Projections really bother me...and they're wrong! (1 Viewer)

[ceo3west]

If there's 50% turnover in the top 10 at RB as you say, then that means:

5 of this year's top 10 will be from last year's top 10

5 will be from last year's 11-1000000000

In other words last year's top 10 is way way more likely to be in this year's top 10 than anyone else. I'm really not sure what your point is here.

My point is that based on standard projections, you have a 50% chance of picking a bust with your top pick and your season being in jeopardy. I use my gut to pick a guy that I think will have a big year, not LJ or Alexander because he's next on the list and I might get ridiculed if I don't take one of them.

I promised MoP to stop stealing his thread but this is just crazy. First off, there are degrees of busts. If I draft somebody at the #8 spot and they finish as the #13 did that kill my season? Is that player a bust? You're not using your gut. You're using information at your disposal. ESPN, FBG's, etc. You're using things to come to conclusions then passing it off as your gut.

The players that drop out are usually because of injury. Since 5 of the top ten players will drop out, please tell me which ones. Since the ones to fall will be injury based, please tell me what kind of injury they will suffer and when it will hit.

But not always. And that's what I try to look at - which guys have circumstances which are not that favorable this year and are not a safe pick in the top tier. Alexander was one last year I hated - the previous year he declined after his big contract, and guys were lining up at #6 to grab this guy last year thinking about a bounceback. I saw an aging RB who got his fat check and doesn't exactly have an LT-type work ethic - it showed up in his stats last year, and guys looking for 2005 numbers got burned. You can call his gimpy ankle a "freak" injury, but the fact is guys who pick down a list of top 10 projections and are expecting previous results without looking at unfavorable factors which affect performance get what they deserve.

 

[twr]

first off, when doing projections, they are basically just an educated GUESS... nothing more.. there are SO many factors that go into how good of a season a player will have. with that being said, successful projections are the ones that account the factors that are visable.. skill(size,speed,strength,will...ect), coaches(training,philosophies), talent around them(good oline, good #2 wr to draw coverage, good qb throwing the ball to me.. ect)... knowledgeable projectors factor in a number of different things, so they can formulate a guess.. this guess could be actual projected numbers... or it could be just a general standing on how well they will do... but both cases are based off of success... which directly corresponds to statistics...

successful projections cant be based off the invisible.. injury, 400 carry myth, rookie regression myth..ect

the only way to account for the invisible would be to have a ceiling and a floor for all players.. which would be very time consuming. however, just as only having successful stats goes, there will alway be variations to them... personally i dont see LJ preforming bad, or getting injured again. so, i wouldnt be inclined to include a floor projection for him... but others might.

this hobby is based off personal belief.. thats why player values are very very different from person to person. i guarantee you will never find another person who feels the exact same way about EVERY player.

FBG does a great job giving a ton load of info, which i then compare to things i already know, and have opinions about

personally i think your argument is very petty because your simply ranking players by feeling, and not stats... but this game is based off stats... and projections are feelings AND stats mixed together to try and PREDICT a future outcome.. in which one CANT factor in the invisible

look at it like this.. alot of things in life have luck... some people dont believe in luck, but i do... hard work diminishes the luck factor, but everything has luck in it...

as an added note, im pretty sure your going to be wrong about the roddy white projection, as well as the dunn projection... but guess what, thats what they are, projections

 

[Chairshot]

This honestly seems like a really stupid argument.

The only difference between guys who use/generate projections and guys who draft "from the gut" (which, by the way, is a absolutely silly term to label that style of drafting as) is that the projection guys actually distill what they think will happen into specific yards or TDs or catches. The "from the guy" guys stop short of doing that because they prefer not to look at the details.

For someone to say that the projection guys don't look at O line information or coaching information or injury history or "intangibles" means one of two things: that person is ignorant of how good projections are generated or they are looking at a set of half-assed projections. No good projections are just taking last years list and putting it back up for this year. I'm sure just as many "from the guy" guys do that as projectors do.

Everyone uses the methodology that they think is best for judging how players will play in the upcoming season. The information projection guys use is different for each projector - that's why there are a million sets of projections out there. It's the same thing for "from the guy" guys - you have your methodology for figuring how players will play in the upcoming season. Each "from the gut" methodology is just as different as all the projectors methodologies, I'm sure.

It's all the same stuff, boys. It just depends on what format you like to see the results in.

That said (and someone mentioned this before), if the turnover for top 10 production each year is 50%, you are *still* better off just going off last years projections - you have 50/50 shot of getting a top 10 player, which is better then if you had to randomly pick players based on chance.

Somewhere it was mentioned that most projections are only different by 10% from the previous season. In my opinion, that's the 10% that wins you your league. If you can get a 10% edge on all the other owners in your league, that's significant.

The goal of any projections or gut feelings is to make that 10% into 15%. Or 20%. Or more. That's how you win a league.

Still, to me, regardless of how you generate your rankings, the most important thing is ADP. Because no matter how you sort your players, you need to know where the value in a draft lies. If you are only marginally right on your guesses for the year but you destroy everyone on ADP, you will have a very good chance to win, I believe.

 

[barista]

There is no right answer with projections. Everyone should discount their lists on their own otherwise where's the fun? I haven't met any infallible ff jockeys. I think I make the most sense to myself... that's why I let me do my own draft.

THIS IS ALSO WHY ADP IS AS IMPORTANT AS ACCURATE PROJ.

