Figuring out Hurt/Suspended Player Value (1 Viewer)

[Yamato]

This year I'm adding games missed to my projections (projected injuries, suspensions).  I'm figuring out per game averages and using average replacement (#18 QB, #30 RB, and so on) for games missed to come up with a yearly average for that starting roster position.  So say a player averages 15 points/week but is suspended for 4 games.  Average replacement scores 5.5 PPG. His average points would then be 12.6 PPG for that postion

Does anyone see this as being inaccurate?  Does anyone else have a different way?  I was thinking of adding something like a dead spot penalty but would have to really sit down and figure out how that would work.     

 

[Snorkelson]

This year I'm adding games missed to my projections (projected injuries, suspensions).  I'm figuring out per game averages and using average replacement (#18 QB, #30 RB, and so on) for games missed to come up with a yearly average for that starting roster position.  So say a player averages 15 points/week but is suspended for 4 games.  Average replacement scores 5.5 PPG. His average points would then be 12.6 PPG for that postion

Does anyone see this as being inaccurate?  Does anyone else have a different way?  I was thinking of adding something like a dead spot penalty but would have to really sit down and figure out how that would work.     

I’m no stats guy but it seems sound in theory mathematically. I try not to base too much off the unpredictability of injuries as a lot of it is chance. Foot injuries worry me but I don’t try to predict whether a guy might miss 2 games this year or 4 games this year. It might make me skip fournette for someone else, but I simply look at the ppg as ppg when they’re playing and hope they stay healthy. Guys who have never been injured have just as much chance at getting injured as a guy who tore his acl last year. 

Do you take the backups and add points because they would benefit from the extra 2-4 games they would get? No, probably not. 

Suspensions are another story but again, if Edelman scores let’s say 12 ppg, I’ll look at him as a 12 ppg guy I can get later than I should because he’s missing 4 games. I don’t downgrade his ppg, he’s still going to get that. 

I guess if you want to extrapolate a number and simply go by that then by all means, but for me this is relying on numbers a bit too much to quantify the inpredictable. 

 

[Ignoratio Elenchi]

This year I'm adding games missed to my projections (projected injuries, suspensions).  I'm figuring out per game averages and using average replacement (#18 QB, #30 RB, and so on) for games missed to come up with a yearly average for that starting roster position.  So say a player averages 15 points/week but is suspended for 4 games.  Average replacement scores 5.5 PPG. His average points would then be 12.6 PPG for that postion

Does anyone see this as being inaccurate?  Does anyone else have a different way?  I was thinking of adding something like a dead spot penalty but would have to really sit down and figure out how that would work.     

It's inaccurate in the sense that it's potentially distributing his points in a misleading way.  Consider an oversimplified example, where you're comparing the player you described to one who's projected for 14 PPG.  14 > 12.6, so you'd draft the 14 PPG guy.  But when you match them H2H for the whole season, the 12.6 PPG guy wins most weeks.  

If you absolutely need to come up with a single number to represent his season-total value, then I think it's right to take his projected points and add in replacement value for the weeks he's expected to miss.  Just don't forget that it's a weekly game, there's roster/positional scarcity, etc.  

 

[Interseptopus]

Seems like you just grabbed a random replacement number. in that case, no, that's not accurate at all. 

 

[Dismattle]

Ready to test your threshold for pain tolerance eh  Carrying the walking wounded and then hoping for the best during the byes  Hope we see some responses that appear rationale because it does seem like a good question  Personally, I would just suggest marking the player as like base and/or high end in which to differentiate as to whether he provides (when playing) a solid average or is more boom/bust  Using such logic the worst case scenario may involve a high draft pick because they would dictate your draft  Now I can tell you that I've read that no adjustment is needed for only two weeks

Here is an older article which may help w/ numbers  https://sports.yahoo.com/news/suspended-injured-players-070100123--nfl.html

(Projected PPG * Expected Games) + (Replacement Player Projected PPG * Expected Games Missed)  That means that the calculation for Gordon in 2013 was (7.4*14) + (6.0*2), or 115.6 points. That’s barely less than the 118.4 points we should have expected from Gordon (based on his ADP) if he played all 16 games.

