Projection Theory Poll #3
Projection Theory Poll #3
Let's assume for a minute we project two players to hit .280, with 20 HR and 80 RBI.
Player #1 is a 12 year veteran with pretty stable statistical sets- his career has generally seen him move anywhere between .265-.290 and 18-22 HR, 70-90 RBI. Not much chance for upside here, but the projection is solid
Player #2 is a rookie who had 45 PA in a cup of coffee in late September. His MLE's support skills that right now represent the projection above, but he hasn't done much in the majors, and he's only had one full season above A ball.
Which poll response best fits your feelings here?
Player #1 is a 12 year veteran with pretty stable statistical sets- his career has generally seen him move anywhere between .265-.290 and 18-22 HR, 70-90 RBI. Not much chance for upside here, but the projection is solid
Player #2 is a rookie who had 45 PA in a cup of coffee in late September. His MLE's support skills that right now represent the projection above, but he hasn't done much in the majors, and he's only had one full season above A ball.
Which poll response best fits your feelings here?
Re: Projection Theory Poll #3
1st player sounds like Adrien Beltre or little less Raul Ibanez
Second player sounds like Travis Snyder...
but I select option # 2
Second player sounds like Travis Snyder...
but I select option # 2
Re: Projection Theory Poll #3
I prefer to assess risk during the draft.
I'll decide if I want Chris Davis or Aramis Ramirez, Adam Jones or Mike Cameron, etc.
I'll decide if I want Chris Davis or Aramis Ramirez, Adam Jones or Mike Cameron, etc.
Re: Projection Theory Poll #3
I voted #2. I tend to discount rookie projections a bit - err on the side of caution. If the projections are based on MLE's, this is an undemonstrated MLB skill set. The veteran is more likely to maintain proven production than the rookie is to provide unproven production.
- alleyoops
- Major League All-Star
- Posts: 424
- Joined: January 3rd, 2009, 8:22 pm
- Preferred Style: 5x5 slow auctions
- Location: La Quinta, CA
- Contact:
Re: Projection Theory Poll #3
I voted #2, but I really feel #5 (which isn't listed). The projections are valid, but there should be some indicator along with the projection to tell me that the first projection is more "solid" than the second one.
If this was a stock, I'd want a volatility indicator. That is, the first one is not very volatile - he'll likely be 10-20% within the projection. The second one is very volatile - could be a total bust, or could exceed the projections by a significant amount. Of course, if this was a stock, they'd both be going down big time in today's market.
If this was a stock, I'd want a volatility indicator. That is, the first one is not very volatile - he'll likely be 10-20% within the projection. The second one is very volatile - could be a total bust, or could exceed the projections by a significant amount. Of course, if this was a stock, they'd both be going down big time in today's market.
-
- Hall of Famer
- Posts: 1574
- Joined: January 3rd, 2009, 12:09 am
Re: Projection Theory Poll #3
I selected option #2. I stated on the old boards that I do not put a lot of weight on September ABs for call ups especially if they were against teams no longer playing for anything and he's on a team that wasn't playing for anything.
These are my views based on my own opinions and observations - your mileage may vary.
"KNOW THY LEAGUE" - the Forum Funklord - 4/13/2009
Fantasy is managing stats ... roto is managing teams
"KNOW THY LEAGUE" - the Forum Funklord - 4/13/2009
Fantasy is managing stats ... roto is managing teams
Re: Projection Theory Poll #3
As I post this I am still the only person to chose option 1. Personally, I'd like to think that this risk is already built into the projection. The 2 biggest components of a projection are 1. that players skill level and 2. his opportunity to use those skills (playing time). If the projectors are projecting the same stat line from 2 very different players I personally would hope that these 2 biggest factors have already been weighed carefully.
Though ultimately I suppose it doesn't matter so much if the projections are based on poll option 1 or poll option 2 as long as you know whether or not this bias is already built in. If I know its not then I can impart that bias myself. But, if I know it has been then I would have to feel very strongly 1 way or the other to impart additional bias on top of the projections built in bias.
wow. as i posted this someone else voted 1. weird.
Though ultimately I suppose it doesn't matter so much if the projections are based on poll option 1 or poll option 2 as long as you know whether or not this bias is already built in. If I know its not then I can impart that bias myself. But, if I know it has been then I would have to feel very strongly 1 way or the other to impart additional bias on top of the projections built in bias.
wow. as i posted this someone else voted 1. weird.