 

[joffer]

This honestly seems like a really stupid argument. The only difference between guys who use/generate projections and guys who draft "from the gut" (which, by the way, is a absolutely silly term to label that style of drafting as) is that the projection guys actually distill what they think will happen into specific yards or TDs or catches. The "from the guy" guys stop short of doing that because they prefer not to look at the details.For someone to say that the projection guys don't look at O line information or coaching information or injury history or "intangibles" means one of two things: that person is ignorant of how good projections are generated or they are looking at a set of half-assed projections. No good projections are just taking last years list and putting it back up for this year. I'm sure just as many "from the guy" guys do that as projectors do.Everyone uses the methodology that they think is best for judging how players will play in the upcoming season. The information projection guys use is different for each projector - that's why there are a million sets of projections out there. It's the same thing for "from the guy" guys - you have your methodology for figuring how players will play in the upcoming season. Each "from the gut" methodology is just as different as all the projectors methodologies, I'm sure.It's all the same stuff, boys. It just depends on what format you like to see the results in. That said (and someone mentioned this before), if the turnover for top 10 production each year is 50%, you are *still* better off just going off last years projections - you have 50/50 shot of getting a top 10 player, which is better then if you had to randomly pick players based on chance.Somewhere it was mentioned that most projections are only different by 10% from the previous season. In my opinion, that's the 10% that wins you your league. If you can get a 10% edge on all the other owners in your league, that's significant.The goal of any projections or gut feelings is to make that 10% into 15%. Or 20%. Or more. That's how you win a league.Still, to me, regardless of how you generate your rankings, the most important thing is ADP. Because no matter how you sort your players, you need to know where the value in a draft lies. If you are only marginally right on your guesses for the year but you destroy everyone on ADP, you will have a very good chance to win, I believe.

what he said

 

[NoFBinLA]

I use projections as merely another tool to get the best lineup drafted as possible. Rankings are nice - but sometimes I like to see what the rankers (ranker-people? rank and file?) think a player will do as well.

It's one thing to see someone rank McGahee at 10 and Lynch at 11. It's another to see that the guy also predicted 500 more yards and three more TDs for McGahee as opposed to 50 yards and a TD. Shows me where the separation is and how severe they think it is.

No they aren't infallible by a long shot. But if a guy throws out projections WITH his rankings, it can give me a clearer understanding as to how serious they are about one guy or another.

All just tools for my ownself though. In the end, it's my team and my choices.

 

[Gatorman]

This honestly seems like a really stupid argument.

It might, but it isn't and it comes up every year about this time.The gut guys vs. math masturbators is really just guessing, and it all depends on where in the process you want to guess.Even Kid C's Average value stuff can give you great numbers that matchup well long term, but then you need to "guess" at who fits the particular slot.It is all guesswork, and most of it is educated, but it still should be looked at in more than one way...

 

[ratbast]

I use tiers and have for many years, for the same reason you don't like projections, however I "project" what tier they qualify for.

I do not apply numbers but slot players into tiers based on what lineup slot I project him to be on my roster (RB1, RB2 etc.). If I feel confident in their ability to be an RB1 then they are slotted RB1 (my RB1 tier usually has a few sub tiers). I sometimes brake down the tiers into RB4U (U for upside) and plain RB4 but that is as much about my keeper standards as anything.

The beauty of my tiers is as I go along I can see if a player is available that shouldnt be, I see where positions run deep AND my tiers can adapt to how my draft is turning out at that point.

ADP is fine if you play in a hyperleague but ADP goes out the window when it comes to my 15 year old keeper leagues as many early picks are out in left field. So "average draft position" makes no sense as these owners are not polled and are by far not average.

 

[notdarkyet]

Is MOP arguing that there are people that draft strictly on their top200 list no matter what, without evaluating several talents at the same "level"?

 

[T Bell]

This honestly seems like a really stupid argument. The only difference between guys who use/generate projections and guys who draft "from the gut" (which, by the way, is a absolutely silly term to label that style of drafting as) is that the projection guys actually distill what they think will happen into specific yards or TDs or catches. The "from the guy" guys stop short of doing that because they prefer not to look at the details.For someone to say that the projection guys don't look at O line information or coaching information or injury history or "intangibles" means one of two things: that person is ignorant of how good projections are generated or they are looking at a set of half-assed projections. No good projections are just taking last years list and putting it back up for this year. I'm sure just as many "from the guy" guys do that as projectors do.Everyone uses the methodology that they think is best for judging how players will play in the upcoming season. The information projection guys use is different for each projector - that's why there are a million sets of projections out there. It's the same thing for "from the guy" guys - you have your methodology for figuring how players will play in the upcoming season. Each "from the gut" methodology is just as different as all the projectors methodologies, I'm sure.It's all the same stuff, boys. It just depends on what format you like to see the results in. That said (and someone mentioned this before), if the turnover for top 10 production each year is 50%, you are *still* better off just going off last years projections - you have 50/50 shot of getting a top 10 player, which is better then if you had to randomly pick players based on chance.Somewhere it was mentioned that most projections are only different by 10% from the previous season. In my opinion, that's the 10% that wins you your league. If you can get a 10% edge on all the other owners in your league, that's significant.The goal of any projections or gut feelings is to make that 10% into 15%. Or 20%. Or more. That's how you win a league.Still, to me, regardless of how you generate your rankings, the most important thing is ADP. Because no matter how you sort your players, you need to know where the value in a draft lies. If you are only marginally right on your guesses for the year but you destroy everyone on ADP, you will have a very good chance to win, I believe.

