Rearview SOS article (1 Viewer)

[TheLem]

Let me begin by saying that Rearview SOS article was one of the most refreshing articles i've read in a long time...I wish the FBG mag would carry those types of reads more.

here's an angle i'd like the author to take a look at though:

instead of giving me fppg/defense vs the offensive player.

how about telling us how these guys scored against tiers of point spreads?

for example...a top QB like Manning...yeah...he can tear up teams that they are favored by >7...but what about when they were the underdog?

I like this because: 1. it is easier to find the point spread than the fppg given up by defense per position.

mind you i'm not lazy...i like that research and I feel it separates good owners from worse ones...

but secondly, It gives you the opportunity to analyze how a player performs against good, bad or even teams vs HIS TEAM at that time...not just how that team finished for the season. In other words, the point spread is a better indicator of how 2 teams fare vs each other that week. There could be a key injured player on D - say a probowl cb...thus the qb might be expected to perform better than average. The point spread would reflect that. To me, this would help explain 'outlier' games.

I'd like to see these qb's on the following tiers:

how they scored when favored by 7

from 3 to 7

from -3 to +3

underdog -3 to -7

underdog by more than 7

If a team is favored to be beaten by 8 points...then I don't expect the rb from the underdog to have a good game....but if he did, then to me, that shows this guy is a difference maker.

i'm not very good at explaining in type because my mind runs faster than my fingers.

however, i do think this would be a good way to at least look at how offensive players performed under these circumstances.

 

[Chase Stuart]

If a team is favored to be beaten by 8 points...then I don't expect the rb from the underdog to have a good game....but if he did, then to me, that shows this guy is a difference maker.

I dunno about this. If the Texans played the Chiefs or the Colts (2004 version), I'd expect Dom Davis to be a heavy underdog and to do well. The point spread only shows the strength of the opposition, which isn't always that important. For example, the Jets and Browns and Packers were all bad teams last year, but let up very few FPs to opposing QBs.

 

[TheLem]

Let me rephrase what I originally wrote:

I think looking at a fantasy player's output in relation to point spreads would be an interesting way to see how a player performs.

What your article shows is player performance vs expected performance - which was great.

What i believe is that you can see trends of how a player performs in situations related to how his team is expected to perform.

Does P.Manning play well in games where Indy is the Underdog? how do his fppg relate to the average qb in a game where they are favored by 10? I know that defenses might be great vs the run and poor vs the pass and that isn't reflected in the point spread...but i'm not worried about that factor...i'm looking at it on a more macro level.

I do believe these are just as useful and relative as the other stats that were provided in your article - which again, I might add, was superb.

 

[Chase Stuart]

Let me rephrase what I originally wrote:

I think looking at a fantasy player's output in relation to point spreads would be an interesting way to see how a player performs.

What your article shows is player performance vs expected performance - which was great.

What i believe is that you can see trends of how a player performs in situations related to how his team is expected to perform.

Does P.Manning play well in games where Indy is the Underdog? how do his fppg relate to the average qb in a game where they are favored by 10? I know that defenses might be great vs the run and poor vs the pass and that isn't reflected in the point spread...but i'm not worried about that factor...i'm looking at it on a more macro level.

I do believe these are just as useful and relative as the other stats that were provided in your article - which again, I might add, was superb.

I think I see what you're getting at here. I might be able to look into this. You're going to run into sample size problems though. How many games was Indy an underdog this year? One? Two? None?And I doubt that there's a big correlation between how a player performs when he's a 7+ point favorite one year, and how he performs the next. The sample sizes are just going to be too small to be real meaningful I think.

But I'm wrong lots of times. If you can give me a better idea of what your thesis is, I'll see what I can run with the data I've got.

 

[tomarken]

I think I see what you're getting at here. I might be able to look into this. You're going to run into sample size problems though. How many games was Indy an underdog this year? One? Two? None?

And I doubt that there's a big correlation between how a player performs when he's a 7+ point favorite one year, and how he performs the next. The sample sizes are just going to be too small to be real meaningful I think.

But I'm wrong lots of times. If you can give me a better idea of what your thesis is, I'll see what I can run with the data I've got.

I think he was going one way, and then switched to a different direction in the second post.My first interpretation was that he wanted to see the players' performance adjusted for the strength of the defense at that point in the season. For example, instead of using the end-of-season SOS stats, you would actually try to pin down the strength of the opposing defense during that week (if Manning played the Bears in week 7, what was the strength of the Bears defense during week 7 - i.e., were they particularly weak at that point due to injuries?). I understand the idea there, but I think it would add some subjectivity to the study, and it would make the whole process a lot more difficult. I don't think the extra work (and guesswork) involved would be worth the result - or even produce a significantly different result for that matter.

Then it seemed like he was looking for more of a split stats kind of thing. For example, you have splits on Manning's passing numbers for when he's "Leading Big", "Leading Close", "Tied", etc. What if you could do splits on Manning's performance for games in which he was expected to win big, win close, lose close, etc. That's where the whole point spread thing comes in I think. For example, is it safe to assume that Manning will have a good day fantasy-wise if the Colts are favored by more than 7? How well does he typically perform in those situations?

I'd be interested to see if there were any significant results from this second interpretation, although my gut tells me there wouldn't be. My guess is that an individual player's fantasy success and his team's point spread are much too unrelated to be useful.