2009 projections (1 Viewer)

[moleculo]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

 

[mlball77]

I'm also sort of a math/stats guy… and I've only done player projects for a season one time. It was very time consuming and took a lot of work and thought (encouraging thought was a good thing though).

I used FBG's already complete projections as my base and reworked them entirely… I copied them into Excel, team by team.

It really was a helpful process because there were some players that stood out to me that I was too high on and that probably wouldn't see the touches I had originally anticiapted, and of course the inverse was the case in several instances as well.

Basically, I started by looking at a team's passing/rushing/receiving stats over the last couple of years and made adjustments to their total passing or rushing or receiving projections based on any changes to personnel, coaching, schemes/philosophies etc. A new pass happy coach might cause a team's passing attempts and yardage to go up while rushing attempts and yds go down and perhaps yards per carry would go up oh so slightly. Or maybe the addition of a vet OG and a good rookie OT would bolster the ground game. These were the types of things I would look at and use to adjust at team's total numbers from the previous couple of years. - Once I came to a projection for what a given team would likely produce, I then started divvying up things to the individual players.

It was a pretty fun process but just too much of a time demand, unfortnately.

 

[evilempire]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

Haven't gotten into the numbers too deeply as they aren't my strong suit, but I have done my own rankings based on three seasons worth of stats. I tailor things to my league's scoring system, and the first thing I do is rank the players. I average three years worth of stats and come up with an average FP depending on where a guy is ranked ... the QB1 in my league has scored 383/454/351, so I would average those totals and get 396 ... which isn't out of line for QBs in our league. Whomever I felt was the top QB would get 396 points, and so on and so on. The numbers are less of an exact science than the rankings, really ... I go with personnel changes, coaching changes, situation better or worse and just a gut feel. I've certainly been wrong about the players, but the numbers have been fairly close for the player and his corresponding ranking within the position. So you aren't guessing as much about numbers as you are about your rankings ... but that's why they're yours, right?

 

[moleculo]

I'm also sort of a math/stats guy… and I've only done player projects for a season one time. It was very time consuming and took a lot of work and thought (encouraging thought was a good thing though).I used FBG's already complete projections as my base and reworked them entirely… I copied them into Excel, team by team.It really was a helpful process because there were some players that stood out to me that I was too high on and that probably wouldn't see the touches I had originally anticiapted, and of course the inverse was the case in several instances as well.Basically, I started by looking at a team's passing/rushing/receiving stats over the last couple of years and made adjustments to their total passing or rushing or receiving projections based on any changes to personnel, coaching, schemes/philosophies etc. A new pass happy coach might cause a team's passing attempts and yardage to go up while rushing attempts and yds go down and perhaps yards per carry would go up oh so slightly. Or maybe the addition of a vet OG and a good rookie OT would bolster the ground game. These were the types of things I would look at and use to adjust at team's total numbers from the previous couple of years. - Once I came to a projection for what a given team would likely produce, I then started divvying up things to the individual players.It was a pretty fun process but just too much of a time demand, unfortnately.

That's what I do as well. Some other things include projecting a team YPC, and then look at individual players historical deviation from the team YPC - i.e. as a team SD averages say 4.6 YPC, and LT typically averages -0.3 YPC less than the team average (made up numbers). So, if I anticipate a better running game for SD, I'll bump up SD to 5.0 YPC, and LT then gets a 4.4 YPC...a rising tide lifts all boats type of approach. Some things I have trouble with include injuries - I don't think it's right to project everyone out 16 games, but it's not right to project injuries either. Even if you do assume a player plays all 16 games when he did not last year, it's tough to figure out the right number of carries everyone on the team should get.I also struggle with YPcatch, making that jive with QB passing yards, and making this all balance.

 

[compone]

I do my own projections. Check out www.ddhg.org

The program is being updated, but that's a general idea...

Here are the stats I use:

Team offensive stats; for the past 3 years, projections and tweaks...

Off those projections I give a % to each player for each stat...

Player tweaks for; Age, BMI, average games played, SOS, Team stats...

Good stuff!

 

[Inspiration]

Some things I have trouble with include injuries - I don't think it's right to project everyone out 16 games, but it's not right to project injuries either. Even if you do assume a player plays all 16 games when he did not last year, it's tough to figure out the right number of carries everyone on the team should get.

I tend to project numbers on a per-game basis, rather than for an entire season. It doesn't exactly equate for injuries, but it does display projected production while leaving room to recognize potential injuries.

 

[MAC_32]

I'm a math/stats guy and probably spend 100's of hours doing my projections for baseball.

That said, I do not create projections for football. To this day I still have not seen a reliable manner at which to create a good data set.

 

[moleculo]

I'm a math/stats guy and probably spend 100's of hours doing my projections for baseball.That said, I do not create projections for football. To this day I still have not seen a reliable manner at which to create a good data set.

how do you define a good data set?

 

[moleculo]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

Haven't gotten into the numbers too deeply as they aren't my strong suit, but I have done my own rankings based on three seasons worth of stats. I tailor things to my league's scoring system, and the first thing I do is rank the players. I average three years worth of stats and come up with an average FP depending on where a guy is ranked ... the QB1 in my league has scored 383/454/351, so I would average those totals and get 396 ... which isn't out of line for QBs in our league. Whomever I felt was the top QB would get 396 points, and so on and so on. The numbers are less of an exact science than the rankings, really ... I go with personnel changes, coaching changes, situation better or worse and just a gut feel. I've certainly been wrong about the players, but the numbers have been fairly close for the player and his corresponding ranking within the position. So you aren't guessing as much about numbers as you are about your rankings ... but that's why they're yours, right?