 

[Liquid Tension]

Every year for the past 5 years I have tried to scream from every available bullhorn, in every preseason ranking thread I start, in every possible way I know of I have tried to voice my displeasure with the entire projection system that permeates these boards. I even see websites out their boasting about how great they were at the projections. I would really have to see what their summer projections were and what they ended with to see how some of these sites are scoring 100. I do not believe for a second that they got all the stats right down to the yds and TD for all the players. It is impossible.There are so many holes in the FBG projections, and I'm not picking on them but I pay for a membership every year and this is where I do the majority of my bantering so I'll use them as an example. For instance, Warrick Dunn had his worst year as a RB last year playing on a sinking ship team in the Atlanta Flacons and he still finished 30th. He has been between 15-24 for most of his career. A very undervalued RB2 for the past 10 years in this league. I'll take this a step further...when he was in TB he had reception totals of 44, 68, 44, 54 and then 50 his 1st season in Atlanta...then I don't know if it was Vick, the HC Mora, or what happened but he became invisible in the passing department where his totals slipped under 30 a year. Now he signs a $3 million deal with the Bucs who did not sign him to simply watch him walk around the sidelines as some websites suggest I assure you. What will his totals be? I haven't the foggiest of clues but I do know that I would bet money he finishes well ahead of the ranking he has at 45 this year, and likely will far outperform his draft spot of somewhere in the 12th round or later. I can't put everything I am talking about into his stats but I do think thru some cognitive thinking that one should easily derive that Dunn = value this year, tons of it I might add. What track record does Earnest Graham bring to the table? Dunn could easily find himself getting 12-15 carries and 3-5 catches a game quite a bit this season. So let's land the plane a bit and focus on what I am talking about becasue this deserves some serious debate. I understand it is easy to project out for guys like LT, TO, and the like because they are so GD consistent you can't be off by more than 10% one way or the other, fantastic. But then I see people just loading in Roddy White's stats from a year ago and I want to scream WTF are you thinking? Projections are very dangerous to put on players that have no track record or history to really go on. And if you think the same top20 at WR last season is going to be the same top20 WR this season, you are delusional and you are going to miss the playoffs this season because you won't get your head out of your you know what. Think people, think!Put the stats down and try and think logically about what you are going to do on draft day. What do you do, you say? Let me try and steer you into some different thinking. Get a firm grip on the OL for all 32 teams, rank em, file em, put em down on paper. Use all resources you can to determine strength on these very important pieces of the puzzle. Don't fall for what everyone is writing if it goes against what you believe. For example, I am not sold on the Bucs and Raiders OL the way others seem to be. I could be way off but I am not ready to say these OL are good-great just yet. Next you need to seriously think about the coaching staff and what/whether they like to run/pass. Offensive Philosophy is so critical. There is actually 2 people running the OC position in Detroit...did you know that? Has this ever worked before? And they have come out and said flat out they are not going to pass the way Martz did...they are sending out the "Conservative Approach" if you listen to anything they have to say. What does that do to Roy Williams who some in here think is bound for 1,300 yds and 10 TD...what about Calvin Johnson and his progression in year 2? I don't have all the answers but these are the things you need to think about. Not just crunching out Roy Williams numbers if and when he plays a full 16 games.So you got the OL, coaching philosophy down, next I would look at surrounding talent in the skill positions especially QB. Next I would look at the schedule and see if you can find a soft schedule and also who has the harder schedules as it will matter. And finally I might look at the intangibles. LT for example simply has intangibles that work in his favor. He could have no OL, QB, a pass happy OC, and a tough schedule, he would still finish top5...you catch what I'm saying here?Then I would take these categories...and you can make up others but I try to limit it to around 5, then divide up the 5 categories and give them different weighted averages, sort of like the grades you got in school. Then I look and see what I have, put them into tiers based on what I uncover with a nonprojection system, then I figure out where I have to draft which player in order to get who I want. I'll hang up and listen.

A couple of pointsunn: I think projecting points and possibly even finish may not lead to the value he brings to a fantasy team. The truth is that a guy who puts up 7 points a game will finish with 112 points and finish about 30th. But what kind of value does that really bring to a fantasy team? I would much rather have a guy who finished below Dunn but when he played he was more productive. D. Ward didn't have 100 points for the year but he would be way more valuable to a team than a guy like Dunn. Replacement value is more important than total points.I would caution against predicting defenses too much at this early stage. Every year we see a change in team defenses and while a very solid D will not get torched, it is not prudent to base a pick on the teams defensive schedule as we see in hindsight they are not very accurate.While you can't simply load last year stats into the data base for next year (your Roddy White example), looking FOR track record will make you a hindsight owner instead of setting the trend. Sure it is best to see a guy who had an off year that people have soured on but you know there were extenuating circumstances and he will do better if healthy this year, but if you can see "greatness" in a player before others this is how you really jump ahead.I am not really sure what you are displeased about or would like to see, but I do agree with the other factors you take into account

 

[Liquid Tension]

I think I understand what they're getting at. I know VERY SUCCESSFUL high stakes players that have NEVER projected stats.

ADP is their main warrior and they go off gut feel from there. Statistical projections are helpful, but I find a system of balancing upside and downside to be useful as well.

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right? I mean, your best guess at what they'll produce is what you should go with right? Not so fast... I'll take STEWART all day long due to the upside potential. Why is Forte's upside so low? The offense in Chicago... Why is Stewart's upside so high? The offense in Carolina, and countless other scenarios and possibilities that affect your upside / downside rankings. Maybe it's a backup with an injury prone starter, or maybe its a player who hasn't proven it, but a history of that offensive system shows a big time number potential (Broncos).

Simple projections are good for those wanting a cheatsheet and a quick reference before draft day, but for those of us more serious about it, I think MOP and Gatorman are on to something. Maybe more information like above would help.

Very

 

[Uruk-Hai]

I've become less a "do you own projections for every player" guy as the years have passed. Part of that is laziness & part is the fallibility of them (fully realize that Mr Gut can be the most fallible being on Earth) - in other words, I just don't want to spend the time on something I'm going to be mostly wrong on anyway. If I want to look at projections, I come here - more for the macro view than for any individual player - as the staff can do at least as good a job as I can.

That being said.....

