What's new
Fantasy Football - Footballguys Forums

This is a sample guest message. Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Why is the standard to use end of year totals (1 Viewer)

Max Power

Footballguy
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.

On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases.

Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.

Points per game seems like a more accurate way to rank players, don't you think?

 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
 
The arrogant answer is "Because fantasy teams didn't score points on the games when they were out".

The simple answer is "because that's how most league software sites rank players by default, and user's dont want the extra click to sort".

The better answer is that neither is sufficient for complete analysis on its own. Some of the players you gave are good examples of this, as Ronnie certainly doesn't make many lists at RB10, but likely doesn't fall all the way to RB39 either.

In the end what I really want is a guys with both high PPG and 16 games a year, so both stats are useful.

 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
Not for me. While it's the default way to sort in my league, I always resort by average for my rankings.
 
You can't completely ignore the missed games. Take Santonio Holmes for example. If you think he will have 15 points per game in a redraft league, you shouldn't take him over someone healthy who will get you 13 ppg. It depends a bit on your roster but the 13 ppg will most likely have a higher expected value.

If the 2 guys are close, I agree that ppg is a better indicator of future value. Your example of Mendenhall/Williams demonstrates that but that only applies if you expect the players to play the same amount of games next year.

 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
I look at the whole picture because every situation is different. Even your Mendenhall example is flawed since PPG doesn't take into account the fact that he really wasn't handed the full time starting job until week 4. So by sorting PPG you get paltry stats from Weeks 1 and 2 worked in when he only had a few carries those games. Another example is Frank Gore. While his total points rank was high, points-per-game is higher because he missed two games. But that's not the whole story. Week 3 he only had 1 carry before getting injured. But the computer doesn't know that and it counts Week 3 for the PPG calculation. Additionally, when he came back week 7 he wasn't exaclty fully recovered and only put up 32 yards against a weak Texans defense. If you throw out Weeks 1 and 7 and recalculate the points-per-game numbers he finishes 2nd behind Chris Johnson. That's useful information considering his valuse seems to be fading for no apprent reason lately.When people (myself included) reference someone as finishing as RB14, or whatever, it is typically a general point of reference as opposed to grounds for making an argument related to value.Anyway, in general, I agree taht points per game is more useful than total points... but even more useful than that is having the entire picture and taking into account partial games and such.
 
Honestly, PPG can at times be deceiving, as well...

The #13 QB in my league in PPG last season? Dennis Dixon, with one game for 16.5 points.

The #26 WR was Craig Davis, with one game for 12.1 points.

You really have to look at a mixture of the 2. If a player misses a couple games (a la Frank Gore), his #5 RB point total looks even better on a PPG basis, where he was #3, over Ray Rice and ADP. But if that player also misses those games on a consistent basis as he has (2 games missed in 2009, 2008, 2005; 1 game missed in 2007), then you cannot automatically put his projections on that same PPG amount over the full 16 games.

I generally like to use PPG for players who get an opportunity at the end of the season to possibly see sleepers for next season, but over the entire season, I feel that total points are just as important.

