What's new
Fantasy Football - Footballguys Forums

Welcome to Our Forums. Once you've registered and logged in, you're primed to talk football, among other topics, with the sharpest and most experienced fantasy players on the internet.

Total Points or Points Per Game? (1 Viewer)

Do you prefer Total Points or Points Per Game when ranking players?

  • Total Points

    Votes: 0 0.0%
  • Points Per Game

    Votes: 0 0.0%
  • Ranking Players is for Sissies

    Votes: 0 0.0%

  • Total voters
    0

GroveDiesel

Footballguy
I've seen a couple of guys advocating the use of PPG now. Just interested on getting some opinions and seeing some some arguments for and against each.

Personally I'm still leaning towards total points. PPG would seem to ignore injury risk completely. It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.

 
I voted PPG because I want to know what kind of product to 'expect' on a weekly basis. Total points gives you no indication of what their consistency level was. What if they had 3 big weeks to end the season. Doesn't help your team much if you didn't get production early on and are now out of the playoffs.

 
Maybe I'm just confused at how you're calculating PPG.

Isn't PPG just total points divided by number of games played? Or are you using a different way to average weekly scores or something?

Because if it's just total points divided by number of games played, isn't that the same as total points only factoring out injuries?

 
Because if it's just total points divided by number of games played, isn't that the same as total points only factoring out injuries?
In a word - yes.Factoring out injuries is extremely important in getting relevant comparisons between players - IMNSHO.
 
GroveDiesel,

Yes it is but think of it this way. Let's assume a 10 game season (simply because I'm too lazy this am to do any real math).

Do you want a guy who score 100 points for the over all season or a guy who will get you 10 pts per week for 10 games? I want the guy who will get me the 10 every week, not the player who may get me 15 one week and 5 the next.

I typically do my projections and then do 2 tables. One that sorts by total points and another that sorts by PPG. when you sort by PPG you will see alot of other talent jump up the rankings. I look at it as a way to get some players that may be way down someone elses board because their TP didn't rank them that high.

 
PPG would seem to ignore injury risk completely. It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.
Total Points includes all injuries from last year. If you're going to use those numbers, you are building in the assumption injuries next year will duplicate what happened in the previous year. Which doesn't seem to me a good assumption.It isn't a problem that PPG "ignores" injury risk. The opposite -- it is a benefit, you don't want it in there for the reason I just stated. The problem is if YOU ignore injury risk. PPG removes all injury risk so you can add it back in some meaningful fashion that matches what you think it should be... not what it was last year.

PPG also has to be applied meaningfully, with some kind of threshhold for number of games played. But the same is true with Total Points. Neither should be used blindly.

And PPG has the obvious benefit of quantifying how much LJ outperformed a guy who played 16 games and scored the same FP when LJ wasn't the starter.

I really think you need to look at both, though a well-put together PPG analysis is probably more meaningful. The only big detraction for it is that it is more work. You can't just divide by number of games if your goal is a quality analsys. You need to do something with games where the guy played, but may not have been used as a contributor like you expect him to the next year.

 
Maybe I'm just confused at how you're calculating PPG.

Isn't PPG just total points divided by number of games played?  Or are you using a different way to average weekly scores or something?

Because if it's just total points divided by number of games played, isn't that the same as total points only factoring out injuries?
It's useful for evaluating guys that are injured and play in lesser games. It's not useful for guys that are active in many games and come on late....say the starter designated in week #11 after playing special teams for the first ten weeks.Just another tool, like Last 6 Weeks, Last 3 Weeks sorts. You get a different chocolate from each sort.

As to which is more important: it's how you assign players a value and rank them in a draft scenario that counts. If Marc Bulger averaged 22.5 pts/game and you want to rank him, do you use this information over a 16 game or a 14 game multiplier to establish his total season points scored?

You rate him as a 16 game player, rank him as the #5 QB on the board. If you draft him, he'll probably come as a value pick after you've picked up a top 10 QB as your #1 starter......cause he's a true #1 starter that's injury prone, good for an average of about 12.5 games for the last two years, and a playoff killer if he's the only #1 calibur QB on your roster.

It's all in how you evaluate them and where you obtain a weekly average extended to a 16 game schedule that counts. Is Bradie James LB-DAL draftable as the #30 LB on your board? If so, look at the previous years sort information, find the points scored by the 30th, or so LB, and insert his tag about there in the Dominator.

He'll appear right when you need some value.... and be relative to the rest of the players still on the board.

:banned: Drink The Cool-Aide

:popcorn:

 
Last edited by a moderator:
I was in a 16 team dynasty league with a full 53 man roster and now I'm in a 32 team 40 man roster league and both were supposed to be full of "sharks". They kept making fun of me because I had a roster full of players that averaged mediocre points. My team was in the middle or bottom of the pack in total points scored for the year but they made a big deal out of it and said my team stinks and nobody would trade me. I ONLY ended up winning the championship my first year in the 16 team one and made it to the conference championship in my second year with those "middle of the pack" guys. I kept trying to tell them that I don't want those guys that are gonna put up 100 points one week, then 6 the next. I'll take the guys that averaged 40 points week in and week out. And that's what worked. I'm following that same plan for my 32 team dynasty that I'm starting this year.