Is it possible that you can specify the player(s) in question?  For instance, I'd dare say nearly no adjustment for Ingram based on previous performance and special notes like the teams' early bye week  Personally, I would not much care to go, Edelman, unless it's at a good cost to savings because of mixed results and somewhat obviously be starting in Belichick's doghouse  (It took Edelman five years to clear over 750 yds in a Season  I like that number to be within three if not two yrs)

 

[Yamato]

The replacement number is bases on AVT from my league standings.  Or for PPR I use this.

Seems like you just grabbed a random replacement number. in that case, no, that's not accurate at all. 

The replacement number is bases on AVT from my league scoring.  Or for PPR I use this.

https://www.footballdiehards.com/ledger/AVT_ledger.cfm

The actual number for my league is 5.44 for RB's and WR's.  Also why the hell would I go though all this to just pull a number out of my ###?

 

[Interseptopus]

The replacement number is bases on AVT from my league standings.  Or for PPR I use this.

The replacement number is bases on AVT from my league scoring.  Or for PPR I use this.

https://www.footballdiehards.com/ledger/AVT_ledger.cfm

The actual number for my league is 5.44 for RB's and WR's.  Also why the hell would I go though all this to just pull a number out of my ###?

I dont think it accurately represents what that player is capable of. if you have a guy who gets 15 points every single week, but misses 2 weeks, they should not be valued at 13.8

 

[Yamato]

It's inaccurate in the sense that it's potentially distributing his points in a misleading way.  Consider an oversimplified example, where you're comparing the player you described to one who's projected for 14 PPG.  14 > 12.6, so you'd draft the 14 PPG guy.  But when you match them H2H for the whole season, the 12.6 PPG guy wins most weeks.  

If you absolutely need to come up with a single number to represent his season-total value, then I think it's right to take his projected points and add in replacement value for the weeks he's expected to miss.  Just don't forget that it's a weekly game, there's roster/positional scarcity, etc.  

Regarding roster/positional scarcity I did say I was considering a dead roster spot penalty.

My example was based on a SUSPENDED player who will be out a specified number of games.  People keep saying you can't predict injuries? I'm not.  I'm projecting risk.  If a player for the past 3 years has never been below 1000 yards and 6 TDs I would think he has a better than average change to that that the next year.  If a player has missed 3 games a year for the past 3 years...would projecting he'll miss 2 games this year be wrong?

 

[Interseptopus]

Regarding roster/positional scarcity I did say I was considering a dead roster spot penalty.

My example was based on a SUSPENDED player who will be out a specified number of games.  People keep saying you can't predict injuries? I'm not.  I'm projecting risk.  If a player for the past 3 years has never been below 1000 yards and 6 TDs I would think he has a better than average change to that that the next year.  If a player has missed 3 games a year for the past 3 years...would projecting he'll miss 2 games this year be wrong?

if you did this for Keenan Allen in 2017, yes, you would have been wrong. there are a lot of examples where this is flawed logic. you are lumping all injuries into one category, when they are all different. 

this is a perfect example of how people have mislabeled Dion Lewis as injury prone. when his injuries were just as much of a fluke as Aaron Rodgers' injury last year. 

is it fair to say Jordan Reed and Dion Lewis are the same injury risk? I say no. 

 

[Ignoratio Elenchi]

Regarding roster/positional scarcity I did say I was considering a dead roster spot penalty.

My example was based on a SUSPENDED player who will be out a specified number of games.  People keep saying you can't predict injuries? I'm not.  I'm projecting risk.  If a player for the past 3 years has never been below 1000 yards and 6 TDs I would think he has a better than average change to that that the next year.  If a player has missed 3 games a year for the past 3 years...would projecting he'll miss 2 games this year be wrong?

I didn’t say anything about injuries. But if you’re asking me, then yes, as a general rule I think it would be wrong to project injuries like that. 

 

[Yamato]

I dont think it accurately represents what that player is capable of. if you have a guy who gets 15 points every single week, but misses 2 weeks, they should not be valued at 13.8

How do you value Jordan Reed?  Last 3 years he hasn't played more than 14 games.  Gronk hasn't played more than 13 games the past 2 seasons.  Do you project them for 16 games? How do you rank Eifert?

Edelman and Engram will miss 4 games each.  So unless they get hurt they'll get you 12 games.  Should they lose 25% of their value? You aren't getting zero for that position.  You're using a replacement.  Edelman averaged 14.5/PPR in 2016, Engram 17.75. The average PPR replacement is about 6.3 PPG.  Using just total points their value is 10.88 and 12.53. Using a replacement player their average is 12.53 and 14.97.  The question is what is the best value.    