 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
I look at the whole picture because every situation is different. Even your Mendenhall example is flawed since PPG doesn't take into account the fact that he really wasn't handed the full time starting job until week 4. So by sorting PPG you get paltry stats from Weeks 1 and 2 worked in when he only had a few carries those games. Another example is Frank Gore. While his total points rank was high, points-per-game is higher because he missed two games. But that's not the whole story. Week 3 he only had 1 carry before getting injured. But the computer doesn't know that and it counts Week 3 for the PPG calculation. Additionally, when he came back week 7 he wasn't exaclty fully recovered and only put up 32 yards against a weak Texans defense. If you throw out Weeks 1 and 7 and recalculate the points-per-game numbers he finishes 2nd behind Chris Johnson. That's useful information considering his valuse seems to be fading for no apprent reason lately.When people (myself included) reference someone as finishing as RB14, or whatever, it is typically a general point of reference as opposed to grounds for making an argument related to value.Anyway, in general, I agree taht points per game is more useful than total points... but even more useful than that is having the entire picture and taking into account partial games and such.
I totally agree with that, and I know PPG has its flaws as you pointed out. What I would really like to see is a program/spreadsheet that takes those types of things into account and gives you a value. Something that only calculates stats for when a player is starting or expected to get the majority of work. Something that could show if Mendenhall was more valuable per game once handed the starting spot.
 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
I look at the whole picture because every situation is different. Even your Mendenhall example is flawed since PPG doesn't take into account the fact that he really wasn't handed the full time starting job until week 4. So by sorting PPG you get paltry stats from Weeks 1 and 2 worked in when he only had a few carries those games. Another example is Frank Gore. While his total points rank was high, points-per-game is higher because he missed two games. But that's not the whole story. Week 3 he only had 1 carry before getting injured. But the computer doesn't know that and it counts Week 3 for the PPG calculation. Additionally, when he came back week 7 he wasn't exaclty fully recovered and only put up 32 yards against a weak Texans defense. If you throw out Weeks 1 and 7 and recalculate the points-per-game numbers he finishes 2nd behind Chris Johnson. That's useful information considering his valuse seems to be fading for no apprent reason lately.When people (myself included) reference someone as finishing as RB14, or whatever, it is typically a general point of reference as opposed to grounds for making an argument related to value.Anyway, in general, I agree taht points per game is more useful than total points... but even more useful than that is having the entire picture and taking into account partial games and such.
I totally agree with that, and I know PPG has its flaws as you pointed out. What I would really like to see is a program/spreadsheet that takes those types of things into account and gives you a value. Something that only calculates stats for when a player is starting or expected to get the majority of work. Something that could show if Mendenhall was more valuable per game once handed the starting spot.
I see where you are going with this, but with too many "ifs", you also run into trouble. For example, lets say someone could set up said spreadsheet for us. Barber starts, gets 15 carries for 65 yards. Felix Jones (who is not expected to start nor does he have more carries), goes for 10 carries for 85 yards and a TD. Now you have set the sheet up to dicount his efforts (because he did not start nor get the majority of the carries), so his game becomes a Null. In the end, there are just too many loops and "special" efforts that can't be calculated, or recalibrated on a week-by-week basis. I think people look at total points not only because it is easier, but because unless you get a Mendenhall/Greene case, they want to see how someone handles 16 games (if they can).Personally, I look at my "slots" when doing projections (which are very rough). I slot a guy in as my RB1 (say a Ray Rice). I expect him to play about 15 games/season for the upcoming season. Then I grab Ronnie Brown as my RB2...well i am taking a risk here, but if my RB3 (say I only start two), can put up 10 PPG and Brown can put up 16ppg, I am looking at a rough estimate of 13 PPG from said RB2 slot. It is a bonus if Brown can play the full season and hit 16 PPg, but since I am covering for injury (yeah, that is a real stretch with brown), I am getting X points from my RB2 slot which I estimate will be needed in order to make the playoffs.
 
The arrogant answer is "Because fantasy teams didn't score points on the games when they were out".The simple answer is "because that's how most league software sites rank players by default, and user's dont want the extra click to sort".The better answer is that neither is sufficient for complete analysis on its own. Some of the players you gave are good examples of this, as Ronnie certainly doesn't make many lists at RB10, but likely doesn't fall all the way to RB39 either.In the end what I really want is a guys with both high PPG and 16 games a year, so both stats are useful.
:kicksrock:
 
footballsavvy said:
Max Power said:
Michael Fox said:
Max Power said:
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
I look at the whole picture because every situation is different. Even your Mendenhall example is flawed since PPG doesn't take into account the fact that he really wasn't handed the full time starting job until week 4. So by sorting PPG you get paltry stats from Weeks 1 and 2 worked in when he only had a few carries those games. Another example is Frank Gore. While his total points rank was high, points-per-game is higher because he missed two games. But that's not the whole story. Week 3 he only had 1 carry before getting injured. But the computer doesn't know that and it counts Week 3 for the PPG calculation. Additionally, when he came back week 7 he wasn't exaclty fully recovered and only put up 32 yards against a weak Texans defense. If you throw out Weeks 1 and 7 and recalculate the points-per-game numbers he finishes 2nd behind Chris Johnson. That's useful information considering his valuse seems to be fading for no apprent reason lately.When people (myself included) reference someone as finishing as RB14, or whatever, it is typically a general point of reference as opposed to grounds for making an argument related to value.Anyway, in general, I agree taht points per game is more useful than total points... but even more useful than that is having the entire picture and taking into account partial games and such.
you must be psychic in order to dismiss week three like that.
 
This is pet peeve of mine too, especially when we get projections. I think every legitimate provider of fantasy projections should provide a projection for FPG and for games played. Yes, I know it's difficult to predict games played, but take a stand. If you're the type to believe that everyone plays 16 games, then say so. If you're the type that factors in injury risk, then show the level of risk you are predicting.

This is especially important this year because of suspensions to Holmes, Roethlisberger, and possibly VJax. When I see a Big Ben projection, I want to know if the projection is for 10 or 12 games. It makes a HUGE difference.

I would hope that Footballguys, as one of the leaders in the industry, would take a stand and include GPs and FPGs in the standard page for their projections. This isn't hard because their projections in Projections Dominator include this.

 
There is no single complete stat. You need to balance several and do some reworking on your own.

