Some DST stats from last season

[Timbathia]

Since I am in the process of preparing for another season of playing defensive matchups, I have trawled through the stats from last year to see if anything stood out (rather than just blindly relying on Plindsey's formula each week). Thought I would include a few on the interesting stats in here without any real analysis (I guess it is up to you whether it is relevant or not), but note that this is all based on averages for the 2006 season. It also doesnt necessarily mean anything for the upcoming 2007 season.

Note - the DST scoring this is taken from is fairly standard, with slightly higher points for keeping offenses to low scores.

Home defenses - I always thought that defenses at home were significantly better than on the road. Seems not, as last year the average home def scores was 10.6, while on the road it was 10.1. Similarly, defenses scored 65 'big' scores at home compared to 60 on the road (a big score was considered as 15 or more), not a huge difference.

Easy Matchups - We all knew the Raiders offense were cake for defenses each week, giving up an average of 17 both in Oakland and on the road. The only other offenses that were statistically good matchups (classified as when opposing defenses averaged 13 or more) both on the road and at home were the Vikings and Browns. The good matchups for defenses who were playing on their own turf were against the offenses of the Lions, Steelers, Jags, 49ers, Texans and Bucs. Interestingly, it was a good defensive matchup to play the Seahawks in Seattle and the Packers in Green Bay, but not when they were on the road.

Tough Matchups - The teams that were always bad matchups for defenses (classified when defenses averaged 7.5 points or less) were the Chargers, Colts, and more surprisingly the Ravens and Redskins (though they both just scraped in there). The Jags were a bad matchup for the defense only in JAX, and interestingly so were the Texans in HOU (who were cake on the road). Two teams were bad matchups only when they were traveling, which were the Giants and the Saints.

Big Weeks - Playing defensive matchups is all about finding the team that will have a 'big' week. Based on the definition used here the only offense that never gave one up was the Chargers, mainly due to LT's unbelievable season. The Redskins were the only defense that never had a 'big' week (every other team pulled off at least two).

Easy Matchup Locations - This would be defined as a stadium where both the visiting defense and home defense scored noticeably higher points than both of these defenses did at other locations. From the data no such location existed.

I have no intelligent conclusions, except that this stuff seems to support the use of defensive matchups each week (rather than playing the same 'good' defense each week). This is because every defense except one had at least two well-above scoring weeks during the season. Most of the above is pretty obvious, but just keep an eye out for the anomalies (meaning teams that everyone thinks of as easy or hard matchups that possibly arent depending on the location).

[Axl's Army]

hindsight is 20/20. i never get those matchups right. i may spring for a 'good' defense like chicago this year. do you think devin hesters offensive inclusion will detract from his special teams success?

[bagobonez]

I'm all about playing matchups on D. Last year for http://www.ffauthority.com I wrote waiver wire articles every week, where I was supposed to explain which guys from the waiver wire you could pick up and have a good week from. I had a hit or miss record with the skill positions, but on defense it was a piece of cake. Nearly every week there is a waiver wire defense out there that gets to play Houston, Oakland or Cleveland, and they almost always had a good week, especially those playing Oakland.

footballguys.com also encourages playing matchups. THeir theory is to draft two defenses that will have a pretty easy schedule all the way through the season if the matchups are played right. This year they suggest drafting both Denver and Miami. Basically one of those D's faces a crappy offense every week of the year, so you just play whichever defense has the good matchup that week. The flawed part of this is that both Denver and Miami are ranked fairly high, so you wouldn't be able to wait too long after the Ravens and Bears are off the board. Plus you gotta use up 2 roster spots for the length of the season.

But either way, I think it's a much better alternative than taking Baltimore's D in round 9 when you could be unearthing the next Marques Colston or Frank Gore with that pick and getting your defense later on.

[a4ce]

NIce post. Very informative.

I usually play match-ups and for the most part it works but this season I might spring for the BAL or CHI D/ST early. Last season, CHI was scoring as much points as a top 10 RB with our scoring system. And that was almost on a weekly basis. Another thing is that in my league, every add/trade costs me $2 bucks. It ain't much but that adds up over 16 weeks in addition to the other positions that I add. Eh, we'll see. I guess you gotta spend money to make money.