So you basically rank directly. I think that has some merit, and can certainly be simpler than full fledged projections. After all, the end result of projections is a set of rankings.However, I believe that when one projects and then derives rankings from those projections, you have the following advantages:1. your rankings will be more objective than subjective. You will know why you think Steve Smith will catch 80 balls this year - less emphasis on passing game, for example.2. one set of projections can be used across multiple league formats: PPR , TD=4 points, TE catches = 2 points, etc.3. when you get past the first 15 or so players, differences between guys gets a little fuzzy and it's tough to discern diamonds from rough. Doing straight rankings leads one to focus more on the FBG player du jour, instead of formulating your own opinions.4. The exercise of doing full projections (hopefully) makes your rankings consistent - if a QB is ranked highly, chances are pretty good that he's throwing to someone, and some WR should be ranked highly accordingly. 5. the exercise of projecting all teams makes you better understand all teams, so you will be better prepared come free-agency time - of course, this is a disadvantage too...projecting 32 teams takes too much time.6. you will have a reasonable argument for player projections when the player spotlight threads start up.

 

[mlball77]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

Haven't gotten into the numbers too deeply as they aren't my strong suit, but I have done my own rankings based on three seasons worth of stats. I tailor things to my league's scoring system, and the first thing I do is rank the players. I average three years worth of stats and come up with an average FP depending on where a guy is ranked ... the QB1 in my league has scored 383/454/351, so I would average those totals and get 396 ... which isn't out of line for QBs in our league. Whomever I felt was the top QB would get 396 points, and so on and so on. The numbers are less of an exact science than the rankings, really ... I go with personnel changes, coaching changes, situation better or worse and just a gut feel. I've certainly been wrong about the players, but the numbers have been fairly close for the player and his corresponding ranking within the position. So you aren't guessing as much about numbers as you are about your rankings ... but that's why they're yours, right?

So you basically rank directly. I think that has some merit, and can certainly be simpler than full fledged projections. After all, the end result of projections is a set of rankings.However, I believe that when one projects and then derives rankings from those projections, you have the following advantages:1. your rankings will be more objective than subjective. You will know why you think Steve Smith will catch 80 balls this year - less emphasis on passing game, for example.2. one set of projections can be used across multiple league formats: PPR , TD=4 points, TE catches = 2 points, etc.3. when you get past the first 15 or so players, differences between guys gets a little fuzzy and it's tough to discern diamonds from rough. Doing straight rankings leads one to focus more on the FBG player du jour, instead of formulating your own opinions.4. The exercise of doing full projections (hopefully) makes your rankings consistent - if a QB is ranked highly, chances are pretty good that he's throwing to someone, and some WR should be ranked highly accordingly. 5. the exercise of projecting all teams makes you better understand all teams, so you will be better prepared come free-agency time - of course, this is a disadvantage too...projecting 32 teams takes too much time.6. you will have a reasonable argument for player projections when the player spotlight threads start up.

 

[MAC_32]

I'm a math/stats guy and probably spend 100's of hours doing my projections for baseball.That said, I do not create projections for football. To this day I still have not seen a reliable manner at which to create a good data set.

how do you define a good data set?

In the case of football, I have no idea. I mean, I can tell you what I want - a data set that best represents potential reality - but I'd be shocked if you could convince me your method to create it is an effective one. With the exception of maybe the QB position the sample size is simply too small to ever make an accurate projection for most players at most positions. Baseball's more of an individual sport, each player has a lot of control on their own production, and the sample sizes are quite large which makes projection creation relatively reliable. Football production is so dependent on surrounding factors, many of which cannot be quantified. Anyone can spit out a projection set, but I would not rely on it one bit.

 

[Ministry of Pain]

Projections are a complete wate of time

 

[Biabreakable]

I'm a math/stats guy and probably spend 100's of hours doing my projections for baseball.That said, I do not create projections for football. To this day I still have not seen a reliable manner at which to create a good data set.

how do you define a good data set?

In the case of football, I have no idea. I mean, I can tell you what I want - a data set that best represents potential reality - but I'd be shocked if you could convince me your method to create it is an effective one. With the exception of maybe the QB position the sample size is simply too small to ever make an accurate projection for most players at most positions. Baseball's more of an individual sport, each player has a lot of control on their own production, and the sample sizes are quite large which makes projection creation relatively reliable. Football production is so dependent on surrounding factors, many of which cannot be quantified. Anyone can spit out a projection set, but I would not rely on it one bit.

I agree that projections can be more art than science sometimes. I still think they are useful though.This is a great idea as I definitly trust other peoples projections for teams they are familair with as well as using deductive reasoning and sound numbers. There are always little things that a homer or expert of a team might know about the situation or players that someone with less interest in the team (and thus the division) might miss. And then sharing those projections and having another set of eyes go over and question it ushualy can give a decent projection.There are always outliers and things that happen no one could see coming. Like Brady getting injured last year for example.I don't project static numbers but ranges. This makes it hard to rank them but it shows me tiers. Then there are the 1000 tiebreakers lol.3 year method works pretty well for team projections but you need to really look for outliers when your going over this. Teams that are making substaintal changes make the past less useful. There is a lot of change around the league right now.After the team projections are done then I go to the individual players and I ushualy start with the running game how that will be distributed balancing that with past tendencies of pass/run but tailored to the personel on the team. Once I am done with the running game I have an idea of how many passing attempts there will be and correlate that to the QB completion %. If there have been major changes in personel or young players I might look at things like catch % for WR and I might factor that in to the distribution and also the completions. But that isn't something I give much weight to. I tend to be too conservative in my projections and that is something I would like to work on. I ushualy gravitate to situations where there are more questions and uncertainty. My projections for Washington will likely look pretty similar to what they have been doing for awhile now for example. I don't see huge changes there. But there are guys here who would look a lot deeper into it than me.

 

[moleculo]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

 

[Chase Stuart]

Do a search for Maurile. He's gone into his explanation several times on this board, and I think he and David have the best theory behind their rankings. I'm sure all the other staffers have sound theory, too, but I've at least read and see MT in action. MT's process is controversial in some ways, but I believe it to be correct.

 

[BuckeyeArt]

Do a search for Maurile. He's gone into his explanation several times on this board, and I think he and David have the best theory behind their rankings. I'm sure all the other staffers have sound theory, too, but I've at least read and see MT in action. MT's process is controversial in some ways, but I believe it to be correct.