Players can be valued several ways in a PPG basis. You can look at last years actual PPG on a 16 or 17 game schedule (depending on your league). You can look at points per "expected game", games in which a player would reasonably have been inserted into a fantasy lineup (counts games where a guy was the named starter or 10-ish touch backup... if the regular starter got injured on the second play, that game counts, but the following weeks where he's known to be out wouldn't... and if it's a backup who never plays but gets thrown in when the starter is hurt, that first partial game wouldn't count but the following ones would until the starter is announced as coming back). You could do points per "draft expected" game (you draft a guy figuring he'll play in 13 or 14 games, but he goes out for the season after 2, knock his ppg average down accordingly). You could combine the above. Each one tells a different part of the story, but no single stat can tell the whole thing. When you distill information you can't re-expand it and still get it all back.

 
There are two main purposes for statistics- description, and prediction. If you're trying to use stats descriptively, then PPG isn't all that great. For instance, you can't just look at PPG and say that Ronnie Brown had the 10th best season, because the simple fact is that he DID miss those games, and it DID impact his fantasy owner. For description, I like to use an RB's total points, plus a number that represents what an average backup could have been expected to score in the games that the RB missed. To use Frank Gore as an example: Frank Gore scored 230.6 fantasy points in 14 games. What would an average backup score in the 2 games he missed? Well, RB30 last year (i.e. average RB3) scored 9.19 points per game. Just run a quick check to see if it passes the smell test- the RBs in the 28-32 range in PPG last year were Ahmad Bradshaw, Lawrence Maroney, Brandon Jacobs, Willis McGahee, and Carnell Williams, which makes sense- those are all guys who I could completely see a team having rostered as mediocre RB3s. So, anyway, give Frank Gore 9.19 times two games missed and add that to his season total and he would have scored... 249, which puts him in a virtual tie with Ray Rice for RB4. Seems like a totally fair representation of how much value Gore really provided last year. Giving Michael Turner a similar treatment puts him at 196.5, which would have ranked him as RB11 last year... which seems a bit high, but you have to remember that that's because you aren't giving the other RBs the same benefit. If you do it to all the other RBs, too, then Cedric Benson, Jamaal Charles, and DeAngelo Williams will jump back in front of Turner, dropping him down to 14th, which seems a bit more of a fair representation of how much value he provided last season.

If you're trying to use stats predictively, then I don't think PPG gives the necessary granularity. For instance, Frank Gore had a "game" where he got injured on his first carry and finished with 4 yards. I seem to recall Hester having a similar game. PPG will underrate players like that. Points per Quarter helps a lot, but that still underrates guys like Charles, Stewart, and Mendenhall, who saw their status change dramatically over the course of the year. And, of course, PPG overrates other guys who played such a small sample that they didn't have a chance to have some normal games to pull them back to the mean. When trying to predict future performance, I don't think PPG does that good of a job. I think you're better off sticking to the age-old "talent + opportunity" formula.

PPG is an interesting look, but it's not really any less flawed than total points, it's just differently flawed. As long as you take it with a grain of salt it can be a useful tool, but it's certainly not the "one stat to rule them all", so to speak.

 
13 game can be useful if thinking totals. Sometimes weeks 14-17 can alter things, not often but....

 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.

 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.
points per start? Unless a deep league...really?
 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.
points per start? Unless a deep league...really?
Yes, really. There's a good thread with a Retrospective Mock Draft that suggests that it's important to figure out players put up totals that make them worthy of being in your starting lineup. The challenge is, of course, in predicting that on a prospecive basis with any accuracy. But you can't do that at all without a prediction of games played.
 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.
points per start? Unless a deep league...really?
Yes, really. There's a good thread with a Retrospective Mock Draft that suggests that it's important to figure out players put up totals that make them worthy of being in your starting lineup. The challenge is, of course, in predicting that on a prospecive basis with any accuracy. But you can't do that at all without a prediction of games played.
give me an example please
 
Ppg is my favorite stat and if there was a points per start somewhere id obsess over it. the only real problem is how distorted from reality the small sample size guys are gonna be.

guys like miles austin and jamaal charles definitely won leagues for their owners. not sure exactly how to demonstrate that, but looking at a season total is totally foreign to figuring out what creates fantasy success.

 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.
points per start? Unless a deep league...really?
Yes, really. There's a good thread with a Retrospective Mock Draft that suggests that it's important to figure out players put up totals that make them worthy of being in your starting lineup. The challenge is, of course, in predicting that on a prospecive basis with any accuracy. But you can't do that at all without a prediction of games played.
give me an example please
Retrospective Mock Draft
 
I'm not saying PPG is the end all and be all stat. But it helps greatly in evaluating a picture. When FBG or any ther site doesn't predict Games Played, they prevent you from analyzing their prediction as well as possible. Let the user decide whether to conver to points per game, points per start, points per quarter, etc. But you need that important fact to get better picture of predictions.
points per start? Unless a deep league...really?
Yes, really. There's a good thread with a Retrospective Mock Draft that suggests that it's important to figure out players put up totals that make them worthy of being in your starting lineup. The challenge is, of course, in predicting that on a prospective basis with any accuracy. But you can't do that at all without a prediction of games played.
give me an example please
Retrospective Mock Draft
originally points per game and points per start were both listed. I get PPG being useful. What I don't see is PPS being so very useful, not in that thread either.If it's a part time player like a backup QB who started 5 games and is suddenly a starter in 2010, OK it's useful in that instance but with opportunity in FF you'd have been rolling with a player that is a new starting QB probably anyway. The resulting higher ranking is expected without even considering PPS.