How do the non-projection guys decide between a RB, QB, or WR when all are at the top of their respective tiers (assuming "need" is equal)?

 

[ceo3west]

I've become less a "do you own projections for every player" guy as the years have passed. Part of that is laziness & part is the fallibility of them (fully realize that Mr Gut can be the most fallible being on Earth) - in other words, I just don't want to spend the time on something I'm going to be mostly wrong on anyway. If I want to look at projections, I come here - more for the macro view than for any individual player - as the staff can do at least as good a job as I can.That being said.....How do the non-projection guys decide between a RB, QB, or WR when all are at the top of their respective tiers (assuming "need" is equal)?

My league is heavily weighted towards RB's, so I would probably lean towards the RB in that case. But it would really depend on who the guys are at the other positions and how strongly I felt about them.

 

[Uruk-Hai]

I've become less a "do you own projections for every player" guy as the years have passed. Part of that is laziness & part is the fallibility of them (fully realize that Mr Gut can be the most fallible being on Earth) - in other words, I just don't want to spend the time on something I'm going to be mostly wrong on anyway. If I want to look at projections, I come here - more for the macro view than for any individual player - as the staff can do at least as good a job as I can.That being said.....How do the non-projection guys decide between a RB, QB, or WR when all are at the top of their respective tiers (assuming "need" is equal)?

My league is heavily weighted towards RB's, so I would probably lean towards the RB in that case. But it would really depend on who the guys are at the other positions and how strongly I felt about them.

As I said, let's assume need is equal & you perceive the value of each to be in the same general ballpark. With projections (filtered thru VBD), someone can use math to determine his next pick (I'm not advocating that, btw, but I know it's done). But if you're not using hard #s, what do you do when all things are pretty even?

 

[bicycle_seat_sniffer]

in regards to your question about 2 people running an offense and has it happened before, the answer is yes:

Sean Payton/Parcells ( Dallas)

Joe Gibbs/Al Saunders

Holmgren/Haskell

among other coaching staffs...

I'm quite sure Bellichick uses the same approach with his coaching staff, even when he had Charlie Weiss in town..some coaches are very hands-on like that..

most predictions involve fuzzy math ( Dodds) , while other 'experts' are homers pimping the players from their favorite teams...most fantasy mags and websites use generic , cookie-cutter `me-too' lists, i.e., ever notice that nearly every magazine has an identical top RB/WR/QB. I'd bet the farm that Dodds' top 50 list ( overall) would be nearly player-for-player with Ian Allan's ( fantasy index) list, or Mike Fabiano's ( NFL.com) list, or Mike Nazarek's..

How many people didn't list Witten in the TE top 10 last season ?? they were all wrong!

many of the magzines,especially the ones out in late June/early July, are watered down fluff pieces and they're all the same..The publishers never seem to go out on a limb and make a bold prediction, simply because they want to keep people coming back year after year, which won't happen if they make a lot of predictions that fall short..we all know the unwritten rule,that 50% of the top 10 players from a given position at year's end, will not finish in the top 10 at the end of the following season..

not one of the fantasy mags that I've EVER purchased, takes that into account..Not one 2007 mag had Adrian Peterson listed as their #1 RB last July..Not one magazine had David Garrard ranked in the top 10, or Witten, or Derrick Mason, or Welker..Many mags didn't even have Moss ranked in the WR top 10 in July 2007..

anyone can churn out a WR list with Reg Wayne,CJ, T.O..and I know injuries do happen..but these 'experts' just plain got it wrong with Moss, Welker + Co.

but at a price of $8.99, these publishers need to come up with something new, I can't imagine that people will continue to buy such mags if the continue to churn out fluff piece articles and stats, the same things I can get FOR FREE , online...

I don't mind paying for one mag per year, but thats it..the rest are cookie cutter 'me too' publications..

take a look at FBG's top 20 RB list and tell me it doesn't look like Fantasy index, or Football Forecast, or FantasyGuru... it probably looks nearly identical to their 2007 edition...same guys in the top 10 at every position.

You might want to do some fact-checking next time.FBG had Witten as TE7 last year.

not only that in the perfect draft article this was the TE that he told everyone to target after the big boys were off the board

 

[rzrback77]

This has been a lively discussion, but seems to keep going in circles. No projections are perfect, but they are helpful. The spotlights are some of my favorite threads because people don't just spout numbers, they provide their reasons for why they project the way they do. The more information taken into considation for the projections the better they tend to be. But, you still don't draft just by the specific projections. The best drafters key on ADP and take players no earlier than they have to. And they take the guys they "feel" are best suited to outperform thier ADP or projections. And they grab them just before their opponents do.

All of these differing opinions and methhods are what make the draft and this hobby so much fun, whether you are matching wits with your friends in a local league or with fellow fantasy players across the innernets.

 

[texasheat]

That being said.....How do the non-projection guys decide between a RB, QB, or WR when all are at the top of their respective tiers (assuming "need" is equal)?

I just go with my gut....

 

[ceo3west]

I've become less a "do you own projections for every player" guy as the years have passed. Part of that is laziness & part is the fallibility of them (fully realize that Mr Gut can be the most fallible being on Earth) - in other words, I just don't want to spend the time on something I'm going to be mostly wrong on anyway. If I want to look at projections, I come here - more for the macro view than for any individual player - as the staff can do at least as good a job as I can.

That being said.....

How do the non-projection guys decide between a RB, QB, or WR when all are at the top of their respective tiers (assuming "need" is equal)?

My league is heavily weighted towards RB's, so I would probably lean towards the RB in that case. But it would really depend on who the guys are at the other positions and how strongly I felt about them.