[Timbathia]

NIce post. Very informative.

I usually play match-ups and for the most part it works but this season I might spring for the BAL or CHI D/ST early. Last season, CHI was scoring as much points as a top 10 RB with our scoring system. And that was almost on a weekly basis. Another thing is that in my league, every add/trade costs me $2 bucks. It ain't much but that adds up over 16 weeks in addition to the other positions that I add. Eh, we'll see. I guess you gotta spend money to make money.

Yeah, having to pay for transactions does change things a bit. Based on our settings, the average def score by the Ravens each week last year was approx the same as starting whichever def was playing the Raiders. The drop-off in average score from playing the same def each week after the Ravens was much greater than the drop-off in average score that you would have had by playing the defense that was opposing the worst offenses each week (i.e. the top 5 defenses scored over a point a week less on average than the defenses playing the 5 worst offenses). I guess this means that you really are better off playing matchups rather than the same def each week.

I think for leagues where you pay for transactions (and dont want to fork out cash each week), it is wise to look for defenses that have reasonably complimentary schedules (as suggested in a previous post) so that you maximize the number of weeks that your two defenses are playing weaker offenses.

[Plindsey88]

Nice Stats...

[mystykoekaki]

Nice Stats...

So says the defense guy! ...is that included in your equation?

[Timbathia]

Nice Stats...

Please note, I will still continue to use Plindsey's formula for selecting defensive matchups each week, but since I am bored at work I am now starting several different ones of my own to see how well they work this year.

I have put together a formula now for the expected average score by a defense each week based on strength of opposing offense, average defense points per date, with a half a point added in for home field advantage. The opposing offense rating is more heavily weighted than the previous performance of the defense. It wont work well until the middle of the season (if it even works at all), but it can be easily used for predictions if you think you have a good idea of how offenses will do running or passing the ball.

The strength of opposing offense is based on a combination of the QB rating and average rushing yards for an offense (more heavily weighted to QB rating) - there was a reasonably good fit using this with last years results, which I guess makes sense as if your QB blows or you cant effectively run the ball and tire the defense, then you will tend to give up more FFL points to the defense. A couple of teams dont fit well with this data, but I think it okay to ignore them as most teams do fit.

If anyone has other suggestions of things to include then let me know, as I think I will try several formulas this season ranging from simple to complex to see what works (for just in case Plindsey gets sick of of us and stops posting his ratings each week).

[Paladyr]

Nice Stats...

Please note, I will still continue to use Plindsey's formula for selecting defensive matchups each week, but since I am bored at work I am now starting several different ones of my own to see how well they work this year.

I have put together a formula now for the expected average score by a defense each week based on strength of opposing offense, average defense points per date, with a half a point added in for home field advantage. The opposing offense rating is more heavily weighted than the previous performance of the defense. It wont work well until the middle of the season (if it even works at all), but it can be easily used for predictions if you think you have a good idea of how offenses will do running or passing the ball.

The strength of opposing offense is based on a combination of the QB rating and average rushing yards for an offense (more heavily weighted to QB rating) - there was a reasonably good fit using this with last years results, which I guess makes sense as if your QB blows or you cant effectively run the ball and tire the defense, then you will tend to give up more FFL points to the defense. A couple of teams dont fit well with this data, but I think it okay to ignore them as most teams do fit.

If anyone has other suggestions of things to include then let me know, as I think I will try several formulas this season ranging from simple to complex to see what works (for just in case Plindsey gets sick of of us and stops posting his ratings each week).

Are you sure playing at home is an advantage? I seem to remember last year that some defenses played better in away games for some reason.

[Timbathia]

Are you sure playing at home is an advantage? I seem to remember last year that some defenses played better in away games for some reason.

If you look at the entire NFL, then yes, home field is an advantage on average. Looking at specific teams can change that. There were 4 defenses that were obviously better on the road - Packers, Bengals, Rams and Giants, while there were about 8 that were much better at home - Ravens, Chargers, 49ers, Jaguars, Browns, Falcons, Titans and Raiders.

Most of the other defenses werent too different either at home or on the road.