When is PPS indicative of something you wouldn't have known just by sheer opportunity?

For lack of the right word here, I want to learn something for a stat that I took the time to research and add up or average.

I'm guessing some 3rd WR but the couple I hit at PFR weren't of any predictive value as these good players performed a smidge worse in 2009.

So, please give me an example.

 
footballsavvy said:
Max Power said:
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
What gives you the idea that smart people only use total points?
I tend to see end of year points thrown around more often to back arguements than PPG. EOY rankings seem to be the "go-to" stat around here.
I look at the whole picture because every situation is different. Even your Mendenhall example is flawed since PPG doesn't take into account the fact that he really wasn't handed the full time starting job until week 4. So by sorting PPG you get paltry stats from Weeks 1 and 2 worked in when he only had a few carries those games. Another example is Frank Gore. While his total points rank was high, points-per-game is higher because he missed two games. But that's not the whole story. Week 3 he only had 1 carry before getting injured. But the computer doesn't know that and it counts Week 3 for the PPG calculation. Additionally, when he came back week 7 he wasn't exaclty fully recovered and only put up 32 yards against a weak Texans defense. If you throw out Weeks 1 and 7 and recalculate the points-per-game numbers he finishes 2nd behind Chris Johnson. That's useful information considering his valuse seems to be fading for no apprent reason lately.When people (myself included) reference someone as finishing as RB14, or whatever, it is typically a general point of reference as opposed to grounds for making an argument related to value.Anyway, in general, I agree taht points per game is more useful than total points... but even more useful than that is having the entire picture and taking into account partial games and such.
you must be psychic in order to dismiss week three like that.
I'm not sure what you mean. I didn't predict anything. I saw that he only got 1 carry in Week 3 before getting injured. I therefore determined that it was innacurate to include week three in the point-per-game calculation. If you don't take this into account the avgerage calculated by the site will assume that he played the entire game and only managed to have 1 carry for 6 yards.
 
Chase Stuart said:
I find it a bit shocking to see who's ranked #10 on this list, fumbles and all.
That is a little weird, but it's a bit misleading IMO.Slaton is at #10, but the difference between #9 and #10 is greater than the difference between #10 and #16; there are seven RBs tightly packed between #10 and #17 there. Further, I'm using 0.5 PPR; if you play in a non-PPR league, Slaton moves down to RB21 in FP/4Q. If you're surprised to see Slaton so high, you probably don't play in a PPR league, and that article shouldn't cause you to think that Slaton is undervalued.

 
PPG is an interesting look, but it's not really any less flawed than total points, it's just differently flawed. As long as you take it with a grain of salt it can be a useful tool, but it's certainly not the "one stat to rule them all", so to speak.
:lol: PPG analysis is why Steven Jackson is constantly over-rated from year to year. He has a few "boom" games where he throws down 16 + points (he did this five times last year), but is mostly pedestrian-to-disappointing (12-13 points, 4 games; Under 12, 6 games, 0 points, one game). Basically, the Rams' surrounding offense is so lame that Jackson has trouble scoring TDs (this has been true for three years now, averaging only .5 TDs per game or less over 16 games, with a mere four TDs during 2009) even against just 1/2 decent defenses.The few "boom" games inflate his PPG ranking across the entire season, but doesn't reflect what his owners can expect most of the time.

To get a proper analysis of a fantasy player, you need to consider both PPG and total points. Jeff Pasquino's Quality (insert position here) articles give us another way to break down any given fantasy player's year: LINK to 2009 Article . In that paradigm, you see that Jackson's 2008 season was tied for 12th among all fantasy RBs.

In the end, no one rear-ward looking stat analysis will paint a full picture for future production as there are intangible issues from year to year that impact production going forwards (injury/rehab; coaching changes; changes along an offensive line; off-field issues like Ronnie Browns' Georgia DUI arrest earlier this offseason, new personnel arriving via the NFL draft, etc). Crafting a set of projections and rankings is partly based on past production/track record, but must "fold in" the intangible issues that aren't accounted for by raw stats.