Thanks Chase. That was pretty good. Here is the thread that explains MT's projections if anyone else want s to take a look. He also mentions a FBGs article but I haven't looked for that yet:Here and Here

 

[Ministry of Pain]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

Good question chief.I typically weigh together a combination of offensive philosophy by the coaching staff from the HC to the OC, that is a pretty strong indicator. The next thing I tend to look at is pure talent. There is a reason guys like LT and Westbrook are among the very few in the league that catch 50 balls and run the ball over 250 times a year. Next I look at surrounding talent including the OL which is crucial. Finally I might peer down and look at a shcedule as well. Projections get you in trouble because I rarely see anyone who can acurately predict what someone will finish with who isn't a steady eddie like a Portis or and LT type. I simply like to gauge the upside or downside and go from there. Stats really hold owners up I find, and to be honest i don't know all that many folks that have been playing for a long long time that actually do well with projections, in fact I find mroe and more that folks show up with their own cheatsheets or pools of players and draft from that.To each their own, but I do not like the downside of projections. I do look at them but I always find that folks including myself tend to over project...everyone looks good right now in May but the reality is jsut a handful of players will make the real difference this year.

 

[gianmarco]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective. I've gotten into this with someone else and the issues with projections I have are the following:1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average. In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

 

[moleculo]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

Good question chief.I typically weigh together a combination of offensive philosophy by the coaching staff from the HC to the OC, that is a pretty strong indicator. The next thing I tend to look at is pure talent. There is a reason guys like LT and Westbrook are among the very few in the league that catch 50 balls and run the ball over 250 times a year. Next I look at surrounding talent including the OL which is crucial. Finally I might peer down and look at a shcedule as well. Projections get you in trouble because I rarely see anyone who can acurately predict what someone will finish with who isn't a steady eddie like a Portis or and LT type. I simply like to gauge the upside or downside and go from there. Stats really hold owners up I find, and to be honest i don't know all that many folks that have been playing for a long long time that actually do well with projections, in fact I find mroe and more that folks show up with their own cheatsheets or pools of players and draft from that.To each their own, but I do not like the downside of projections. I do look at them but I always find that folks including myself tend to over project...everyone looks good right now in May but the reality is jsut a handful of players will make the real difference this year.

that's a good point. Biabreakable says he projects ranges instead of static numbers - i assume that accounts for the variability you are talking about WRT upside or downside...guys like portis and LT should be pretty easy to forecast, but guys w/o an established role have a high ceiling and a high floor...where you want them ranked is a tough decision.

 

[Ministry of Pain]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective.

I've gotten into this with someone else and the issues with projections I have are the following:

1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.

Exhibit "A" your honor

2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.

Get down with your bad self

3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.

Excellent!!!

4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average.

Right

In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.

I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

GianMarco holds a PhD in MOP philosophy, or maybe it is the other way around...I couldn't have said it any better.

 

[Snotbubbles]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

Good question chief.I typically weigh together a combination of offensive philosophy by the coaching staff from the HC to the OC, that is a pretty strong indicator. The next thing I tend to look at is pure talent. There is a reason guys like LT and Westbrook are among the very few in the league that catch 50 balls and run the ball over 250 times a year. Next I look at surrounding talent including the OL which is crucial. Finally I might peer down and look at a shcedule as well. Projections get you in trouble because I rarely see anyone who can acurately predict what someone will finish with who isn't a steady eddie like a Portis or and LT type. I simply like to gauge the upside or downside and go from there. Stats really hold owners up I find, and to be honest i don't know all that many folks that have been playing for a long long time that actually do well with projections, in fact I find mroe and more that folks show up with their own cheatsheets or pools of players and draft from that.To each their own, but I do not like the downside of projections. I do look at them but I always find that folks including myself tend to over project...everyone looks good right now in May but the reality is jsut a handful of players will make the real difference this year.

This is what I do. I never do projections. I find it's better to actually watch the players play, keep track of who went where, what coaches are running things, who's coming off injuries and stuff like that. There are some players that I watch play and just don't like (like Willis McGahee) whether it be that they don't run well, they aren't used enough, their passing offense poor. Another useful tool is the player spotlights here on FBGs. It's good to hear from people who watch teams play on a more consistent basis. The Eagles are a perfect example this year of how information can be a useful tool to get a reading on a player. I love Brian Westbrook (I know Homer). But the Eagles replaced both their tackles (and quite possibly upgrade both positions), have an All-Pro player in Shaun Andrews returning from injury, had Westbrook get off-season surgery where he reports he feels the best he's felt in years. Information like that can really shoot a player up my list. Of course, I would like to see them play first to see if they have "it". I really can't describe what "it" is, but it's just a quickness when they are running with the ball. Westbrook last year didn't have it, but Correll Buckhalter did. So when I watch Westbrook this off-season, I'll be looking to see if he got "it" back. Knowing your leaguemates is another must. I've been doing this for so many years with the same people, that there are certain players that I know nobody in my league will take and those positions I can get a real steal. For the last 3 years in my PPR league, for some reason, no one wants Derrick Mason. Every year I keep picking him up and winning the league when he finishes 15th or so among WRs. And I don't have to go out and reach for a TO who maybe out-scores him by 15 points over the course of 16 games.