 

[Interseptopus]

How do you value Jordan Reed?  Last 3 years he hasn't played more than 14 games.  Gronk hasn't played more than 13 games the past 2 seasons.  Do you project them for 16 games? How do you rank Eifert?

Edelman and Engram will miss 4 games each.  So unless they get hurt they'll get you 12 games.  Should they lose 25% of their value? You aren't getting zero for that position.  You're using a replacement.  Edelman averaged 14.5/PPR in 2016, Engram 17.75. The average PPR replacement is about 6.3 PPG.  Using just total points their value is 10.88 and 12.53. Using a replacement player their average is 12.53 and 14.97.  The question is what is the best value.    

Ingram not Engram

Sure, this is fine for suspended players, but when you try to project injuries it loses its validity. 

Prior to last year you could have projected Dion Lewis to miss 8 games. but he didnt. thus you missed out on some great value. and hes completely mislabeled as injury prone based on a history of fluke injuries.

what about Cook? he missed almost the entire season. surely you arent projecting 12 missed games? any? if so, why? and if not, what's your justification? 

I see what you're trying to do, and honestly I think you're doing too much when you try to determine how many games a player might miss. you're trying to make your rankings as objective as you can, but the basis for your final PPG is based on a subjective assessment of expected games played/missed. that's flawed from the get go and is no longer objective

best of luck to your strategy. hope it works out for you. 

 

[Yamato]

Ingram not Engram

Sure, this is fine for suspended players, but when you try to project injuries it loses its validity. 

Prior to last year you could have projected Dion Lewis to miss 8 games. but he didnt. thus you missed out on some great value. and hes conpletely mislabeled as injury prone. 

what about Cook? he missed almost the entire season. surely you arent projecting 12 missed games? any? if so, why? and of not, what's your justification? 

I see what you're trying to do, and honestly I think you're doing too much when you try to determine how many games a player might miss. you're trying to make your rankings as objective as you can, but the basis for your final PPG is based on a subjective assessment of expected fames played. that's flawed from the get go. 

best of luck to your strategy. hope it works out for you. 

The purpose for this post is to figure out how to rank players who are consistently injured.  II'm looking for ideas not trying to press my methods. The questions I asked I'd like to talk about.

How do you value Jordan Reed?  Last 3 years he hasn't played more than 14 games.  Gronk hasn't played more than 13 games the past 2 seasons.  Do you project them for 16 games? How do you rank Eifert?

What I'm hearing is players who are consistently injured every year should not be penalized and should be given value as if they will play all 16 games.  I'm starting to think projecting them to play 16 games as their ceiling and their projected games played value as their floor.

 

[Ignoratio Elenchi]

To a certain extent, whether or not you can predict injuries is a math question.  Injury history either is or isn't a reliable predictor of future time missed due to injuries, so rather than guess at it you can just pull the data and see how much weight you should put on it.  I've never done that so I don't know what the answer is, but if you're trying to bake it into your projections that's probably the first step.  

 

[Interseptopus]

The purpose for this post is to figure out how to rank players who are consistently injured.  II'm looking for ideas not trying to press my methods. The questions I asked I'd like to talk about.

How do you value Jordan Reed?  Last 3 years he hasn't played more than 14 games.  Gronk hasn't played more than 13 games the past 2 seasons.  Do you project them for 16 games? How do you rank Eifert?

What I'm hearing is players who are consistently injured every year should not be penalized and should be given value as if they will play all 16 games.  I'm starting to think projecting them to play 16 games as their ceiling and their projected games played value as their floor.

I put guys like Gronk and Reed on my do not draft list to avoid this headache. 

No, someone who is injured every year for little things like a strained groin are not to be expected to play 16 games, but I wont even try to determine how many games those guys could miss. 

I dont have a right answer on this. I have no idea how to project a player who is frequently injured. 

sportsinjurypredictor tries to do this, but I have no idea what their algorithm is and how they can predict missed time. I think there is a lot of room for error when you start making objective claims off subjective information. if you're okay with that, then go nuts. 

I think your method of taking an average replacement to average in for a player who is suspended 4 games is a great way to determine value. where it goes sideways is arbitrarily determining how many games a particular player is likely to miss this year. but at the end of the day these are your rankings for you. 