As I said, let's assume need is equal & you perceive the value of each to be in the same general ballpark. With projections (filtered thru VBD), someone can use math to determine his next pick (I'm not advocating that, btw, but I know it's done). But if you're not using hard #s, what do you do when all things are pretty even?

FWIW, you're original question mentioned nothing about value, only that the need is equal, which is why I would have chosen the RB. Assuming now that the value of these players is equal, I would choose the one I feel has the best chance of success this year based on gut feeling with other circumstances surrounding that player also being considered (team he plays for, playoff schedule, etc.)

 

[Dirty Weasel]

When it comes down to it, every FF'er does projections, whether they put it on paper or not. I will exaggerate to make my point, but think of it this way - if I ask you if Tomlinson will get at least 100 carries, you will say yes, right? How about 200 carries? It's still yes, right? What about 225 carries? Eventually I will hit a number you will say no to. Repeat the same questions for rushing yards, receptions, receiving yards and TD's and there you have it. You have just made your projections. As I said, you don't have to put it on paper for it to be a projection.

 

[rzrback77]

When it comes down to it, every FF'er does projections, whether they put it on paper or not. I will exaggerate to make my point, but think of it this way - if I ask you if Tomlinson will get at least 100 carries, you will say yes, right? How about 200 carries? It's still yes, right? What about 225 carries? Eventually I will hit a number you will say no to. Repeat the same questions for rushing yards, receptions, receiving yards and TD's and there you have it. You have just made your projections. As I said, you don't have to put it on paper for it to be a projection.

:( From the gut = projections not written down

 

[FantasyTrader]

I read the whole thread, great discussion. But I think Chairshot's post above ended the thread if you're willing to step back and think logically. The idea that projections are useless because they're always wrong is just...wrong.

The lie that's perpetuated over and over again is that the league champ was the guy that was right. It's just not true. The winner is just the person that was LEAST WRONG.

 

[Ministry of Pain]

I read the whole thread, great discussion. But I think Chairshot's post above ended the thread if you're willing to step back and think logically. The idea that projections are useless because they're always wrong is just...wrong. The lie that's perpetuated over and over again is that the league champ was the guy that was right. It's just not true. The winner is just the person that was LEAST WRONG.

Projections are probably right on players like LT, TO, and folks like that...but a lot of the projections that websites put together are not right...and i think I got a little wound up when I saw sights bragging about waht they did and also some site scored a 100...But this has been some very spirited debate on both sides of the aisle and I think both sides deserve credit and also they have both good and bad merits with both.

 

[Maurile Tremblay]

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right?

The answer is no.If Julius Jones has an easier schedule this year than LaDainian Tomlinson, would SOS leave you drafting Jones every time? Obviously not, right? There are other factors to consider as well.

It's the same with projections. Projections wouldn't leave me drafting Forte over Stewart any more than SOS would leave me drafting Julius Jones over Tomlinson. That's not a knock against the value of either projections or SOS.

 

[Pictus Cat]

Nice thread.

I wonder, if I tiered A, B, C, D, E, F. Then sub categoried -, +, --, ++...and so on, until I had a top 200...what is that?

A projection to me differs from that because it gives me an understanding of how the player is going ot do it.

That's important because I play in some complicated scoring systems. I create my leagues that way.

The DD is great for adjusting to new scoring systems and letting me tweak the projections where my opinion varies.

The PD is great for allowing me to start with projections feeding the DD that are close to my thoughts by emphasizing the staff I agree with most.

The hover explanations and Staff articles and posting help me understand the staff that make projections and have an idea how to incorperate staff that do not make projections, but rankings are more similar with my thoughts (gut.)

My thought gut comes from all the other information I gather, some of the most important to me are staff/expert opinions, articles, player rightups, latest news, initial scout reports, camp writeups....

Once the projections can give me fantasy points for the scoring system,

tiering can begin.

Then the ADP is looked at. I'm not interested in who's drafting who in meaningless mocks in unrated scoring and roster leagues. I attempt to find ADP that will be close to my league to fall back on AND tweak it with my own predictions of who will be taken. The average magazine ranking and DD's ability to assign draft tendancies has worked well for me as a fallback.

Armed with my tiers(adjusted for scoring with projections) and ADP draft predictions,

I'll make my pick,

with my gut (everything I know about the players in the tier, including floor/ceiling varience...would like a little more mathmatical FBG opinion on that)

considering what kind of overall team I can draft (given my predictions of who will be gone in the draft for my next picks .)

and my teams needs

At the end of the draft I have two groups of players, ones that should make their ADP and "my horses" I'm betting on to exceed their ADP. The most important step after that: know when to shoot the horse and pick a new one throughout the season.

Does it work? Who cares. I just think it's fun.

...and I go lucky and won half my leagues last year.

...and I think I'm getting better.

 

[thayman]

I didn't read all the thread, I also dislike projections this time of year. However there is nothing else to do Fantasy Football related so why not? It brings up some great talks and debates on various players, homers chime in on the various player spotlight threads giving and changing my opinion on different players.

Projections spark some good discussion.

 

[The Keyboard of Reason]

Every year for the past 5 years I have tried to scream from every available bullhorn, in every preseason ranking thread I start, in every possible way I know of I have tried to voice my displeasure with the entire projection system that permeates these boards. I even see websites out their boasting about how great they were at the projections. I would really have to see what their summer projections were and what they ended with to see how some of these sites are scoring 100. I do not believe for a second that they got all the stats right down to the yds and TD for all the players. It is impossible.

There are so many holes in the FBG projections, and I'm not picking on them but I pay for a membership every year and this is where I do the majority of my bantering so I'll use them as an example. For instance, Warrick Dunn had his worst year as a RB last year playing on a sinking ship team in the Atlanta Flacons and he still finished 30th. He has been between 15-24 for most of his career. A very undervalued RB2 for the past 10 years in this league.

