year the numbers were pretty similar to what I had this past week, a couple point difference on average between the top 10 and bottom 10.

I'll go into a little more on what goes into my formula but I'd rather keep the actual formula just for me, spent alot of time getting it to the point that I feel it actually works. There are essentially 3 parts that I take into account. The first part is the kicker's actual stats and accuracy. The second part is the offensive production of the team that each kicker plays on including things like redzone opportunities, how often an offense stalls in the redzone, extra points, etc. The last part is the defensive matchups. This includes things like how many FGs a team has given up, how good the redzone defense is, etc.

From what I've found, the biggest part of determining who to play from week to week is the defensive matchups. There are obviously going to be some exceptions, as there will be with any formula, but for the most part, playing defensive matchups ends up working quite well.

That being said, the stats still aren't updated on nfl.com and i doubt i'll be able to update it in time before the games start. But i'll be sure to report back on how the formula would have worked for this week.

[milo]

Did you run the numbers for this week yet? Thanks!

[Humpback]

While not everyone in my top 10 had great days, the top 2 guys on my list in standard scoring had the 2nd and 3rd highest points respectively and half of the players in the top 10 reached double digits.

Here's how the average scores worked out for the formula:
Top 10: 9.90 pts
Bottom 10: 7.40 pts

While it may not look like 2.5 pts is a huge difference for an average with these numbers, when you're only looking at this few points total, it does end up making a pretty good size difference overall.

NFL.com hasn't updated all of their stats yet, hopefully they do it soon so i can get an updated list out by tomorrow before I leave for thanksgiving.

If not, here is at least the list of the top 10 matchups for this week:

1. Nate Kaeding vs BAL
2. Phil Dawson vs HOU
3. Kris Brown vs CLE
4. Jeff Reed vs MIA
5. Neil Rackers vs SF
6. Sebastian Janikowski vs KC
7. Josh Brown vs STL
8. Rob Bironas vs CIN
9. Nick Folk vs NYJ
10. Stephen Gostkowski vs PHI

The results don't seem to be worth the effort. The #2 kicker was middle of the pack, #3 and #4 stunk, #5 was middle, #7 stunk, etc. The actual #1, #2, and #4-#7 kickers weren't even on this top 10 list. I'm not knocking your formula, I just think it's nearly impossible to predict with much accuracy from week to week, so it's not really worth the effort IMO.

[mattb47]

While not everyone in my top 10 had great days, the top 2 guys on my list in standard scoring had the 2nd and 3rd highest points respectively and half of the players in the top 10 reached double digits.

Here's how the average scores worked out for the formula:
Top 10: 9.90 pts
Bottom 10: 7.40 pts

While it may not look like 2.5 pts is a huge difference for an average with these numbers, when you're only looking at this few points total, it does end up making a pretty good size difference overall.

NFL.com hasn't updated all of their stats yet, hopefully they do it soon so i can get an updated list out by tomorrow before I leave for thanksgiving.

If not, here is at least the list of the top 10 matchups for this week:

1. Nate Kaeding vs BAL
2. Phil Dawson vs HOU
3. Kris Brown vs CLE
4. Jeff Reed vs MIA
5. Neil Rackers vs SF
6. Sebastian Janikowski vs KC
7. Josh Brown vs STL
8. Rob Bironas vs CIN
9. Nick Folk vs NYJ
10. Stephen Gostkowski vs PHI

The results don't seem to be worth the effort. The #2 kicker was middle of the pack, #3 and #4 stunk, #5 was middle, #7 stunk, etc. The actual #1, #2, and #4-#7 kickers weren't even on this top 10 list. I'm not knocking your formula, I just think it's nearly impossible to predict with much accuracy from week to week, so it's not really worth the effort IMO.

This actually wasn't the true list....this list was based purely on matchups from stats from the week before (not including any other part of my formula) since stats weren't updated in time for me to get it out before i left for thanksgiving so i didn't necessarily anticipate these were going to be all that accurate. I wasn't actually ever able to get all the numbers plugged in for last week because by the time I got back from thanksgiving, the stats were already updated for the teams that had played on Thursday and Sunday.

I made it pretty clear that this wasn't my actual formula when I posted this....I just wanted to give people something to work with since I couldn't get it out and even if the results would have turned out poorly this week....there's NO formula (not even Plindsey's defensive formula) that doesn't have weeks where random things occur that are out of the ordinary. Looking back from last year when I was working on this when I've actually been able to get the information in time to make it usable, the turnout has been quite good on a regular basis.