Re: Projection Theory Poll #3
That would be ME on #1.....first (sorry, I dont mean this negative) but this theory isn't practical at all, There are more factors here, I chose option one, because I feel most projection should take the basic stuff into account...such as team the player is on, hitters park or not, how many AB the player has...did he lose or gain based on other players in the off season(such as who is batting before him and after him) You see a big difference based on players on where they bat.....like Jeff Kent, he got pitched to A TON because of Bonds, sure he never liked to admit it, but its true. Phillies lineup in a great example......oh yea, when I do my own picks and the projections seem dead even, it goes to past knowledge of players....such as Randy Johnson, he is ALWAYS hurt, he is old.....Tulo...he burned me as a keeper last year....he's worth about willy mo pena in my book still......Most of us have players we can't stand to own on our team becaus of something like this.....oh TIP time...find out in your draft if people hate certain players and trade for them or snag them cheaper in an auction draft....sometimes you get some steals.......
Re: Projection Theory Poll #3
Some work is required from the user of the projections whether a stability/risk factor is included in the projections or not.
If a risk/stability factor is integrated into the projections, a player like Chris Davis will project lower. He has a half season in the Majors. How much stability can there be in the projection? I'll call it "The Kevin Maas Factor". Whether you modify the projections on the front end or not, the user still has to decide on the individual player. If the projection is lower due to risk, I have to move Davis up my draft list if I'm a believer. If the projection does not take risk into account, I have to move him down my draft list if I'm not a believer.
I prefer the projections merely reflect current measured skill set and projected AB.
I don't need it, but listing a players age or number of seasons could be added to give an indication of stability/risk.
If a risk/stability factor is integrated into the projections, a player like Chris Davis will project lower. He has a half season in the Majors. How much stability can there be in the projection? I'll call it "The Kevin Maas Factor". Whether you modify the projections on the front end or not, the user still has to decide on the individual player. If the projection is lower due to risk, I have to move Davis up my draft list if I'm a believer. If the projection does not take risk into account, I have to move him down my draft list if I'm not a believer.
I prefer the projections merely reflect current measured skill set and projected AB.
I don't need it, but listing a players age or number of seasons could be added to give an indication of stability/risk.
Re: Projection Theory Poll #3
i would like to know what round we are in...obviously i wouldnt take the kid in the early rounds for the risk factor and blowing a high pick on him if he does fail... & visa versa with the elder statesman.
- alleyoops
- Major League All-Star
- Posts: 424
- Joined: January 3rd, 2009, 8:22 pm
- Preferred Style: 5x5 slow auctions
- Location: La Quinta, CA
- Contact:
Re: Projection Theory Poll #3
I'd like the projection numbers themselves to be the site guru's best guess as to what the player will actually produce this year, based on everything he knows at the time. As things change, the projections should change.
When making that guess, he goes through some process of examining a lot of factors (I hope), and going through this process should allow him to have some kind of idea of whether those numbers he's putting out are "solid", or whether they're "iffy". In other words, how confident he is that the player will actually put up numbers within a relatively narrow range of the projected numbers.
Admittedly this measure of confidence would be subjective, but so are the projections themselves, to some extent. I follow the news and players quite a bit, but I'm sure I wouldn't be able to assign this kind of confidence level to every player - just not enough time. But the guru is already spending that time. Having that subjective factor would be of use to me, in alerting me to the difference in these two players having the same projections.
Todd's already talked about this in an earlier thread. I'm bringing it up again here because it fits the example so well.
When making that guess, he goes through some process of examining a lot of factors (I hope), and going through this process should allow him to have some kind of idea of whether those numbers he's putting out are "solid", or whether they're "iffy". In other words, how confident he is that the player will actually put up numbers within a relatively narrow range of the projected numbers.
Admittedly this measure of confidence would be subjective, but so are the projections themselves, to some extent. I follow the news and players quite a bit, but I'm sure I wouldn't be able to assign this kind of confidence level to every player - just not enough time. But the guru is already spending that time. Having that subjective factor would be of use to me, in alerting me to the difference in these two players having the same projections.
Todd's already talked about this in an earlier thread. I'm bringing it up again here because it fits the example so well.
- viper
- Hall of Famer
- Posts: 1476
- Joined: December 31st, 2008, 11:32 pm
- Preferred Style: Currently in an AL-only league with the Bill James Technical RCA as the single hitting category and ERA as the single pitching category.
- Contact:
Re: Projection Theory Poll #3
One thing is that I want them to be projections and not predictions. I consider there is a difference. We take a site's projections and then quite often modify them in our minds. These projections then become more like predictions.
The avalanche has started. It is too late for the pebbles to vote. -- Ambassador Kosh
Mike Ladd
Buffy, the Umpire Slayer
Mike Ladd
Buffy, the Umpire Slayer
Re: Projection Theory Poll #3
I'm glad the conversation here went as it did. I'd especially like to thank the two who didn't answer with the rest, if only to stimulate the conversation.