I'll take this a step further...when he was in TB he had reception totals of 44, 68, 44, 54 and then 50 his 1st season in Atlanta...then I don't know if it was Vick, the HC Mora, or what happened but he became invisible in the passing department where his totals slipped under 30 a year. Now he signs a $3 million deal with the Bucs who did not sign him to simply watch him walk around the sidelines as some websites suggest I assure you. What will his totals be? I haven't the foggiest of clues but I do know that I would bet money he finishes well ahead of the ranking he has at 45 this year, and likely will far outperform his draft spot of somewhere in the 12th round or later. I can't put everything I am talking about into his stats but I do think thru some cognitive thinking that one should easily derive that Dunn = value this year, tons of it I might add. What track record does Earnest Graham bring to the table? Dunn could easily find himself getting 12-15 carries and 3-5 catches a game quite a bit this season.

So let's land the plane a bit and focus on what I am talking about becasue this deserves some serious debate. I understand it is easy to project out for guys like LT, TO, and the like because they are so GD consistent you can't be off by more than 10% one way or the other, fantastic. But then I see people just loading in Roddy White's stats from a year ago and I want to scream WTF are you thinking? Projections are very dangerous to put on players that have no track record or history to really go on. And if you think the same top20 at WR last season is going to be the same top20 WR this season, you are delusional and you are going to miss the playoffs this season because you won't get your head out of your you know what. Think people, think!

Put the stats down and try and think logically about what you are going to do on draft day. What do you do, you say? Let me try and steer you into some different thinking. Get a firm grip on the OL for all 32 teams, rank em, file em, put em down on paper. Use all resources you can to determine strength on these very important pieces of the puzzle. Don't fall for what everyone is writing if it goes against what you believe. For example, I am not sold on the Bucs and Raiders OL the way others seem to be. I could be way off but I am not ready to say these OL are good-great just yet.

Next you need to seriously think about the coaching staff and what/whether they like to run/pass. Offensive Philosophy is so critical. There is actually 2 people running the OC position in Detroit...did you know that? Has this ever worked before? And they have come out and said flat out they are not going to pass the way Martz did...they are sending out the "Conservative Approach" if you listen to anything they have to say. What does that do to Roy Williams who some in here think is bound for 1,300 yds and 10 TD...what about Calvin Johnson and his progression in year 2? I don't have all the answers but these are the things you need to think about. Not just crunching out Roy Williams numbers if and when he plays a full 16 games.

So you got the OL, coaching philosophy down, next I would look at surrounding talent in the skill positions especially QB. Next I would look at the schedule and see if you can find a soft schedule and also who has the harder schedules as it will matter. And finally I might look at the intangibles. LT for example simply has intangibles that work in his favor. He could have no OL, QB, a pass happy OC, and a tough schedule, he would still finish top5...you catch what I'm saying here?

Then I would take these categories...and you can make up others but I try to limit it to around 5, then divide up the 5 categories and give them different weighted averages, sort of like the grades you got in school. Then I look and see what I have, put them into tiers based on what I uncover with a nonprojection system, then I figure out where I have to draft which player in order to get who I want.

I'll hang up and listen.

A couple of pointsunn: I think projecting points and possibly even finish may not lead to the value he brings to a fantasy team. The truth is that a guy who puts up 7 points a game will finish with 112 points and finish about 30th. But what kind of value does that really bring to a fantasy team? I would much rather have a guy who finished below Dunn but when he played he was more productive. D. Ward didn't have 100 points for the year but he would be way more valuable to a team than a guy like Dunn. Replacement value is more important than total points.

I would caution against predicting defenses too much at this early stage. Every year we see a change in team defenses and while a very solid D will not get torched, it is not prudent to base a pick on the teams defensive schedule as we see in hindsight they are not very accurate.

While you can't simply load last year stats into the data base for next year (your Roddy White example), looking FOR track record will make you a hindsight owner instead of setting the trend. Sure it is best to see a guy who had an off year that people have soured on but you know there were extenuating circumstances and he will do better if healthy this year, but if you can see "greatness" in a player before others this is how you really jump ahead.

I am not really sure what you are displeased about or would like to see, but I do agree with the other factors you take into account

Very

 

[The Keyboard of Reason]

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right?

The answer is no.If Julius Jones has an easier schedule this year than LaDainian Tomlinson, would SOS leave you drafting Jones every time? Obviously not, right? There are other factors to consider as well.

It's the same with projections. Projections wouldn't leave me drafting Forte over Stewart any more than SOS would leave me drafting Julius Jones over Tomlinson. That's not a knock against the value of either projections or SOS.

One of the problems with SOS at this time of year is that some minor differences at where the defense is ranked can really change the way you "would" have drafted. I am not a big fan of using SOS for drafting, unless, I am breaking a tie (which is rare)

 

[Ministry of Pain]

Every year for the past 5 years I have tried to scream from every available bullhorn, in every preseason ranking thread I start, in every possible way I know of I have tried to voice my displeasure with the entire projection system that permeates these boards. I even see websites out their boasting about how great they were at the projections. I would really have to see what their summer projections were and what they ended with to see how some of these sites are scoring 100. I do not believe for a second that they got all the stats right down to the yds and TD for all the players. It is impossible.

There are so many holes in the FBG projections, and I'm not picking on them but I pay for a membership every year and this is where I do the majority of my bantering so I'll use them as an example. For instance, Warrick Dunn had his worst year as a RB last year playing on a sinking ship team in the Atlanta Flacons and he still finished 30th. He has been between 15-24 for most of his career. A very undervalued RB2 for the past 10 years in this league.