 
May we should just sign up in each others leagues and both pick up another team.

NOT ANOTHER TEAM!....AGGGG!

Yup, the total ammunition that you have, the more points you score. Depends on where the points come from, too.....an equally balanced scoring system is the challenge.

I've got one that's just that way so if you've got an open mind and are still considering where to come up with a creative system, PM me.

I fact, I'll PM you with a link so that you can see what's up.

 
Fantasy leagues are almost all scored and won weekly not yearly. That's why ppg has always been and always will be more important. If a player is injured, your fantasy team doesnt stop scoring points at that position. You simply put in another starter. Therefore your yearly output for that position isnt limited to the number of games played by any particuliar player.

Let's throw some math in here.

Team A's first string QB scores 21 ppg but only plays 8 weeks for a yearly total of only 168 points. His backups average 16 ppg over the other 7 weeks. His team's yearly total is 280 points even though his QB1 only scored 168.

Team B's first string QB plays in every game scoring 270 points for the year. That means his starting QB scored over 100 fantasy points more than Team A. Yet his team's output was the same 270 which is 10 fantasy points less than team A for the year. So while QB B outscored QB A, fantasy team A outscored fantasy team B.

Team B's QB1 averaged 18ppg. That's 3 fewer than team A's. What's even more relevant than team A outscoring team B over the season is that team A averaged 3 QB ppg more than team B did for 8 weeks of the season while team B averaged 2 QB ppg for 7 weeks of the season. Not talking about playoff performances, which team was more likely to pull out wins over his weekly opponent due to his QB? The answer is team A.

Of course the playoffs are the most important but that is only affected by injury if a QB is injured those weeks (the majority of injuries arent season ending). The playoffs are also even more affected by matchup. Even if team B's QB1 is healthy during the playoffs and team A's isnt, it doesnt help team B if his QB at the time has an equal or a worse matchup than team A's backup. Not to mention simple logic like having to get to the playoffs first, not being able to predict playoff defenses before the season starts, and the ability of any and all teams to aquire a playoff specific QB as the season progresses.

 
PPG would seem to ignore injury risk completely. It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.
Total Points includes all injuries from last year. If you're going to use those numbers, you are building in the assumption injuries next year will duplicate what happened in the previous year. Which doesn't seem to me a good assumption.It isn't a problem that PPG "ignores" injury risk. The opposite -- it is a benefit, you don't want it in there for the reason I just stated. The problem is if YOU ignore injury risk. PPG removes all injury risk so you can add it back in some meaningful fashion that matches what you think it should be... not what it was last year.

PPG also has to be applied meaningfully, with some kind of threshhold for number of games played. But the same is true with Total Points. Neither should be used blindly.

And PPG has the obvious benefit of quantifying how much LJ outperformed a guy who played 16 games and scored the same FP when LJ wasn't the starter.

I really think you need to look at both, though a well-put together PPG analysis is probably more meaningful. The only big detraction for it is that it is more work. You can't just divide by number of games if your goal is a quality analsys. You need to do something with games where the guy played, but may not have been used as a contributor like you expect him to the next year.
That makes sense and I can buy this argument.Maybe the best solution is to do some type of hybrid of the two. IMO, that would be most effective if you used stats over several years and maybe weighted them more heavily for the more current seasons. That way you end up with one system that takes some injury risk into account and one system that takes into account performance when healthy and up and coming players.

And of course that's always going to be just a starting point. Adjustments based on the situation in the new season are always part of the equation.

I'm still confused with the assertion from other posters that it can give you more consistent scoring. Unless I'm missing something, PPG does no such thing. To figure out something like that you'd need to do a much more complex calculation using some sort of median average instead of the mean average.

 
Fantasy leagues are almost all scored and won weekly not yearly. That's why ppg has always been and always will be more important. If a player is injured, your fantasy team doesnt stop scoring points at that position. You simply put in another starter. Therefore your yearly output for that position isnt limited to the number of games played by any particuliar player.

Let's throw some math in here.

Team A's first string QB scores 21 ppg but only plays 8 weeks for a yearly total of only 168 points. His backups average 16 ppg over the other 7 weeks. His team's yearly total is 280 points even though his QB1 only scored 168.

Team B's first string QB plays in every game scoring 270 points for the year. That means his starting QB scored over 100 fantasy points more than Team A. Yet his team's output was the same 270 which is 10 fantasy points less than team A for the year. So while QB B outscored QB A, fantasy team A outscored fantasy team B.

Team B's QB1 averaged 18ppg. That's 3 fewer than team A's. What's even more relevant than team A outscoring team B over the season is that team A averaged 3 QB ppg more than team B did for 8 weeks of the season while team B averaged 2 QB ppg for 7 weeks of the season. Not talking about playoff performances, which team was more likely to pull out wins over his weekly opponent due to his QB? The answer is team A.

Of course the playoffs are the most important but that is only affected by injury if a QB is injured those weeks (the majority of injuries arent season ending). The playoffs are also even more affected by matchup. Even if team B's QB1 is healthy during the playoffs and team A's isnt, it doesnt help team B if his QB at the time has an equal or a worse matchup than team A's backup. Not to mention simple logic like having to get to the playoffs first, not being able to predict playoff defenses before the season starts, and the ability of any and all teams to aquire a playoff specific QB as the season progresses.
But doesn't your scenario mean that Team B had an extra roster spot or that he was able to draft his #2 or #3 WR in the round that Team A drafted his #2 QB? Having 2 QBs combine to score more points than another team's 1 QB doesn't mean that Team A didn't sacrifice at another position to do so.
 