 

[moleculo]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective. I've gotten into this with someone else and the issues with projections I have are the following:1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average. In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

please don't mis-understand my take here. I acknowledge that projections are subjective. but, along the objective/subjective spectrum, they are more objective than pure rankings. Making projections is simply a method to apply weighted opinions of lots of variables to particular players.IMO, rankings are easy for up to basically RB10, WR10, TE5, and QB5. After that, it gets fuzzy because there isn't always a huge difference down to the next tier of guys. This hits you in the draft in about round 3-10, and this is the most critical phase. anyone can look at any cheet-sheet and get rounds 1 and 2 pretty good...they're all about the same. But, championships aren't won in the first two rounds, they are won in the mid to late rounds. This is where you find the undervalued stud, and this is where you overpay for hype, and I think that this level of players is way under-discussed here. And, this area, hopefully, is where your projections can help you discern differences between two players that are commonly valued similarly.in the Tony Gonzalez example - yeah, projecting his share of catches will be guess-work, no doubt. But, if you have a good set of projections, as you watch pre-season and see how his situation unfolds, you can adjust your projections accordingly. this, I think, is better than "bump Gonzo up"...How much do you bump him up? Past whom?

 

[moleculo]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective.

I've gotten into this with someone else and the issues with projections I have are the following:

1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.

Exhibit "A" your honor

...

GianMarco holds a PhD in MOP philosophy, or maybe it is the other way around...I couldn't have said it any better.

can you show me a set of pre-season rankings, culminated w/o using projections, that at the end of the season was significantly close to being accurate?We're all throwing darts here, this is simply a discussion of technique.

 

[Dr. Octopus]

Projections are a complete wate of time

Not if you use them to formulate your rankings.I hear what you're saying in that, yes, they are likely not going to be worth the paper their printed on by seasons end - but when you do your rankings aren't you kind of projecting the final stats anyway (without the actual numbers)?I enjoy doing it and I'm not going to claim that I take an "exact science" approach, meaning at the end of the day the toatl yards fro the WRs of one team may bot match up exactky with what I have their QB projected to total, but I do try and take all factors into consideration.What I do is look at past history and adjust according to situation (new opportunity, better supporting cast, new coaching philosophy, etc.) At the end of the day my projections entered into a spreadsheet that reflects my league's scoring system formulates my loose rankings.Once this is completed I don't necessariily draft straight off it as I take ADP and risk into consideration come draft day.

 

[gianmarco]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective.

I've gotten into this with someone else and the issues with projections I have are the following:

1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.

Exhibit "A" your honor

...

GianMarco holds a PhD in MOP philosophy, or maybe it is the other way around...I couldn't have said it any better.

can you show me a set of pre-season rankings, culminated w/o using projections, that at the end of the season was significantly close to being accurate?We're all throwing darts here, this is simply a discussion of technique.

I can't seem to find what I did with them, but I'll look around the house

 

[gianmarco]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective.

I've gotten into this with someone else and the issues with projections I have are the following:

1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.

Exhibit "A" your honor

...

GianMarco holds a PhD in MOP philosophy, or maybe it is the other way around...I couldn't have said it any better.

can you show me a set of pre-season rankings, culminated w/o using projections, that at the end of the season was significantly close to being accurate?We're all throwing darts here, this is simply a discussion of technique.

But in all seriousness, I don't think MOP or myself is trying to say that pre-season rankings without projections are better than projections. They are all relatively equal in that they are subjective and based on the skill of evaluating talent and situation among different FF'ers. The main difference is that one takes an extraordinarily longer amount of time to complete. Some enjoy them and that's great. Others do it because they think it's more objective and "better" and I think that this likely isn't the case. I think in the end, if someone is "good" at fantasy football, it doesn't matter if they sit down and do projections for every single player or just come up with their own list. They will likely still be successful more times than not. And the opposite of this is true. Projections are not going to help someone that simply isn't able to weed out the good from the bad.

I think the best way to see if projections help YOU is to come up with a set of rankings on your own without them. Then do your projections. At the end of the year, see which was more accurate. Do this for a couple years. Then you might have an answer whether or not you should continue to do them.

 

[Bankerguy]

I usually take others projections and change them. I look for opportunities between tiers etc. It's a valuable exercise, but NOT the end all be all that some make it out to be.

 

[Lucky Bucky]

Count me as one of those who does projections, although maybe not to the level of detail of others on this board. What I find advantageous about doing projections is it actually forces me to sit and think about each team and things like key personnel or coaching changes. The result is, I quickly see which players or teams I need to do more research on.

Even if the projections aren't all that accurate, I learn a lot while doing them. I guess for me it's more about the journey than the destination.

 

[Snotbubbles]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective. I've gotten into this with someone else and the issues with projections I have are the following:1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average. In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

please don't mis-understand my take here. I acknowledge that projections are subjective. but, along the objective/subjective spectrum, they are more objective than pure rankings. Making projections is simply a method to apply weighted opinions of lots of variables to particular players.IMO, rankings are easy for up to basically RB10, WR10, TE5, and QB5. After that, it gets fuzzy because there isn't always a huge difference down to the next tier of guys. This hits you in the draft in about round 3-10, and this is the most critical phase. anyone can look at any cheet-sheet and get rounds 1 and 2 pretty good...they're all about the same. But, championships aren't won in the first two rounds, they are won in the mid to late rounds. This is where you find the undervalued stud, and this is where you overpay for hype, and I think that this level of players is way under-discussed here. And, this area, hopefully, is where your projections can help you discern differences between two players that are commonly valued similarly.in the Tony Gonzalez example - yeah, projecting his share of catches will be guess-work, no doubt. But, if you have a good set of projections, as you watch pre-season and see how his situation unfolds, you can adjust your projections accordingly. this, I think, is better than "bump Gonzo up"...How much do you bump him up? Past whom?

Your post assumes that your projections are accurate. Whether you bump the player up due to feel, or projections it's all just a "best guess" scenario. What's the difference between saying "I have player X at 1000-7, but now I think he'll get 1200-10 so that will move him to #5 based on my projection" or saying "I have player X at #10, but I really like his situation, I'll move him up to #5". If someone wants to spend their time doing projections, more power to em. This is a hobby and doing projections is fun to some people so let them have their fun. But it really isn't anymore accurate then just listing out players based on feel.