My primary example of the "intangibles" factor this year is the Washington Redskins RBs during Shanahan year one. RBBC? Portis to lead with some support from Parker/Johnson? Johnson leads? Parker leads? At this point, I don't think anyone has a handle on how to project the Washington backfield. What I did is averaged the rushing yards that Denver backs produced during Shanahan's entire Denver tenure and used that as a baseline for projecting what I initially identified as a RBBC jumble between Portis, Johnson and Parker. But all three projections are really placeholders until we get to training camp/preseason and see who, if any, of the Washington backs emerge at the top of the depth chart.

My .02.

 
you must be psychic in order to dismiss week three like that.
I'm not sure what you mean. I didn't predict anything. I saw that he only got 1 carry in Week 3 before getting injured. I therefore determined that it was innacurate to include week three in the point-per-game calculation. If you don't take this into account the avgerage calculated by the site will assume that he played the entire game and only managed to have 1 carry for 6 yards.
in order to dismiss the data, you'd have to know that he was going to miss the game. Since he misses one or more games every year, then you have to take into account the bad game since you're going to start him... unless you're psychic.
 
PPG is an interesting look, but it's not really any less flawed than total points, it's just differently flawed. As long as you take it with a grain of salt it can be a useful tool, but it's certainly not the "one stat to rule them all", so to speak.
:rolleyes: PPG analysis is why Steven Jackson is constantly over-rated from year to year. He has a few "boom" games where he throws down 16 + points (he did this five times last year), but is mostly pedestrian-to-disappointing (12-13 points, 4 games; Under 12, 6 games, 0 points, one game). Basically, the Rams' surrounding offense is so lame that Jackson has trouble scoring TDs (this has been true for three years now, averaging only .5 TDs per game or less over 16 games, with a mere four TDs during 2009) even against just 1/2 decent defenses.The few "boom" games inflate his PPG ranking across the entire season, but doesn't reflect what his owners can expect most of the time.
These kind of up and down games usually happen to RBs when you are not playing PPR...compare him to any of his peers and he is right inline with them. He only had 4 TDs last year. Do you think his total is that low again?What made you choose those benchmark numbers for pedestrian vs. disappointing PPG? I'd like to see this same analysis with Gore (6 games under 13 points), Jones-Drew (7 games under 12 points), and DeAngelo (6 of his 13 games 12 points or lower)Basically what your saying is RBs that don't score a TD every game are erratic in non-ppr fantasy football. Correct?
 
you must be psychic in order to dismiss week three like that.
I'm not sure what you mean. I didn't predict anything. I saw that he only got 1 carry in Week 3 before getting injured. I therefore determined that it was innacurate to include week three in the point-per-game calculation. If you don't take this into account the avgerage calculated by the site will assume that he played the entire game and only managed to have 1 carry for 6 yards.
in order to dismiss the data, you'd have to know that he was going to miss the game. Since he misses one or more games every year, then you have to take into account the bad game since you're going to start him... unless you're psychic.
I see what you're saying. But I still think that's a bit convoluted. I prefer to see a clear picture of what a back is capable of when heplays a full game on a consistent basis. Then I can downgrade that ranking based on injury risk, etc. I think it's just as unpredictable to assume he's going to have a game in 2010 where he has 1 carry and leaves the game. He might have 4 games where he does that (or maybe only plays half the game), or he might have a full healthy season. You just don't know. So you look at his production when healthy, compare it to other backs production when healthy, and discount for injury risk after the fact.
 
I really don't understand this. We routinely see posters use the phrase "Player X finished as the #__ ranked running back." Seems to me the major flaw in that line of thought is that not all players play in 16 games. Also a good amount of leagues end in week 16 making week 17 stats irrelevant.On a PPG basis Ronnie Brown was a top 10 back (PPR). His end of year ranking 39. Michael Turner's PPG ranked him 17, his end of year ranking was 34. Obviously the stats get skewed the most when players miss several games, as in these cases. Something a little closer to compare. Rashard Mendenhall finshed the year as the 16th ranked RB with 210.0 points. DeAngelo Williams finished the year as the 17th ranked RB with 209.9 points. .1 no big deal right? Yet on the PPG scale, Mendenhall was 24th at 13.12, and Williams was 8th at 16.15. Now 3 points per game is a notable difference.Points per game seems like a more accurate way to rank players, don't you think?
:lmao: W/o reading the rest of the posts yet, 100% agree w/ this.It may not be the prevailing factor, but its a factor.It's a simple, rhetorical question you asked, but well worth the post.It is trully a much better indicator to plan off of for the nxt season.It's pretty much the only way I sort players when studying up or evaluating.It's w/i reason of course, a player cant miss excessive games and be judged by ppg.It is fairly obvious to the savvy owner, but you make a great point for the more inexperienced owners to note.You have to evaluate consistency and ppg is a great way to do it.
 