I'm still confused with the assertion from other posters that it can give you more consistent scoring. Unless I'm missing something, PPG does no such thing. To figure out something like that you'd need to do a much more complex calculation using some sort of median average instead of the mean average.
I agree, knowing someone scored 10 PPG doesn't tell you any more about the distribution of his points that knowing he scored 160 total points does. Knowing the standard deviation would help with that though, with a lower number relative to the size of his PPG meaning more of his scores fall close to the average than not.One other note on that though, studies done by FBGs (Drinen I believe) found that consistency in one year isn't a predictor of consistency in the next. So I personally don't much worry about consistency in scoring so much. I would focus more on how I see the player being used the next year, and is it similar to how he was used in the past.

 
PPG would seem to ignore injury risk completely.  It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.
Total Points includes all injuries from last year. If you're going to use those numbers, you are building in the assumption injuries next year will duplicate what happened in the previous year. Which doesn't seem to me a good assumption.It isn't a problem that PPG "ignores" injury risk. The opposite -- it is a benefit, you don't want it in there for the reason I just stated. The problem is if YOU ignore injury risk. PPG removes all injury risk so you can add it back in some meaningful fashion that matches what you think it should be... not what it was last year.

PPG also has to be applied meaningfully, with some kind of threshhold for number of games played. But the same is true with Total Points. Neither should be used blindly.

And PPG has the obvious benefit of quantifying how much LJ outperformed a guy who played 16 games and scored the same FP when LJ wasn't the starter.

I really think you need to look at both, though a well-put together PPG analysis is probably more meaningful. The only big detraction for it is that it is more work. You can't just divide by number of games if your goal is a quality analsys. You need to do something with games where the guy played, but may not have been used as a contributor like you expect him to the next year.
That makes sense and I can buy this argument.Maybe the best solution is to do some type of hybrid of the two. IMO, that would be most effective if you used stats over several years and maybe weighted them more heavily for the more current seasons. That way you end up with one system that takes some injury risk into account and one system that takes into account performance when healthy and up and coming players.

And of course that's always going to be just a starting point. Adjustments based on the situation in the new season are always part of the equation.

I'm still confused with the assertion from other posters that it can give you more consistent scoring. Unless I'm missing something, PPG does no such thing. To figure out something like that you'd need to do a much more complex calculation using some sort of median average instead of the mean average.
It's actually quite simple. Let's say you're doing research on possible players to draft. Let's use a 5 week schedule for length purpose. Here's an example:Player A

Game 1 Game 2 Game 3 Game 4 Game 5

50 pts 5 pts 100 pts 15 pts 15 pts

total points = 185 points in 5 games

Man, he probably helped you win in week 1 and week 3. But on those down weeks you better hope someone else stepped up and scored more points than they usually do.

Player B

Game 1 Game 2 Game 3 Game 4 Game 5

35 pts 40 pts 30 pts 40 pts 30 pts

total points = 175 points in 5 games

Even though player B scored 10 points less in that span he was a more consistent scorer and you can put him in your lineup and get a good nights sleep. If your team is composed of "consistent scorers" you have a better chance to win. I never look at total points for the year. I always look at how many points a guy averaged per game. :thumbup:

 
I was in a 16 team dynasty league with a full 53 man roster and now I'm in a 32 team 40 man roster league and both were supposed to be full of "sharks". They kept making fun of me because I had a roster full of players that averaged mediocre points. My team was in the middle or bottom of the pack in total points scored for the year but they made a big deal out of it and said my team stinks and nobody would trade me. I ONLY ended up winning the championship my first year in the 16 team one and made it to the conference championship in my second year with those "middle of the pack" guys. I kept trying to tell them that I don't want those guys that are gonna put up 100 points one week, then 6 the next. I'll take the guys that averaged 40 points week in and week out. And that's what worked. I'm following that same plan for my 32 team dynasty that I'm starting this year.
Again, I think you're talking about something like average per week instead of just points per game. Points per game would just be Points/Games. So a guy that scores 15 one game and 5 the next has the exact same value as a guy that scores 10 each game.

I'm not very good with statistics so I don't know the formula, but figuring out which guys are more consistent seems like a much more complex calculation. Something like figuring out what the standard deviation for each player is or something like that.

ETA: GregR beat me to it. What you're talking about is comparing standard deviations from their PPG which is a more complex calculation. I agree with the belief that it's better to have more consistent scorers, but PPG alone doesn't do that. I'm also with GregR in that predicting consistency is difficult to do. It's nice to know which guys are more consistent, but outside of the RB position I think it's almost useless.

 
Last edited by a moderator:
Fantasy leagues are almost all scored and won weekly not yearly. That's why ppg has always been and always will be more important. If a player is injured, your fantasy team doesnt stop scoring points at that position. You simply put in another starter. Therefore your yearly output for that position isnt limited to the number of games played by any particuliar player.