 

[gonzobill5]

I like to do my own each year but I'm not naive enough to think that they are perfect. I usually start with someone else's rankings and I not only adjust based on changes and personal opinions, but also I adjust up or down for things that appear to be "flukey." For instance I thought Roethlisberger's TDs in 07 were way too high - and the pittsburgh RBs TDs were way too low based on that teams historical trends. That ranking kept me from avoiding Ben at his ADP but led me to drafting Parker too early in one league.

 

[gianmarco]

Count me as one of those who does projections, although maybe not to the level of detail of others on this board. What I find advantageous about doing projections is it actually forces me to sit and think about each team and things like key personnel or coaching changes. The result is, I quickly see which players or teams I need to do more research on.



Even if the projections aren't all that accurate, I learn a lot while doing them. I guess for me it's more about the journey than the destination.

That to me is probably the most beneficial part of doing them.

 

[moleculo]

Projections are a complete wate of time

so you prefer blind, completely subjective rankings, or do you just go with what the FBG staff says?How do you determine your pre-draft rankings then, MoP?

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective. I've gotten into this with someone else and the issues with projections I have are the following:1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average. In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

please don't mis-understand my take here. I acknowledge that projections are subjective. but, along the objective/subjective spectrum, they are more objective than pure rankings. Making projections is simply a method to apply weighted opinions of lots of variables to particular players.IMO, rankings are easy for up to basically RB10, WR10, TE5, and QB5. After that, it gets fuzzy because there isn't always a huge difference down to the next tier of guys. This hits you in the draft in about round 3-10, and this is the most critical phase. anyone can look at any cheet-sheet and get rounds 1 and 2 pretty good...they're all about the same. But, championships aren't won in the first two rounds, they are won in the mid to late rounds. This is where you find the undervalued stud, and this is where you overpay for hype, and I think that this level of players is way under-discussed here. And, this area, hopefully, is where your projections can help you discern differences between two players that are commonly valued similarly.in the Tony Gonzalez example - yeah, projecting his share of catches will be guess-work, no doubt. But, if you have a good set of projections, as you watch pre-season and see how his situation unfolds, you can adjust your projections accordingly. this, I think, is better than "bump Gonzo up"...How much do you bump him up? Past whom?

Your post assumes that your projections are accurate. Whether you bump the player up due to feel, or projections it's all just a "best guess" scenario. What's the difference between saying "I have player X at 1000-7, but now I think he'll get 1200-10 so that will move him to #5 based on my projection" or saying "I have player X at #10, but I really like his situation, I'll move him up to #5". If someone wants to spend their time doing projections, more power to em. This is a hobby and doing projections is fun to some people so let them have their fun. But it really isn't anymore accurate then just listing out players based on feel.

Well, one difference is that if you bump a player up, that should have an impact on his teammates - in the case of a WR, he may be bumped up because you project him to have a bigger role (and therefore all of his teammates should have a decreased role and therefore should be bumped down), or you expect him to do more with what he gets (in which case his QB should see a slight bump).If we decide to bump up gonzo, should we bump down Roddy White? What impact will Gonzo have on Ryan? Turner?

 

[gianmarco]

They aren't "blind". They are based on watching football and paying attention each year what particular players do and what particular teams do.As to the subjective part, even projections are subjective. Just because you're using actual #'s and stats and basing your projections on those, you are still doing your OWN SUBJECTIVE TWEAKS based on what YOU think the change will be. That is not objective. It's an attempt to cover up subjectivity and make it LOOK objective. I've gotten into this with someone else and the issues with projections I have are the following:1. They usually just aren't accurate. If someone can show me a preseason set of projections and then at the end of the season show me how the majority of those fell right into place or were significantly close, I'll start to listen. Until then, I'm just not buying it.2. 3 different people can come up with 3 completely different sets of projections for a player and there can be good enough reason to justify all 3. That, in and of itself, is why projections are still subjective. Pull up any player spotlight thread and you'll see one guy projecting Steve Smith for 90/1300/9 and another guy projecting him for 60/850/5. Obviously one guy likes him and one guy doesn't. Sure, it LOOKS objective, but it's not.3. Certain team changes and personnel changes are just too drastic to try and apply projections using previous years as a baseline. The example that came up with someone else was Tony Gonzalez in Atlanta. You simply can't take what Atlanta did last year with their TE's and use that as a projection for what Gonzo will do. Justin Peele led all TE's for Atlanta with 15 catches. Any sort of projection that is made using that is going to be sheer guessing. Whether you think Gonzo has 40 catches or 60 catches or 80 catches or 100 catches, it's a complete "close your eyes and throw a dart" projection. Unless you start looking at previous years when Crumpler was in town (but with a different coaching staff) or another team that is modeled similarly to Atlanta's current situation (SD? Dallas? Houston?), these types of projections simply aren't objective.4. The projections that are most likely to be accurate are the ones that are relatively unchanged from last year (i.e. static situations). In that case, you don't NEED projections. If you can tell that things will most likely remain unchanged), then I don't need a projection that takes last year's #'s and shifts it 5-10% in either direction based on the way the wind is blowing. I can tell that things won't change drastically and use last year's #'s as a good average. In the end, I feel that talented players are going to get the ball no matter what with very few exceptions (and those exceptions are going to be obvious for the most part). It is VERY RARE for true talents to not perform well due to lack of opportunity as long as they are on the field. I don't need to do passing projections for Atlanta to know that Gonzo is going to get the ball. I likewise don't need to do them for Tampa Bay because I know that Winslow is going to get the ball. If Boldin were to leave town, while he obviously won't repeat his crazy Arizona #'s unless he lands in an equally ideal passing situation (Indy, NE), he will still get balls thrown to him even if he were to land on a team like Tenn or Baltimore. Good players get the ball. The few exceptions to that are when you have a guy like Lee Evans who just didn't have a QB that COULD get him the ball.I'm not saying projections are 100% useless. For some people, they provide a detailed and organized means to rank players and sort out situations by seeing it. For others, having the experience and practice of seeing what works and what doesn't, which player is going to succeed and which won't, which situation is going to lend itself to production and which won't, projections only serve as a way of complicating an otherwise uncomplicated process.