I'll take this a step further...when he was in TB he had reception totals of 44, 68, 44, 54 and then 50 his 1st season in Atlanta...then I don't know if it was Vick, the HC Mora, or what happened but he became invisible in the passing department where his totals slipped under 30 a year. Now he signs a $3 million deal with the Bucs who did not sign him to simply watch him walk around the sidelines as some websites suggest I assure you. What will his totals be? I haven't the foggiest of clues but I do know that I would bet money he finishes well ahead of the ranking he has at 45 this year, and likely will far outperform his draft spot of somewhere in the 12th round or later. I can't put everything I am talking about into his stats but I do think thru some cognitive thinking that one should easily derive that Dunn = value this year, tons of it I might add. What track record does Earnest Graham bring to the table? Dunn could easily find himself getting 12-15 carries and 3-5 catches a game quite a bit this season.

So let's land the plane a bit and focus on what I am talking about becasue this deserves some serious debate. I understand it is easy to project out for guys like LT, TO, and the like because they are so GD consistent you can't be off by more than 10% one way or the other, fantastic. But then I see people just loading in Roddy White's stats from a year ago and I want to scream WTF are you thinking? Projections are very dangerous to put on players that have no track record or history to really go on. And if you think the same top20 at WR last season is going to be the same top20 WR this season, you are delusional and you are going to miss the playoffs this season because you won't get your head out of your you know what. Think people, think!

Put the stats down and try and think logically about what you are going to do on draft day. What do you do, you say? Let me try and steer you into some different thinking. Get a firm grip on the OL for all 32 teams, rank em, file em, put em down on paper. Use all resources you can to determine strength on these very important pieces of the puzzle. Don't fall for what everyone is writing if it goes against what you believe. For example, I am not sold on the Bucs and Raiders OL the way others seem to be. I could be way off but I am not ready to say these OL are good-great just yet.

Next you need to seriously think about the coaching staff and what/whether they like to run/pass. Offensive Philosophy is so critical. There is actually 2 people running the OC position in Detroit...did you know that? Has this ever worked before? And they have come out and said flat out they are not going to pass the way Martz did...they are sending out the "Conservative Approach" if you listen to anything they have to say. What does that do to Roy Williams who some in here think is bound for 1,300 yds and 10 TD...what about Calvin Johnson and his progression in year 2? I don't have all the answers but these are the things you need to think about. Not just crunching out Roy Williams numbers if and when he plays a full 16 games.

So you got the OL, coaching philosophy down, next I would look at surrounding talent in the skill positions especially QB. Next I would look at the schedule and see if you can find a soft schedule and also who has the harder schedules as it will matter. And finally I might look at the intangibles. LT for example simply has intangibles that work in his favor. He could have no OL, QB, a pass happy OC, and a tough schedule, he would still finish top5...you catch what I'm saying here?

Then I would take these categories...and you can make up others but I try to limit it to around 5, then divide up the 5 categories and give them different weighted averages, sort of like the grades you got in school. Then I look and see what I have, put them into tiers based on what I uncover with a nonprojection system, then I figure out where I have to draft which player in order to get who I want.

I'll hang up and listen.

A couple of pointsunn: I think projecting points and possibly even finish may not lead to the value he brings to a fantasy team. The truth is that a guy who puts up 7 points a game will finish with 112 points and finish about 30th. But what kind of value does that really bring to a fantasy team? I would much rather have a guy who finished below Dunn but when he played he was more productive. D. Ward didn't have 100 points for the year but he would be way more valuable to a team than a guy like Dunn. Replacement value is more important than total points.

I would caution against predicting defenses too much at this early stage. Every year we see a change in team defenses and while a very solid D will not get torched, it is not prudent to base a pick on the teams defensive schedule as we see in hindsight they are not very accurate.

While you can't simply load last year stats into the data base for next year (your Roddy White example), looking FOR track record will make you a hindsight owner instead of setting the trend. Sure it is best to see a guy who had an off year that people have soured on but you know there were extenuating circumstances and he will do better if healthy this year, but if you can see "greatness" in a player before others this is how you really jump ahead.

I am not really sure what you are displeased about or would like to see, but I do agree with the other factors you take into account

Wholeheartedly disagree. You can go back and look on NFL.com/Stats or pro-football-reference.com and you will see that a lot of the same teams pop up year in and year out in the top10 and bottom 10. For example, the Lions are in the bottom 5-10 of pass defenses almost year in and year out...Baltimore, Pitt and Minnesota have finished in the top3 in rush defenses for the last 2 years in a row, and I think most see the Vikings to continue to dominate agianst the run with their pillars at DT. You're correct that it is not 100% foolproof but that is where you look and see if there have been any upgrades or if you see rookies that should improve in year 2...but actually predicting defenses is not a total crapshoot at all. We know the AFC North is going to have a tough schedule for example.

 

[Dinsy Ejotuz]

Projections are probably right on players like LT, TO, and folks like that...but a lot of the projections that websites put together are not right...and i think I got a little wound up when I saw sights bragging about waht they did and also some site scored a 100...

I don't disagree with you at all here re: projections. But that '100' score you're talking about wasn't a perfect score. They just took the best performing set of projections and assigned it a score of 100. Then based all the other scores off that one, using it as an anchor.

 

[Ministry of Pain]

Projections are probably right on players like LT, TO, and folks like that...but a lot of the projections that websites put together are not right...and i think I got a little wound up when I saw sights bragging about waht they did and also some site scored a 100...

I don't disagree with you at all here re: projections. But that '100' score you're talking about wasn't a perfect score. They just took the best performing set of projections and assigned it a score of 100. Then based all the other scores off that one, using it as an anchor.