Last edited by a moderator:
you must be psychic in order to dismiss week three like that.
I'm not sure what you mean. I didn't predict anything. I saw that he only got 1 carry in Week 3 before getting injured. I therefore determined that it was innacurate to include week three in the point-per-game calculation. If you don't take this into account the avgerage calculated by the site will assume that he played the entire game and only managed to have 1 carry for 6 yards.
in order to dismiss the data, you'd have to know that he was going to miss the game. Since he misses one or more games every year, then you have to take into account the bad game since you're going to start him... unless you're psychic.
I see what you're saying. But I still think that's a bit convoluted. I prefer to see a clear picture of what a back is capable of when heplays a full game on a consistent basis. Then I can downgrade that ranking based on injury risk, etc. I think it's just as unpredictable to assume he's going to have a game in 2010 where he has 1 carry and leaves the game. He might have 4 games where he does that (or maybe only plays half the game), or he might have a full healthy season. You just don't know. So you look at his production when healthy, compare it to other backs production when healthy, and discount for injury risk after the fact.
oh, my bad.I think that you're saying that you'd like to create a ceiling for him with this data? Yes, I can see how you'd keep the injury risk as a separate variable and factor it in at a different point.

 
PPG is an interesting look, but it's not really any less flawed than total points, it's just differently flawed. As long as you take it with a grain of salt it can be a useful tool, but it's certainly not the "one stat to rule them all", so to speak.
:unsure: PPG analysis is why Steven Jackson is constantly over-rated from year to year. He has a few "boom" games where he throws down 16 + points (he did this five times last year), but is mostly pedestrian-to-disappointing (12-13 points, 4 games; Under 12, 6 games, 0 points, one game). Basically, the Rams' surrounding offense is so lame that Jackson has trouble scoring TDs (this has been true for three years now, averaging only .5 TDs per game or less over 16 games, with a mere four TDs during 2009) even against just 1/2 decent defenses.The few "boom" games inflate his PPG ranking across the entire season, but doesn't reflect what his owners can expect most of the time.
These kind of up and down games usually happen to RBs when you are not playing PPR...compare him to any of his peers and he is right inline with them. He only had 4 TDs last year. Do you think his total is that low again?What made you choose those benchmark numbers for pedestrian vs. disappointing PPG? I'd like to see this same analysis with

Gore (6 games under 13 points), Jones-Drew (7 games under 12 points), and DeAngelo (6 of his 13 games 12 points or lower)

Basically what your saying is RBs that don't score a TD every game are erratic in non-ppr fantasy football. Correct?
Not at all, I was addressing the notion that PPG is a superior yardstick vs. total points. It isn't. It is a tool, but is only one of those in the box (just as Quality games is another useful tool, it isn't the "superior" method). Regarding Jackson 2010, I think mid single digits is his ceiling in TDs with a rookie QB under center behind that awful line, with no credible, experienced pro receivers to throw to (Donnie Avery, Brandon Gibson, Laurent Robinson, Mardy Gilyard, Keenan Burton (inj), Danny Amendola (PR), Brooks Foster (inj), Nick Moore, Jordan Kent = YUCK).

I think Jackson's PPR value last year was inflated by an unusually high number of passes (due to the fact that most of the WR corps was walking wounded all year long). I think those numbers come back down more in line with his career norms (55-60 targets and 40ish receptions per year), and probably dip below that territory somewhat (say 45-50 targets and 35 receptions) due to the fact that Bradford is going to stink in year one, spend a lot of time on his back after getting drilled by blitzers, like most rookies thrown into the NFL fire do.

Bradford 2010 reminds me of David Carr 2002 - and I think the Rams' offense will look similar to the 2002 Texans in output when it's all said and done : 32nd in the league in passing (2225 yards), 32nd in TD passes (11), 31st in rushing (1347), 31st in rushing TDs (6) was what they managed that year. Jackson will rush more effectively than that Texans' team did, but I think his scoring opportunities (trips inside the 20) will ultimately be similarly depressingly infrequent.

The above is why I am pessimistic about Jackson's chances to finish among the top tier of fantasy RBs during 2010.

 
Not at all, I was addressing the notion that PPG is a superior yardstick vs. total points. It isn't. It is a tool, but is only one of those in the box (just as Quality games is another useful tool, it isn't the "superior" method).

Regarding Jackson 2010, I think mid single digits is his ceiling in TDs with a rookie QB under center behind that awful line, with no credible, experienced pro receivers to throw to (Donnie Avery, Brandon Gibson, Laurent Robinson, Mardy Gilyard, Keenan Burton (inj), Danny Amendola (PR), Brooks Foster (inj), Nick Moore, Jordan Kent = YUCK).

I think Jackson's PPR value last year was inflated by an unusually high number of passes (due to the fact that most of the WR corps was walking wounded all year long). I think those numbers come back down more in line with his career norms (55-60 targets and 40ish receptions per year), and probably dip below that territory somewhat (say 45-50 targets and 35 receptions) due to the fact that Bradford is going to stink in year one, spend a lot of time on his back after getting drilled by blitzers, like most rookies thrown into the NFL fire do.