please don't mis-understand my take here. I acknowledge that projections are subjective. but, along the objective/subjective spectrum, they are more objective than pure rankings. Making projections is simply a method to apply weighted opinions of lots of variables to particular players.IMO, rankings are easy for up to basically RB10, WR10, TE5, and QB5. After that, it gets fuzzy because there isn't always a huge difference down to the next tier of guys. This hits you in the draft in about round 3-10, and this is the most critical phase. anyone can look at any cheet-sheet and get rounds 1 and 2 pretty good...they're all about the same. But, championships aren't won in the first two rounds, they are won in the mid to late rounds. This is where you find the undervalued stud, and this is where you overpay for hype, and I think that this level of players is way under-discussed here. And, this area, hopefully, is where your projections can help you discern differences between two players that are commonly valued similarly.in the Tony Gonzalez example - yeah, projecting his share of catches will be guess-work, no doubt. But, if you have a good set of projections, as you watch pre-season and see how his situation unfolds, you can adjust your projections accordingly. this, I think, is better than "bump Gonzo up"...How much do you bump him up? Past whom?

Your post assumes that your projections are accurate. Whether you bump the player up due to feel, or projections it's all just a "best guess" scenario. What's the difference between saying "I have player X at 1000-7, but now I think he'll get 1200-10 so that will move him to #5 based on my projection" or saying "I have player X at #10, but I really like his situation, I'll move him up to #5". If someone wants to spend their time doing projections, more power to em. This is a hobby and doing projections is fun to some people so let them have their fun. But it really isn't anymore accurate then just listing out players based on feel.

Well, one difference is that if you bump a player up, that should have an impact on his teammates - in the case of a WR, he may be bumped up because you project him to have a bigger role (and therefore all of his teammates should have a decreased role and therefore should be bumped down), or you expect him to do more with what he gets (in which case his QB should see a slight bump).If we decide to bump up gonzo, should we bump down Roddy White? What impact will Gonzo have on Ryan? Turner?

Again, I don't need projections to try to answer those questions. When Gonzo got traded to Atlanta, I didn't just think of the impact it had on Gonzo. I thought of what it did for Ryan, Roddy, Turner, and others. If you need projections to figure that out, that's fine. Nothing wrong with it. I have heard the argument that there won't be enough targets to go around with him in town. Well, we'll just see. But just because I decide to bump up Gonzo upon his arrival in Atlanta doesn't mean I can't try to figure out what it does to the others without having to do projections. To me, it's more helpful to compare Atlanta's situation to another team with a similar situation to see how it might affect others. Looking at teams like SD, Dallas, Houston, and Cleveland from 2 yrs ago shows me that you can have a successful top RB, WR, and TE and still have all do well at the same time. In fact, most of the top TE's have had exactly that situation with the exception of Gonzo. It's pointless to me to base it on what Atlanta did in 2008 with Hartsock and Peelle as the TE's.

 

[gonzobill5]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

 

[moleculo]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

 

[Snotbubbles]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

If I only had to do one team, I'd do the Eagles. I could probably spit that out in about 30 minutes.

 

[mlball77]

I agree with the several posters that have commented on the fact that preparing projections, at least for me, is not at all about trying to get things exactly right for each player as much as it is an in depth review and analysis of all of the things I can think of that will impact a player's production (and a guesstimate as to what level). The process really makes me examine why I feel the way I do about a player, and might actually change that initial feeling.

The one year I did full projections, I feel I greatly benefitted from it and had a very good understanding going into the season of what each team was likely to do on offense... but of course that all changes with injuries and various other things.

But point being, I feel that going through the process of projections was very beneficial to me and helped me understand things more thoroughly. That said; I don't do them typically, as the benefit probably wasn't worth the amount of work required. And I definitely understand the fact that many folks can really wrap their brains around how a team and player is likely to fair in the upcoming season without going through the painstaking process of projections. However, for me, and probably many out there that do their own projections, it provides definite benefit.

 

[boubucarow]

How do some of you that do not do any sort of projections account for different fantasy scoring systems?

 

[gonzobill5]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

With all the interest generated from the mock draft, I'd be surprised if we couldn't get 32 people together. I'd be willing to take on any team (I think it would be easier to avoid the Eagles so my personal biases wouldn't be represented). What can we do to align our methods? Perhaps we need to make sure our teams totals can't go above or below a certain percentage of previous years totals.

 

[gianmarco]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

With all the interest generated from the mock draft, I'd be surprised if we couldn't get 32 people together. I'd be willing to take on any team (I think it would be easier to avoid the Eagles so my personal biases wouldn't be represented). What can we do to align our methods? Perhaps we need to make sure our teams totals can't go above or below a certain percentage of previrs totals.

That's the problem and limits of projections. Some things can and do change drastically. Limiting what you do is just taking last year and tweaking. It doesn't take into account key personnel changes. When Martz arrived in Detroit, you couldn't just take the previous year and tweak it. You had to make major changes. Likewise when Turner went to Atlanta or T. Jones to NYJ. Same will happen with Gonzo in Atlanta and Winslow in TB.

 

[moleculo]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

With all the interest generated from the mock draft, I'd be surprised if we couldn't get 32 people together. I'd be willing to take on any team (I think it would be easier to avoid the Eagles so my personal biases wouldn't be represented). What can we do to align our methods? Perhaps we need to make sure our teams totals can't go above or below a certain percentage of previous years totals.

the way I envision it is some GUI type sliders embeded in Excel. you move sliders, and the various parameters update accordingly. As long as the limits for the sliders are set appropriately, people will be limited from projections that are unrealistic - i.e. no one should project an established RB to have a YPC of > than their career max * 10%, or a teams run/pass ratio can't be more than 15% different than it was last year...I have a rough draft version of this done, but it needs some tweaking still (I've been thinking about how to improve this for quite a while now...)I think that if we have an established template that everyone agreed to and worked within, it could all come together nicely.