 

[Long Ball Larry]

To a certain extent, whether or not you can predict injuries is a math question.  Injury history either is or isn't a reliable predictor of future time missed due to injuries, so rather than guess at it you can just pull the data and see how much weight you should put on it.  I've never done that so I don't know what the answer is, but if you're trying to bake it into your projections that's probably the first step.  

I've never subscribed, so I don't know exactly how accurate it is, but this guy claims that he can at least give insight into injury likelihood: https://sportsinjurypredictor.com/

 

[Yamato]

I've never subscribed, so I don't know exactly how accurate it is, but this guy claims that he can at least give insight into injury likelihood: https://sportsinjurypredictor.com/

I have those numbers.  I'm much more conservative.  Here are some of the players I have given games missed from Injury:

Tevin Coleman 1 game

John Brown 2 games

LeSean McCoy 1 game

Greg Olsen 1 game

John Ross 1 game

Allen Hurns 2 games

Emmanuel Sanders 1 game

Ty Montgomery 2 games

Deshaun Watson 1 game

Will Fuller 1 game

Andrew Luck 1 game

Marlon Mack 2 games

Leonard Fournette 1 game

Donte Moncrief 2 games

Spencer Ware 2 games

Sammy Watkins 1 game+ (if he stays healthy in preseason)

Mike Williams 1 game+ (if he stays healthy in preseason)

Robert Woods 2 games

Ryan Tannehill 2 games

Frank Gore 1 game

DeVante Parker 1 game

Danny Amendola 2 games

Dalvin Cook 1 game

Stefon Diggs 1 game

Tom Brady 1 game

Rex Burkhead 2 games

Sony Michel 2 games

Chris Hogan 1 game+ (if he stays healthy in preseason)

Julian Edelman 1 game+ (if he stays healthy in preseason) 5 total with suspension.

Eric Decker 1 game

Rob Gronkowski 2 games

Drew Brees 1 game

Cameron Meredith 1 game+ (if he stays healthy in preseason)

Ben Watson 1 game

Quincy Enunwa 1 game+ (if he stays healthy in preseason)

Marshawn Lynch 1 game

Carson Wentz 1 game

Jay Ajayi 1 game

Zach Ertz 1 game

Ben Roethlisberger 1 game

LeVeon Bell 1 game

Antonio Brown 1 game

Vance McDonald 2 games

C.J. Prosise 4 games

Doug Baldwin 1 game

Matt Breida 1 game

George Kittle 1 game

DeSean Jackson 2 games

Marcus Mariota 1 game

Dion Lewis 2 games

Corey Davis 1 game

Delanie Walker 1 game

Samaje Perine 1 game

Chris Thompson 2 games

Josh Doctson 1 game

Jordan Reed 3 games

There are others but these are the main ones.

 

[ZWK]

Fantasy playoff games matter a lot more than fantasy regular season games (roughly 5x), so you know that a player is going to miss games at the start of the season that is not as bad as the raw numbers suggest. Whereas injury-prone players who start the season healthy could miss time at any point in the season, and the late-season fantasy playoff games are actually the games that they're most likely to miss.

Roster space matters, but it's tricky to take that into account in a systematic way. The costs increase faster as the number of dead spots goes up, because managing your roster becomes increasingly tight.

5.5 ppg seems too low for the replacement player in most leagues. e.g., If you know that one of your starting WRs will miss week 1, that just means that you have to start your WR4 in your first game and he is probably projected for more than that.

 

[Yamato]

Fantasy playoff games matter a lot more than fantasy regular season games (roughly 5x), so you know that a player is going to miss games at the start of the season that is not as bad as the raw numbers suggest. Whereas injury-prone players who start the season healthy could miss time at any point in the season, and the late-season fantasy playoff games are actually the games that they're most likely to miss.

Roster space matters, but it's tricky to take that into account in a systematic way. The costs increase faster as the number of dead spots goes up, because managing your roster becomes increasingly tight.

5.5 ppg seems too low for the replacement player in most leagues. e.g., If you know that one of your starting WRs will miss week 1, that just means that you have to start your WR4 in your first game and he is probably projected for more than that.

I play in a non PPR league that gives 4 points for all TD's.  We start 2 RB, 2 WR, and 2 flex.  The 5.4 PPG is for the 84th best player but should be 5.7 for 78th best player.