....
But doesn't your scenario mean that Team B had an extra roster spot or that he was able to draft his #2 or #3 WR in the round that Team A drafted his #2 QB? Having 2 QBs combine to score more points than another team's 1 QB doesn't mean that Team A didn't sacrifice at another position to do so.
I think this is a good FF subject, though I don't think it has much bearing on the worth of PPG vs Total Points. But in a general sense, I think there are things people can do to improve their valuation of backups vs starters.For example, there's an article out there that goes over how many games on average a starting player at each position misses. Between that and your league's parameters, you can figure out how much on average your backup QB/RB/WR will play. And knowing that, you can better decide if you should take a backup RB now who scores X, or take the #2 or #3 WR now and take the backup next round.

 
Last edited by a moderator:
If your team is composed of "consistent scorers" you have a better chance to win. I never look at total points for the year. I always look at how many points a guy averaged per game. :thumbup:
Which would indeed be good, if consistency were predictable. I dug up Doug's article which is a few years old, but I don't know that I'd expect it to not still be valid. From the conclusions section:

Update (7/7/2000): I ran some numbers on this, and it turns out that my guess was on the mark. That is, consistency in year N is a very weak indicator of consistency in year N+1. To be specific, I looked at all players from 1995-1998 who had over 50 fantasy points, and I measured their game-by-game standard deviation for that year and the next year. For RBs, I got a correlation coefficient of .09. For WRs, it was .29, and for QBs, it was -.11. Players who were consistent one year showed no strong tendency to be consistent the following year.
 
I've seen a couple of guys advocating the use of PPG now. Just interested on getting some opinions and seeing some some arguments for and against each.

Personally I'm still leaning towards total points. PPG would seem to ignore injury risk completely. It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.
You should rank players by total points, but PPG is a more meaningful #. These two thoughts do not contradict each other.
 
BTW, for those voting for PPG due to "consistancy", that is not why you should be voting for PPG.

 
I've seen a couple of guys advocating the use of PPG now. Just interested on getting some opinions and seeing some some arguments for and against each.

Personally I'm still leaning towards total points. PPG would seem to ignore injury risk completely. It could also over-emphasize the statistics of someone that replaced an injured player if the replacement faced some select easy defenses.
You should rank players by total points, but PPG is a more meaningful #. These two thoughts do not contradict each other.
Let me steal bagger's words and add to them:You should rank players by total points you think they will score next year, but PPG from last year is a more meaningful # than Total Points from last year. These two thoughts do not contradict each other.

Going back to our discussion of including injury... the total points you think they will score next year would incorporate back in your expectations of injuries, risk of losing starting job, etc.

 
Last edited by a moderator:
Man! You guys put way too much analysis in this stuff. When you are breaking out quantum physics math equations you've gone over the edge. Maybe that's the problem. Some people over analyze and they take out the "gut feeling factor"...or also known as "luck". I would never spend the amount of time needed to come up with these equations and such. To me that seems like borderline madness. The extent of my research is basically composed of listening to the news and keeping up to date in injuries and various transactions and such. THen I go back and research say the players last 3 years and look at how he scored each game in that 3 year span. Then you start to find "trends" or consistencies that will help you make safer choices. When you start adding alpha characters to your numbers when trying to find a fantasy player...that's crazy. The only thing pie is good for is eating...not math.

 
It's actually quite simple. Let's say you're doing research on possible players to draft. Let's use a 5 week schedule for length purpose. Here's an example:

Player A

Game 1 Game 2 Game 3 Game 4 Game 5

50 pts 5 pts 100 pts 15 pts 15 pts

total points = 185 points in 5 games

Man, he probably helped you win in week 1 and week 3. But on those down weeks you better hope someone else stepped up and scored more points than they usually do.

Player B

Game 1 Game 2 Game 3 Game 4 Game 5

35 pts 40 pts 30 pts 40 pts 30 pts

total points = 175 points in 5 games

Even though player B scored 10 points less in that span he was a more consistent scorer and you can put him in your lineup and get a good nights sleep. If your team is composed of "consistent scorers" you have a better chance to win. I never look at total points for the year. I always look at how many points a guy averaged per game. :thumbup:
Others have already pointed out that the total points and PPG are equivalent in these scenarios; PPG is a measure of consistency only insofar as it eliminates weeks where the player didn't play.But there's another point to make about consistency; you're probably better off with not all your players being consistent. Once you get to the playoffs, you need to have some big scores; if you're consistently scoring in the middle of the pack, someone will beat you. Probably the best thing is a mix of consistent scorers and "big game" players. If you start three WRs, your third WR consistently scoring 6 points isn't going to win you a lot of games, but if he scores 18 one week and averages 2 for the next three, he probably won you one game and didn't cost you any of the others.

 
i like to look at all factors but I also go back and look at PPG for players that got injured early in the season and would have projected out very well and also do the same looking at players that got jobs at the end of the season and then project that out....

more then likely players that were injured early will be forgotten and can be decent value picks in the following yrs draft

IE:

Eric Johnson TE SF

Jeff Ulbrich LB SF

just two examples of players that missed the end of the yr last yr that will be nice value picks this yr

 
Man! You guys put way too much analysis in this stuff. When you are breaking out quantum physics math equations you've gone over the edge. Maybe that's the problem. Some people over analyze and they take out the "gut feeling factor"...or also known as "luck". I would never spend the amount of time needed to come up with these equations and such. To me that seems like borderline madness. The extent of my research is basically composed of listening to the news and keeping up to date in injuries and various transactions and such. THen I go back and research say the players last 3 years and look at how he scored each game in that 3 year span. Then you start to find "trends" or consistencies that will help you make safer choices. When you start adding alpha characters to your numbers when trying to find a fantasy player...that's crazy. The only thing pie is good for is eating...not math.
Been sitting here deciding how to reply. From the response I feel like what I said was taken personally somehow. It certainly wasn't meant that way.I used to put stock in consistency making that same assumption, that guys were consistent from one year to the next. So I was glad that Doug took the time to see if that assumption was valid or not.