Bradford 2010 reminds me of David Carr 2002 - and I think the Rams' offense will look similar to the 2002 Texans in output when it's all said and done : 32nd in the league in passing (2225 yards), 32nd in TD passes (11), 31st in rushing (1347), 31st in rushing TDs (6) was what they managed that year. Jackson will rush more effectively than that Texans' team did, but I think his scoring opportunities (trips inside the 20) will ultimately be similarly depressingly infrequent.

The above is why I am pessimistic about Jackson's chances to finish among the top tier of fantasy RBs during 2010.
Those 40ish reception years were when he played in 12 games. He's been a top 15 (non-ppr) RB for 5 years in a row in pretty bad situations and not playing all the games. The situation certainly can't get any worse than last year with Keith Null playing QB. Its fine to say Bradford will struggle but did you see the QB play last year? It was historically terrible. That 2002 Texans team scored 48 more points than the 2009 Rams

You have been so down on Jackson for so long I'm sure at some point he'll live down to your standards but for 5 years in a row (I assume, don't recall all your projections) he has outplayed your projections...it might be time to adjust.

He is one of about 4 or 5 RB left that dominate their teams touches which makes him have a very high floor. To say his PPG production is too erratic is not accurate, in my opinion anyway.

 
Not at all, I was addressing the notion that PPG is a superior yardstick vs. total points. It isn't. It is a tool, but is only one of those in the box (just as Quality games is another useful tool, it isn't the "superior" method).

Regarding Jackson 2010, I think mid single digits is his ceiling in TDs with a rookie QB under center behind that awful line, with no credible, experienced pro receivers to throw to (Donnie Avery, Brandon Gibson, Laurent Robinson, Mardy Gilyard, Keenan Burton (inj), Danny Amendola (PR), Brooks Foster (inj), Nick Moore, Jordan Kent = YUCK).

I think Jackson's PPR value last year was inflated by an unusually high number of passes (due to the fact that most of the WR corps was walking wounded all year long). I think those numbers come back down more in line with his career norms (55-60 targets and 40ish receptions per year), and probably dip below that territory somewhat (say 45-50 targets and 35 receptions) due to the fact that Bradford is going to stink in year one, spend a lot of time on his back after getting drilled by blitzers, like most rookies thrown into the NFL fire do.

Bradford 2010 reminds me of David Carr 2002 - and I think the Rams' offense will look similar to the 2002 Texans in output when it's all said and done : 32nd in the league in passing (2225 yards), 32nd in TD passes (11), 31st in rushing (1347), 31st in rushing TDs (6) was what they managed that year. Jackson will rush more effectively than that Texans' team did, but I think his scoring opportunities (trips inside the 20) will ultimately be similarly depressingly infrequent.

The above is why I am pessimistic about Jackson's chances to finish among the top tier of fantasy RBs during 2010.
Those 40ish reception years were when he played in 12 games. He's been a top 15 (non-ppr) RB for 5 years in a row in pretty bad situations and not playing all the games. The situation certainly can't get any worse than last year with Keith Null playing QB. Its fine to say Bradford will struggle but did you see the QB play last year? It was historically terrible. That 2002 Texans team scored 48 more points than the 2009 Rams

You have been so down on Jackson for so long I'm sure at some point he'll live down to your standards but for 5 years in a row (I assume, don't recall all your projections) he has outplayed your projections...it might be time to adjust.

He is one of about 4 or 5 RB left that dominate their teams touches which makes him have a very high floor. To say his PPG production is too erratic is not accurate, in my opinion anyway.
Actually, he's played down to my expectations for the past three seasons, and disappointed given most others' projections/rankings. He's finished 14th, 13th, and 10th in total points (non PPR) since his SINGLE top-5 appearance during 2006, when he had a ridiculous 436 touches on the ball and 16 TDs, (with 111 targets). He's never even come CLOSE to that level of touches in years since - even last year, when he WAS the offense in St. Louis, he "only" managed 375 touches. He was 8th in total points in the PPR format last year among RBs, even with the 375 touches to his credit (LINK).By the way, the 2009 Rams' offense scored 16 total TDs, and the 2002 Texans' offense scored 17. So unless you somehow include field goals or defensive points scored (for that 48 point gap you cited) in the Texans' 2002 production at the non-PK skill positions, they were approximately equal in historically bad fantasy performance from the non-PK skill positions.

And I expect the 2010 Rams to be equally inept or even worse on offense with Bradford under center. Which will definitely impact Jackson's ceiling in many ways. I don't see a high floor here, especially given a late-April back surgery. In terms of total touches and yardage production week-to-week, I expect Jackson to struggle to match 2009 - he may have a few more TDs (mid-single-digits), but double-digits are likely out of reach IMO.

:goodposting: End of thread hijack.