 

[Biabreakable]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

With all the interest generated from the mock draft, I'd be surprised if we couldn't get 32 people together. I'd be willing to take on any team (I think it would be easier to avoid the Eagles so my personal biases wouldn't be represented). What can we do to align our methods? Perhaps we need to make sure our teams totals can't go above or below a certain percentage of previous years totals.

the way I envision it is some GUI type sliders embeded in Excel. you move sliders, and the various parameters update accordingly. As long as the limits for the sliders are set appropriately, people will be limited from projections that are unrealistic - i.e. no one should project an established RB to have a YPC of > than their career max * 10%, or a teams run/pass ratio can't be more than 15% different than it was last year...I have a rough draft version of this done, but it needs some tweaking still (I've been thinking about how to improve this for quite a while now...)I think that if we have an established template that everyone agreed to and worked within, it could all come together nicely.

I will do a team once this is settled. Should make for some interesting discussion. Thanks for the great idea!

 

[gonzobill5]

Curious to see who does their own projections and what kind of tips/tricks you use. Being a math/stats guy, I'm actually just as interested in the process as I am the results; the process is rarely talked about and I think that's a shame.

I'm also interested to see if anyone is willing to participate in a group effort to lessen the burden. I've got some ideas for how to automate a system, where people can only focus on teams they are comfortable with. I think it's too tedious to do make accurate projections for 32 teams - it's just too much for one person to be an expert on all teams.

I forgot to say - I'd be willing to help in doing this. For it to be worthwhile, we should come to a clear consensus on our methods so it doesn't become a bunch of homers overvaluing their favorite players.

Thanks. I agree about homers overvaluing their players. Ideally, the way I envision things is that we get 32 guys - one for each team. Then, the homer aspect cancels itself out. From the tone of this thread though, I don't think I'll be able to find 30 other people who are interested in doing projections.any help is welcome though...I definitely think a collaborative effort is the way to go. Last year my projections broke down at about WR40 or so, and I was left drafting WR's from teams that I had no real understanding on what was going on - Seattle and Chicago, specifically. I was drafting injured players because I wasn't up to date on their status, I had Hester too high because someone had to catch the ball (right?)...that's where it got tough, and I realized that there was no way I could be competent for 32 teams.

With all the interest generated from the mock draft, I'd be surprised if we couldn't get 32 people together. I'd be willing to take on any team (I think it would be easier to avoid the Eagles so my personal biases wouldn't be represented). What can we do to align our methods? Perhaps we need to make sure our teams totals can't go above or below a certain percentage of previrs totals.

That's the problem and limits of projections. Some things can and do change drastically. Limiting what you do is just taking last year and tweaking. It doesn't take into account key personnel changes. When Martz arrived in Detroit, you couldn't just take the previous year and tweak it. You had to make major changes. Likewise when Turner went to Atlanta or T. Jones to NYJ. Same will happen with Gonzo in Atlanta and Winslow in TB.

My bad, that was a dumb statement. On second thought we could just let each individual support his own projections and let the ensuing debates settle anything that seems out of whack.

 

[Jasonb]

Some things I have trouble with include injuries - I don't think it's right to project everyone out 16 games, but it's not right to project injuries either. Even if you do assume a player plays all 16 games when he did not last year, it's tough to figure out the right number of carries everyone on the team should get.I also struggle with YPcatch, making that jive with QB passing yards, and making this all balance.

Anyone thought about averaging the number of games played per Year, to get an overall Points per Gm and then averaging that number for the past 3 years. This should account for injury prone players.The really hard part is ranking ROOKIES.Personally I have tried all sorts of ranking systems and have found the best system is to find a good base projection for the up coming year. Build your excel file to account for your scoring system...for example my system awards points for every yard earned...not the 1pt for every 20. This sets the bar, I then include the Strength of Schedule for weeks 1-14 and also for 15-16 for the playoffs. Note: my SOS is based on pts per position not by team. This allows me to rank my players by both potential and opponents they will face...Because FF is all about who your players are facing... If your best players are always facing stout defenses...then you can't truly expect high numbers from those players. This may skew your projections if that doesn't appear to already be factored in. Finding the studs with weak schedules is how I have won my leagues. You have to analysis your ranking to their schedule. For example 2 years ago JON KITNA was the QB for Detriot...and was ranked really low for QBs. I picked up late in the draft and he ended up being really productive. Why was he productive...because he had an easy schedule, and quality WRS and no running game. Good chance they were going to have to throw a lot with their defense and nonetheless that is eactly what they did. He ended up ranking 3rd in total points for QB just behind Manning and Brees. Best part is I didn't draft him until my 2nd to last pick because everyone else had picked up their QBs and I was able to fill depth at my other positions rather than spend a mid to high pick on a high ranked QB. These are the gems you have to find using research to win! Note: who claims the QB position is DEN is going to be a top-10 QB - defense really didn't get better...not until they fix the D-line. So they will be scored on a lot this year...and that will force DEN to pass, despite there 45 RBs. Should be a great sleeper this year.

 

[MAC_32]

How do some of you that do not do any sort of projections account for different fantasy scoring systems?

Take my master file, 'save as' league with customized scoring, review positions bumping up and down according to that league's specialized scoring. A review of previous season's scoring serves as a good guideline as to what positions may be more valuable than others. Two years ago I took the time to create projections because of this factor and felt the time to create the projections (A LOT of hours) was not worth it. I think I'd benefit more spending that time analyzing and reading.

 

[Carolina Hustler]

Doing Projections can be a very important tool. Sure it isn't going to be 100% accurate, but neither will making a list in order of who you like. And ultimately, making a list in order of who you like is going to be based on some projection right? Or else how would you determine the order? You're guessing who will do better based upon stats, age, team tendencies... etc.... projecting...

So, the Pros-

These projections can be then calculated into any scoring format. How many of us use the same scoring format in EVERY league? Not many I'm sure...

You also have given some thought to every situation. Double checking your work...