There are lots of assumptions that seem logical but don't hold true. Consistency. That missing time last year means the player is much more likely to miss time next year than a guy who played 16 games, etc.

 
Let me steal bagger's words and add to them:

You should rank players by total points you think they will score next year, but PPG from last year is a more meaningful # than Total Points from last year. These two thoughts do not contradict each other.

Going back to our discussion of including injury... the total points you think they will score next year would incorporate back in your expectations of injuries, risk of losing starting job, etc.
Now how are others going to learn if you do all the work for them? ;)

 
Man!  You guys put way too much analysis in this stuff.  When you are breaking out quantum physics math equations you've gone over the edge.  Maybe that's the problem.  Some people over analyze and they take out the "gut feeling factor"...or also known as "luck".  I would never spend the amount of time needed to come up with these equations and such.  To me that seems like borderline madness.  The extent of my research is basically composed of listening to the news and keeping up to date in injuries and various transactions and such.  THen I go back and research say the players last 3 years and look at how he scored each game in that 3 year span.  Then you start to find "trends" or consistencies that will help you make safer choices.  When you start adding alpha characters to your numbers when trying to find a fantasy player...that's crazy.  The only thing pie is good for is eating...not math.
Been sitting here deciding how to reply. From the response I feel like what I said was taken personally somehow. It certainly wasn't meant that way.I used to put stock in consistency making that same assumption, that guys were consistent from one year to the next. So I was glad that Doug took the time to see if that assumption was valid or not.

There are lots of assumptions that seem logical but don't hold true. Consistency. That missing time last year means the player is much more likely to miss time next year than a guy who played 16 games, etc.
Nah...I didn't take anything personally. I am just truly amazed at the algorithms and mathematical equations people come up with all in the name of fantasy football. I don't know if it's jealousy or just misunderstanding. I never have the time in a day to do that type of research or math logic. Wild stuff.
 
I use a system where I take the player's name, give points per letter (A=1, Z=26, etc.) add them up, and draft accordingly.

Ian Allen is a STUD

 
GroveDiesel,

Yes it is but think of it this way. Let's assume a 10 game season (simply because I'm too lazy this am to do any real math).

Do you want a guy who score 100 points for the over all season or a guy who will get you 10 pts per week for 10 games? I want the guy who will get me the 10 every week, not the player who may get me 15 one week and 5 the next.

I typically do my projections and then do 2 tables. One that sorts by total points and another that sorts by PPG. when you sort by PPG you will see alot of other talent jump up the rankings. I look at it as a way to get some players that may be way down someone elses board because their TP didn't rank them that high.
Ummm... sorting both of those players by points per game would do nothing. Both players scored exactly the same PPG.The first player scored 100 points in 10 games. That's 100/10 ppg. The second player scored 10 points a game for 10 games, which equals 100 total points, which equals 100/10 points per game.

So you're saying you sort by points per game to separate the guys who score 10 points per game from the guys who also score 10 points per game?

If both players play the exact same number of points in the exact same number of games, their PPG will be the same, whether one of them scored all the points in one week and the other was consistant.

Personally, I prefer points per game, but I always have to have a sanity cap. Anyone who started 6 or fewer games gets a gigantic red flag. Anyone who started 8 or fewer games gets a gigantic red flag if they had a couple of high outliers (i.e. 3 TD games) pulling up their average. Anyone who did the majority of their damage over the last half of the season (Kevin Jones and Tatum Bell 2 years ago, Willie Green the year before that) gets another big red flag, because they seem to frequently have trouble duplicating the success (actually, I need to make a note to do a study on that this offseason, whether players who scored 80% of their points over the final 8 weeks tend to see a drop in their PPG the following season...). I don't really know why, perhaps it's because they're playing against tired competition while they're still fresh. Anyway, points per game is useful, but it's no replacement for common sense.

As for inconsistancy... I always try to start inconsistant, boom-or-bust types as my WR3. To borrow from your 10 game example, I figure my choices at WR3 are going to be a guy who scores 6 points a week for the season, and a guy who puts up a pair of 20 point games, and then 2.5 points per game for the rest of the year. I don't think the extra 3.5 points per game from my WR3 slot are going to help me win many games... but those 20 point games sure will.

Basically, if a player is going to be high scoring, I want him to be consistantly high scoring. If a player is going to be low scoring, I want him to be INCONSISTANTLY low scoring.