Glad you mentioned that...but I also will make this point...sometimes the preseason mags will score the different positions at like 6.5, 8.0, or 7.5...there never seems to be a perfect score of 10.0...not sure why but they don't ever use the "anchor" as you call it and base everything off of that. Thanks for pointing that out though, I saw the 100 score and jumped to conclusions.

 

[gheemony]

I think I understand what they're getting at. I know VERY SUCCESSFUL high stakes players that have NEVER projected stats.

ADP is their main warrior and they go off gut feel from there. Statistical projections are helpful, but I find a system of balancing upside and downside to be useful as well.

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right? I mean, your best guess at what they'll produce is what you should go with right? Not so fast... I'll take STEWART all day long due to the upside potential. Why is Forte's upside so low? The offense in Chicago... Why is Stewart's upside so high? The offense in Carolina, and countless other scenarios and possibilities that affect your upside / downside rankings. Maybe it's a backup with an injury prone starter, or maybe its a player who hasn't proven it, but a history of that offensive system shows a big time number potential (Broncos).

Simple projections are good for those wanting a cheatsheet and a quick reference before draft day, but for those of us more serious about it, I think MOP and Gatorman are on to something. Maybe more information like above would help.

This is why I think all FBG projections should come with a low, high, and mean projection. And the projections dominator should let me print out the low, high, and mean projection for the 5 projectoins provided. I want to know the range that the projections might fall in.

 

[Ministry of Pain]

I think I understand what they're getting at. I know VERY SUCCESSFUL high stakes players that have NEVER projected stats.

ADP is their main warrior and they go off gut feel from there. Statistical projections are helpful, but I find a system of balancing upside and downside to be useful as well.

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right? I mean, your best guess at what they'll produce is what you should go with right? Not so fast... I'll take STEWART all day long due to the upside potential. Why is Forte's upside so low? The offense in Chicago... Why is Stewart's upside so high? The offense in Carolina, and countless other scenarios and possibilities that affect your upside / downside rankings. Maybe it's a backup with an injury prone starter, or maybe its a player who hasn't proven it, but a history of that offensive system shows a big time number potential (Broncos).

Simple projections are good for those wanting a cheatsheet and a quick reference before draft day, but for those of us more serious about it, I think MOP and Gatorman are on to something. Maybe more information like above would help.

This is why I think all FBG projections should come with a low, high, and mean projection. And the projections dominator should let me print out the low, high, and mean projection for the 5 projectoins provided. I want to know the range that the projections might fall in.

It would be cool to have a high side/low side...however FBG does do best case scenario/worst case scenario. They do an in depth player analysis on almost every player that will be drafted. They do a spotlight on almost everyone as well. I think when you add it all up you can see why they are making certain projections.

 

[Pictus Cat]

I think I understand what they're getting at. I know VERY SUCCESSFUL high stakes players that have NEVER projected stats.

ADP is their main warrior and they go off gut feel from there. Statistical projections are helpful, but I find a system of balancing upside and downside to be useful as well.

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right? I mean, your best guess at what they'll produce is what you should go with right? Not so fast... I'll take STEWART all day long due to the upside potential. Why is Forte's upside so low? The offense in Chicago... Why is Stewart's upside so high? The offense in Carolina, and countless other scenarios and possibilities that affect your upside / downside rankings. Maybe it's a backup with an injury prone starter, or maybe its a player who hasn't proven it, but a history of that offensive system shows a big time number potential (Broncos).

Simple projections are good for those wanting a cheatsheet and a quick reference before draft day, but for those of us more serious about it, I think MOP and Gatorman are on to something. Maybe more information like above would help.

This is why I think all FBG projections should come with a low, high, and mean projection. And the projections dominator should let me print out the low, high, and mean projection for the 5 projectoins provided. I want to know the range that the projections might fall in.

Like it. I would prefer columns in the dominator for fantasy points. Projection, Ceiling, Floor FP.Maybe a replacement article on potential fantasy points (Standard & PPR) for backups if they became starters.

 

[gheemony]

I think I understand what they're getting at. I know VERY SUCCESSFUL high stakes players that have NEVER projected stats.

ADP is their main warrior and they go off gut feel from there. Statistical projections are helpful, but I find a system of balancing upside and downside to be useful as well.

For example: (not a real comparison)

(assuming health is not an issue)

Let's say Jonathan Stewart, has an upside of 250 points and a downside of 80 but is projected at 130.

Let's say Matt Forte, has an upside of 180 points and a downside of 110 but is projected at 150.

Projections will leave you drafting MATT FORTE every time right? I mean, your best guess at what they'll produce is what you should go with right? Not so fast... I'll take STEWART all day long due to the upside potential. Why is Forte's upside so low? The offense in Chicago... Why is Stewart's upside so high? The offense in Carolina, and countless other scenarios and possibilities that affect your upside / downside rankings. Maybe it's a backup with an injury prone starter, or maybe its a player who hasn't proven it, but a history of that offensive system shows a big time number potential (Broncos).

Simple projections are good for those wanting a cheatsheet and a quick reference before draft day, but for those of us more serious about it, I think MOP and Gatorman are on to something. Maybe more information like above would help.

This is why I think all FBG projections should come with a low, high, and mean projection. And the projections dominator should let me print out the low, high, and mean projection for the 5 projectoins provided. I want to know the range that the projections might fall in.

It would be cool to have a high side/low side...however FBG does do best case scenario/worst case scenario. They do an in depth player analysis on almost every player that will be drafted. They do a spotlight on almost everyone as well. I think when you add it all up you can see why they are making certain projections.

Yes, but it is not all in one place. I want it available in PD and DD. I'm a numbers guy and would love to see the ranges with the projections.Also, it would be great to see more explanations with projections. The projections threads were a good start as I now understand how Henry, Tremblay, etc. come up with their projections (and it's not a dart board). But that needs to be built in to one app.