The cons -

Took a little extra time? Or did it really? Now you only have to rank them once, then use an excel sheet to calculate the ranking for each league.....

I am far to busy/lazy to do this but if I had the time to commit I believe I would do it like this -

1) Start with your favorite site's player projections, go line by line and make adjustments with all things considered and tailor to your calculations or suspicions.

2) Transfer these projections to a cheat sheet for each league based upon particular league scoring. (excel sheet)

3) Run through that sheet and color code the players for tiers and potential variances due to injury, stability, age...etc

4) On a separate worksheet I'd take current ADP and make adjustments for each league based upon league tendencies/history

5) Compare the adjusted ADP sheet with the league calculated rankings sheet, and look for value.

I think it's a great idea, having several people collaborate, but it would be hard for me to trust rankings I didn't create if they where based on only one persons opinion. In order to make this work I think you'd have to have some over lap. Maybe 15 people doing 6 teams each so that each team is considered by 3 different people and use the average.... that would somewhat make up for a few bad projections.

 

[moleculo]

so I've got a template that can be used to (hopefully) easily generate realistic projections for any team. I use last years stats as a starting point, and have some GUI buttons that vary parameters such as YPC or % of teams rushes. it's all set up to balance out - i.e. passing yards = sum of all receiving yards, pass completions = sum of all receptions, etc.

It starts with establishing total # of plays and run/pass ratio, and I have some guidelines for how the league ran in general (avg +/- stdev, max, min) for these parameters. After these are ok, you move to adjusting the parameters for the players. To help set parameters, it shows data from the past 3 years for any player of interest (using a drop-down box), just to provide some context.

The players listed for each team are not up to date. But, all you have to do is manually type in the name of the players in the cell that the existing players are listed in, and their stats update accordingly.

I have made the convention that for initial values at least, % of workload (carries, passes attempted, or catches) stays static with the position and team - i.e. RB1, RB2, etc. Performance parameters (YP rush attempt, YP pass attempt, YP catch), move with the player. Again, these are to establish initial values only, the parameters can be updated as you see fit.

If you are interested in seeing this, send me a PM with what team you want to work with and your e-mail, and I'll be happy to send you the files. I'd like to use May and early June to debug, and start with live projections some time in late June. If you want to help with projections when it's time to go live that's great. If you don't and you just want to see the files and provide some feedback, that's fine too. It's not perfect, but I'm liking the direction this is going.

 

[CBower4545]

I think some of you are missing the point about doing your own projections. Once you put your thoughts into numbers you can test those numbers, put them in historical context and see if they make sense. I find this makes projections more interactive. You can test and revise them yourself over and over again. This makes for a good mental exercise and sharpens your perceptions imo. When I make a list I don't find myself revising it nearly as much. Another advantage of projecting every player, is you get to look at the big picture and analyze what that means about your projections. With simply doing rankings sometimes you can't see the forest for the trees.

 

[robb]

so I've got a template that can be used to (hopefully) easily generate realistic projections for any team. I use last years stats as a starting point, and have some GUI buttons that vary parameters such as YPC or % of teams rushes. it's all set up to balance out - i.e. passing yards = sum of all receiving yards, pass completions = sum of all receptions, etc.It starts with establishing total # of plays and run/pass ratio, and I have some guidelines for how the league ran in general (avg +/- stdev, max, min) for these parameters. After these are ok, you move to adjusting the parameters for the players. To help set parameters, it shows data from the past 3 years for any player of interest (using a drop-down box), just to provide some context.The players listed for each team are not up to date. But, all you have to do is manually type in the name of the players in the cell that the existing players are listed in, and their stats update accordingly.I have made the convention that for initial values at least, % of workload (carries, passes attempted, or catches) stays static with the position and team - i.e. RB1, RB2, etc. Performance parameters (YP rush attempt, YP pass attempt, YP catch), move with the player. Again, these are to establish initial values only, the parameters can be updated as you see fit.As a fellow Bronco fan who lives in Estes Park, I would be interested in what you have done so far. Sounds like a lot of work!Is yours "customizable" so I could plug in my league's scoring system from my league and arrive at my own VBD spreadsheet? I might trade you for my customized Defense By Committee spreadsheet.Do any of you others out there have this so I could plug in my league's scoring system also?Thanks everyone!If you are interested in seeing this, send me a PM with what team you want to work with and your e-mail, and I'll be happy to send you the files. I'd like to use May and early June to debug, and start with live projections some time in late June. If you want to help with projections when it's time to go live that's great. If you don't and you just want to see the files and provide some feedback, that's fine too. It's not perfect, but I'm liking the direction this is going.

 

[Biabreakable]

so I've got a template that can be used to (hopefully) easily generate realistic projections for any team. I use last years stats as a starting point, and have some GUI buttons that vary parameters such as YPC or % of teams rushes. it's all set up to balance out - i.e. passing yards = sum of all receiving yards, pass completions = sum of all receptions, etc.It starts with establishing total # of plays and run/pass ratio, and I have some guidelines for how the league ran in general (avg +/- stdev, max, min) for these parameters. After these are ok, you move to adjusting the parameters for the players. To help set parameters, it shows data from the past 3 years for any player of interest (using a drop-down box), just to provide some context.The players listed for each team are not up to date. But, all you have to do is manually type in the name of the players in the cell that the existing players are listed in, and their stats update accordingly.I have made the convention that for initial values at least, % of workload (carries, passes attempted, or catches) stays static with the position and team - i.e. RB1, RB2, etc. Performance parameters (YP rush attempt, YP pass attempt, YP catch), move with the player. Again, these are to establish initial values only, the parameters can be updated as you see fit.If you are interested in seeing this, send me a PM with what team you want to work with and your e-mail, and I'll be happy to send you the files. I'd like to use May and early June to debug, and start with live projections some time in late June. If you want to help with projections when it's time to go live that's great. If you don't and you just want to see the files and provide some feedback, that's fine too. It's not perfect, but I'm liking the direction this is going.

PM sent.This should be fun.