 
From the conclusions section:

Update (7/7/2000): I ran some numbers on this, and it turns out that my guess was on the mark. That is, consistency in year N is a very weak indicator of consistency in year N+1. To be specific, I looked at all players from 1995-1998 who had over 50 fantasy points, and I measured their game-by-game standard deviation for that year and the next year. For RBs, I got a correlation coefficient of .09. For WRs, it was .29, and for QBs, it was -.11. Players who were consistent one year showed no strong tendency to be consistent the following year.
I just want to get on record as saying that, since writing that particular paragraph, I have decided that I'm not particularly happy with the methodology I used. I don't know if I should be unhappy that I ever wrote it or happy that I'm smarter than I used to be. But either way, I don't like it anymore.But the good news is: since then I've done a few other studies (with methodology I do like) to test the same thing, and they've all come to the same conclusion, which is bolded above.

 
But doesn't your scenario mean that Team B had an extra roster spot or that he was able to draft his #2 or #3 WR in the round that Team A drafted his #2 QB? Having 2 QBs combine to score more points than another team's 1 QB doesn't mean that Team A didn't sacrifice at another position to do so.
No. This looks like a newbie question. Everybody has the same number of roster spots. Nobody knows if their first string QB is going to be injured unless he drafted an already injured player as his QB1. Its like you are assuming that you can predict which players will or wont be injured. You cant. Nobody can. That's not even part of the debate. Everyone drafts, carries, and aquires more backups every season. That applies to both team A and team B in my simple example. The example is only to show that a fantasy scoring isnt based on a player's yearly total. Rather its based on his weekly score which is why ppg is the more informational statistic.
 
One thing about ppg is that I normally dont start really judging a player's stats until he has around 25 starts. Yes, that means I'm not valueing players based on last season... try not to fall into shock! I make exceptions to this especially when looking at RBs (out of neccesity). I think we all agree that you cant judge a player by only a handful of games. I'll also contend that you cant judge a player on only one season. So regardless of whether I was looking at ppg or pps, I'm not judging the player who only has a history of 6 starts by his stats.

 
How about using PPG with a little common sense?

As others have mentioned, PPG and consistency don't belong in the same sentence. PPG is just an average and doesn't differ at all from total points if everyone plays the same number of games.

The NFL is constantly changing. Players get hurt, players switch teams, coaches switch teams, players rise or fall on the depth chart, etc. If I look at last year's stats, I prefer PPG since some players simply didn't play a full season for whatever reason.

Of course, I'll revise the list based on some of those changes, the things I've learned about the player or the player's downside/upside. Its an art, not a science.

 
Last edited by a moderator:
I use PPG to get an idea of expectations and then take a look at the injury a player may have recieved. But you need to take out injuries if you are going to look at how well a player performed, because when they were hurt they didn't perform.

 
But doesn't your scenario mean that Team B had an extra roster spot or that he was able to draft his #2 or #3 WR in the round that Team A drafted his #2 QB? Having 2 QBs combine to score more points than another team's 1 QB doesn't mean that Team A didn't sacrifice at another position to do so.
No. This looks like a newbie question. Everybody has the same number of roster spots. Nobody knows if their first string QB is going to be injured unless he drafted an already injured player as his QB1. Its like you are assuming that you can predict which players will or wont be injured. You cant. Nobody can. That's not even part of the debate. Everyone drafts, carries, and aquires more backups every season. That applies to both team A and team B in my simple example. The example is only to show that a fantasy scoring isnt based on a player's yearly total. Rather its based on his weekly score which is why ppg is the more informational statistic.
I'm aware that everyone has the same number of roster spots. But if you have an injured QB that means you now need 3 QBs on your roster to account for byes. If I don't have an injured QB I don't need 3 QBs on my roster. Again, if you need 2 guys to equal or slightly better the scoring of 1 guy, you are wasting roster spots. To get 2 QBs that can combine to give you numbers better than my 1 QB, that means that when you were picking your second QB, I was picking my #2 or #3 WR that is going to start for my team all year long and beat your #2 or #3 WR by a point or two all season long. So while your QB position may slightly outscore me, I'll have an even larger advantage at WR because you needed 2 QBs to beat my one.

 
I use PPG to get an idea of expectations and then take a look at the injury a player may have recieved. But you need to take out injuries if you are going to look at how well a player performed, because when they were hurt they didn't perform.
I disagreeCertain guys find a way to get hurt, are unlucky, are risk takers...they get hurt "every" year

 
I use PPG to get an idea of expectations and then take a look at the injury a player may have recieved.  But you need to take out injuries if you are going to look at how well a player performed, because when they were hurt they didn't perform.
I disagreeCertain guys find a way to get hurt, are unlucky, are risk takers...they get hurt "every" year
Sorry I didn't make this more clear, but that is why I look at the player and their injury I agree with you just didn't write it well.
 
rank the number of good weeks first. Pick something as your definition of a good week.
Good GamesIf you want to see just # of games where a certain total yards was reached, then set tds = 10, and yds = "your desired number".

If you want to see just # of games where a certain total tds was reached, then set yds = 1000 and tds = "your desired number"

If you want to see # of games where a certain # of tds were scored OR exceeded a certain total yardage total, then just adjust the numbers to the level that you want.

PPG Calculator

change the ppg number at the end of the querystring to whatever level you want to see.