 
This is pet peeve of mine too, especially when we get projections. I think every legitimate provider of fantasy projections should provide a projection for FPG and for games played. Yes, I know it's difficult to predict games played, but take a stand. If you're the type to believe that everyone plays 16 games, then say so. If you're the type that factors in injury risk, then show the level of risk you are predicting.This is especially important this year because of suspensions to Holmes, Roethlisberger, and possibly VJax. When I see a Big Ben projection, I want to know if the projection is for 10 or 12 games. It makes a HUGE difference.I would hope that Footballguys, as one of the leaders in the industry, would take a stand and include GPs and FPGs in the standard page for their projections. This isn't hard because their projections in Projections Dominator include this.
Good points here. In my projections, I have always projected games played. I fully realize there are different camps on this subject, but I have always felt more comfortable projecting some risk/tolerance into my numbers. For suspensions, it's not a gray area as we know players like Holmes, Roethlisberger, etc will miss X games. So, projecting the difference is just math. Predicting injuries is a slippery slope, but at least IMO, it helps one account for risk and to a lesser extent the likelihood a player will miss any games based on things like past injury history, age, etc..Where that comes in handy for me is differentiating between projections and rankings. Take Roethlisberger as an example. If he's projected to score within a reasonable amount of points to other QBs playing 16 games, then on a PPG basis, I'd much prefer Roethlisberger knowing that I'll have to have a starter for those other games he'll miss. Yet down the stretch, I'll have a player scoring 3-4 more PPG on my roster than a similarly projected QB (spanning 16 games).I don't know how easy/difficult it is to implement the GP column for the projections on the site, but I'll check to see if it's possible. There is an awful lot going on right now with respect to the site redesign that this might need to get moved out until things settled down, but it's a good suggestion IMO.On the larger topic of the thread, using PPG, total points, SOS and other stats is important. To best analyze players and differentiate between them on draft day we need to look at them through various lenses. PPG is an important one, but it has to be done in the context of total pts scored, schedule, etc. Ignoring it would be a shame.
 
This is pet peeve of mine too, especially when we get projections. I think every legitimate provider of fantasy projections should provide a projection for FPG and for games played. Yes, I know it's difficult to predict games played, but take a stand. If you're the type to believe that everyone plays 16 games, then say so. If you're the type that factors in injury risk, then show the level of risk you are predicting.This is especially important this year because of suspensions to Holmes, Roethlisberger, and possibly VJax. When I see a Big Ben projection, I want to know if the projection is for 10 or 12 games. It makes a HUGE difference.I would hope that Footballguys, as one of the leaders in the industry, would take a stand and include GPs and FPGs in the standard page for their projections. This isn't hard because their projections in Projections Dominator include this.
Good points here. In my projections, I have always projected games played. I fully realize there are different camps on this subject, but I have always felt more comfortable projecting some risk/tolerance into my numbers. For suspensions, it's not a gray area as we know players like Holmes, Roethlisberger, etc will miss X games. So, projecting the difference is just math. Predicting injuries is a slippery slope, but at least IMO, it helps one account for risk and to a lesser extent the likelihood a player will miss any games based on things like past injury history, age, etc..Where that comes in handy for me is differentiating between projections and rankings. Take Roethlisberger as an example. If he's projected to score within a reasonable amount of points to other QBs playing 16 games, then on a PPG basis, I'd much prefer Roethlisberger knowing that I'll have to have a starter for those other games he'll miss. Yet down the stretch, I'll have a player scoring 3-4 more PPG on my roster than a similarly projected QB (spanning 16 games).I don't know how easy/difficult it is to implement the GP column for the projections on the site, but I'll check to see if it's possible. There is an awful lot going on right now with respect to the site redesign that this might need to get moved out until things settled down, but it's a good suggestion IMO.On the larger topic of the thread, using PPG, total points, SOS and other stats is important. To best analyze players and differentiate between them on draft day we need to look at them through various lenses. PPG is an important one, but it has to be done in the context of total pts scored, schedule, etc. Ignoring it would be a shame.
:coffee: Thank you, Bob! I think it should be relatively easy because in Projections Dominator, you can see games projected by each staffer. So it must be in the system that FBG uses to do projections. I think you could just add two columns to the web versions: GP and FP/G. I don't think it would greatly alter web layout of those pages.Eagerly waiting to see your projections this year, Bob.Thanks again.
 
I don't get it. Who cares how Jackson finishes in total points. Unless you consistently start him, even when he doesn't play, it just doesn't matter. The main (I won't say 'only' because SSOG's point about the fill-in not being as good is a good point) stat is PPG. Who the hell cares what SJax scores over the course of a season. I only care what he does during the games he starts.

Nobody knows how many games a player will lose to injury (inmost cases). Why would you ever include that in your projections.

 

Users who are viewing this thread

Back
Top