 
Last edited by a moderator:
This poll is useless without being specific about what type of league you are talking about.

In 12 team leagues PPG is a lot more important than in 32 team leagues.

 
But doesn't your scenario mean that Team B had an extra roster spot or that he was able to draft his #2 or #3 WR in the round that Team A drafted his #2 QB? Having 2 QBs combine to score more points than another team's 1 QB doesn't mean that Team A didn't sacrifice at another position to do so.
No. This looks like a newbie question. Everybody has the same number of roster spots. Nobody knows if their first string QB is going to be injured unless he drafted an already injured player as his QB1. Its like you are assuming that you can predict which players will or wont be injured. You cant. Nobody can. That's not even part of the debate. Everyone drafts, carries, and aquires more backups every season. That applies to both team A and team B in my simple example. The example is only to show that a fantasy scoring isnt based on a player's yearly total. Rather its based on his weekly score which is why ppg is the more informational statistic.
I'm aware that everyone has the same number of roster spots. But if you have an injured QB that means you now need 3 QBs on your roster to account for byes. If I don't have an injured QB I don't need 3 QBs on my roster. Again, if you need 2 guys to equal or slightly better the scoring of 1 guy, you are wasting roster spots. To get 2 QBs that can combine to give you numbers better than my 1 QB, that means that when you were picking your second QB, I was picking my #2 or #3 WR that is going to start for my team all year long and beat your #2 or #3 WR by a point or two all season long. So while your QB position may slightly outscore me, I'll have an even larger advantage at WR because you needed 2 QBs to beat my one.
Not necessarily. QBs in particular tend to have a very tight scoring range, meaning the difference between the #5 QB and the #15 QB tends to be pretty small. You can get quality starter QBs in the extremely late rounds, or even off of the street, every single season. Recent examples include Kitna, Brees, Brunell, Bledsoe, Griese, etc. Same deal with kickers and defenses.Now, if you want to make the arguement that total points is better because there's such a marked point dropoff between starting-caliber players and backup-caliber players (meaning one would have to invest high draft picks into backups in order to still score once someone leaves due to injury), you're better off using RBs as an example. I believe a case could be made that total points are to some extent more valuable than points per game with RBs (unless the injured player's total points were so close to the uninjured player's total points that a scrub off of the street could still make up the difference).

Regardless, you're talking as if investing a high pick in a backup is the end of the world. Worst case scenario, someone goes out due to injury and you have a quality backup (and, meanwhile, are beating up on all the injured teams without quality subs). Best case scenario, no one gets injured, and you have some veeeeery valuable trade bait going into the playoffs.

 
I'm aware that everyone has the same number of roster spots. But if you have an injured QB that means you now need 3 QBs on your roster to account for byes. If I don't have an injured QB I don't need 3 QBs on my roster. Again, if you need 2 guys to equal or slightly better the scoring of 1 guy, you are wasting roster spots.

To get 2 QBs that can combine to give you numbers better than my 1 QB, that means that when you were picking your second QB, I was picking my #2 or #3 WR that is going to start for my team all year long and beat your #2 or #3 WR by a point or two all season long. So while your QB position may slightly outscore me, I'll have an even larger advantage at WR because you needed 2 QBs to beat my one.
Sorry but there just isnt anything true about what you just said. Draft position does not determine ppg. Never has, never will. My example isnt saying Team A will always outscore Team B. The total points for each team in my example was nearly identicle. What was a huge difference was the total points for each player (more than 100 for the season). Anyone who is using the total points scored to determine those players' values is comeing up with a very wrong value. Despite over 100 more fantasy points, the QB1 on team B didnt help team B score any more points than team A. In my example Team A scored more points from the QB position. I could have just as easily made the total points even or in favor of team B. That doesnt matter. What matters is that QB1 B didnt provide any more fantasy value than QB1 A in my example despite the lopsided difference in total points.Nobody is required to carry 3 QBs unless some rare league rule requires it.

Scoring is done weekly, by position. Team A and B are both likely carrying two QBs on their roster. If team B isnt, its a gamble that has an equal chance of backfiring as providing any benefit. Team A may or may not (usually not) need to use up one roster spot for one or more weeks to pick up a bye week filler if one QB is injured and the other is on bye. He/she isnt under any obligation to keep QB1 and QB2 at that time. He/she isnt obligated to come up with a bye week replacement for more than one week. The likelyhood of team A losing any points at all by having to cut one player for a week or even a few is very low. In fact, having to cut one's worst player in order to come up with a bye filler is nearly just as likely to have a positive reaction by comeing up with an even better player on the next free agent grab. Team B simply doesnt gain any advantage from team A's QB injury except when they play head to head during the injury period. Which has nothing to do with the overall value of each team's QB1.

You have the arguement of one who thinks fantasy football is played with past statistics. Its not. Its played with future stats that arent predictable. The 2006 points for all QBs is zero. No QB has missed one game in 2006. We are talking about a player's ppg and pps which are only good through 2005.

 
I'm aware that everyone has the same number of roster spots. But if you have an injured QB that means you now need 3 QBs on your roster to account for byes. If I don't have an injured QB I don't need 3 QBs on my roster. Again, if you need 2 guys to equal or slightly better the scoring of 1 guy, you are wasting roster spots.

To get 2 QBs that can combine to give you numbers better than my 1 QB, that means that when you were picking your second QB, I was picking my #2 or #3 WR that is going to start for my team all year long and beat your #2 or #3 WR by a point or two all season long. So while your QB position may slightly outscore me, I'll have an even larger advantage at WR because you needed 2 QBs to beat my one.
Sorry but there just isnt anything true about what you just said. Draft position does not determine ppg. Never has, never will. My example isnt saying Team A will always outscore Team B. The total points for each team in my example was nearly identicle. What was a huge difference was the total points for each player (more than 100 for the season). Anyone who is using the total points scored to determine those players' values is comeing up with a very wrong value. Despite over 100 more fantasy points, the QB1 on team B didnt help team B score any more points than team A. In my example Team A scored more points from the QB position. I could have just as easily made the total points even or in favor of team B. That doesnt matter. What matters is that QB1 B didnt provide any more fantasy value than QB1 A in my example despite the lopsided difference in total points.Nobody is required to carry 3 QBs unless some rare league rule requires it.

Scoring is done weekly, by position. Team A and B are both likely carrying two QBs on their roster. If team B isnt, its a gamble that has an equal chance of backfiring as providing any benefit. Team A may or may not (usually not) need to use up one roster spot for one or more weeks to pick up a bye week filler if one QB is injured and the other is on bye. He/she isnt under any obligation to keep QB1 and QB2 at that time. He/she isnt obligated to come up with a bye week replacement for more than one week. The likelyhood of team A losing any points at all by having to cut one player for a week or even a few is very low. In fact, having to cut one's worst player in order to come up with a bye filler is nearly just as likely to have a positive reaction by comeing up with an even better player on the next free agent grab. Team B simply doesnt gain any advantage from team A's QB injury except when they play head to head during the injury period. Which has nothing to do with the overall value of each team's QB1.

You have the arguement of one who thinks fantasy football is played with past statistics. Its not. Its played with future stats that arent predictable. The 2006 points for all QBs is zero. No QB has missed one game in 2006. We are talking about a player's ppg and pps which are only good through 2005.
Of course past performance doesn't dictate future performance. So why even use any numbers from the past then? You seem to simultaneously be saying that you should use PPG to help you make your projections, but you shouldn't use anything at all. I mean heck, it would be fantastic to draft both of the QBs for Team A in the 7th and 13th rounds. But since we can't know that ahead of time we must use some statistical method to project those numbers onto realistic draft positions. We have no choice but to assume that a QB that scores 3 points more each game than another QB will get drafted at a higher draft position.If we both draft 2 QBs to start the season, and your starting QB gets hurt during Week 8, you now only have 1 healthy QB on your roster. If your bye is after that week then you have no choice but to add another QB. So now you have to cut a player at another position and you now have 3 QBs on your roster. Of course, if your QB is on IR you simply cut him. But again, while you took that 21 PPG QB and got 8 games out of him, perhaps I took a RB or WR in that round that outscored your player at that position by 2 points all season long.

I just don't see why you're trying to pretend that your scenario doesn't have major problems. Your entire argument boils down to "injuries don't matter." If your first round pick gets hurt, it WILL matter. If someone else's first round pick is not a bust, they WILL have a significant advantage over you for the rest of the season, even if your total season points at the particular position total more than mine.

If you don't want to admit that there's a serious problem with your scenario that's fine. I understand the benefits of PPG and probably will be using it myself going forward. But you clearly do not understand the limitations of PPG and that's your choice. If you'd like to continue to hide your head in the sand and pretend that I'm the idiot feel free to do so.

And what's the point of using any statistical method if you're just going to throw your hands up at the end and say that you can't predict the future? Why not just throw darts at a dartboard to make your draft picks then? Don't try to argue that PPG means something and then tell me that it doesn't matter because you can't use it for anything useful.

Of course scoring more points at the QB position than another team is a good thing. That seems to be your only point left if I can't use the numbers to make projections. Sure I'd like to have 2 QBs outscore 1 QB if I could do that. But what does that mean when I draft? How should I incorporate that into choosing my players? According to you, it means nothing.

 
Last edited by a moderator:
I just don't see why you're trying to pretend that your scenario doesn't have major problems. Your entire argument boils down to "injuries don't matter." If your first round pick gets hurt, it WILL matter. If someone else's first round pick is not a bust, they WILL have a significant advantage over you for the rest of the season, even if your total season points at the particular position total more than mine.
And so what? This has nothing to do with PPG. This just says that you should avoid drafting players who are going to get injured. And if you know in advance who is going to get injured next season, then for heaven's sake go to Vegas and win some big $$$ instead of wasting your time with ff.
 
Maybe I'm just confused at how you're calculating PPG.

Isn't PPG just total points divided by number of games played? Or are you using a different way to average weekly scores or something?

Because if it's just total points divided by number of games played, isn't that the same as total points only factoring out injuries?
I use total points divided by number of games the player played in. If the season is 16 games long but he only played in 10, I divide by 10 to see how many points per game he scored. Then you consider whether that player is injury prone and how many games he will be injured. Then decide whether or not you want to draft him. If you have a good backup to fill in his place, you should be fine.
 

Users who are viewing this thread

Top