You may be interested to know that I am probably a #1 and a #4 answerer from a theoretical standpoint, and a #2 from a practical standpoint.
I include #4 with #1 because in a perfect world, you could calculate the amount of discount a player should take to their projected skill set in order to properly adjust for risk. So in a way we do cop out by not doing it - because I can't think of a method that does it fairly.
So in our example, I cut the rookie down to .270-15-70 (picking a discount randomly), and now I'm comparing him to other .270-15-70 guys, and the user still has to mentally make the adjustment that from a raw skills standpoint the risky guy is "better" than the safe guy. I suppose it depends on your thought process but you're still making mental adjustments on the fly - there's no way out of it. Just depends on which adjustment you want to make.
Practically speaking, I'd rather see skills on the board as Im drafting than have to mentally divide backwards to include/uninclude risk from a projection as my needs change.
Lastly,
But in many of these cases you're just adding bias to a situation without a lot of hard evidence - and yes you're paying us to make gut calls and we make them all the time, but many of those gut calls are to do nothing and let the data be as it is.
You may be interested to know that I am probably a #1 and a #4 answerer from a theoretical standpoint, and a #2 from a practical standpoint.
I include #4 with #1 because in a perfect world, you could calculate the amount of discount a player should take to their projected skill set in order to properly adjust for risk. So in a way we do cop out by not doing it - because I can't think of a method that does it fairly.
So in our example, I cut the rookie down to .270-15-70 (picking a discount randomly), and now I'm comparing him to other .270-15-70 guys, and the user still has to mentally make the adjustment that from a raw skills standpoint the risky guy is "better" than the safe guy. I suppose it depends on your thought process but you're still making mental adjustments on the fly - there's no way out of it. Just depends on which adjustment you want to make.
Practically speaking, I'd rather see skills on the board as Im drafting than have to mentally divide backwards to include/uninclude risk from a projection as my needs change.
Lastly,
The practical question you may or may not see is that in a player with established history on a team (think late 90's Yanks), the adjustments you desire are already incorporated - but in a new situation of a player changing teams, how much manipulation do you want us making? Put another way, do you want to pay for stats that aren't based on history but more based on a hunch? My answer is if I'm 95% comfortable with something, I'd adjust it. For example, if a mid 1980's Wade Boggs went from hitting leadoff for the Red Sox and had been dealt to the 1985 Cardinals for Tommie Herr, where he would instead hit 2nd or 3rd behind Vince Coleman and Willie McGee, yes, I'd adjust RBI totals upwards.That would be ME on #1.....first (sorry, I dont mean this negative) but this theory isn't practical at all, There are more factors here, I chose option one, because I feel most projection should take the basic stuff into account...such as team the player is on, hitters park or not, how many AB the player has...did he lose or gain based on other players in the off season(such as who is batting before him and after him) You see a big difference based on players on where they bat.....like Jeff Kent, he got pitched to A TON because of Bonds, sure he never liked to admit it, but its true. Phillies lineup in a great example......
But in many of these cases you're just adding bias to a situation without a lot of hard evidence - and yes you're paying us to make gut calls and we make them all the time, but many of those gut calls are to do nothing and let the data be as it is.
Re: Projection Theory Poll #3
GaryJ wrote: The practical question you may or may not see is that in a player with established history on a team (think late 90's Yanks), the adjustments you desire are already incorporated - but in a new situation of a player changing teams, how much manipulation do you want us making? Put another way, do you want to pay for stats that aren't based on history but more based on a hunch? My answer is if I'm 95% comfortable with something, I'd adjust it. For example, if a mid 1980's Wade Boggs went from hitting leadoff for the Red Sox and had been dealt to the 1985 Cardinals for Tommie Herr, where he would instead hit 2nd or 3rd behind Vince Coleman and Willie McGee, yes, I'd adjust RBI totals upwards.
But in many of these cases you're just adding bias to a situation without a lot of hard evidence - and yes you're paying us to make gut calls and we make them all the time, but many of those gut calls are to do nothing and let the data be as it is.
So I take Gary's statment above to mean: like most things in life, its not black or white, but gray. The projections are not completely void of bias (players create their own, and Gary and Todd impart their own when they feel strongly about it) but they are not applying added bias to every projection as not every projection warrants additional bias. Am I completely off here, do I have it nailed, or am I in the gray yet again?AllstonRockCity wrote: Though ultimately I suppose it doesn't matter so much if the projections are based on poll option 1 or poll option 2 as long as you know whether or not this bias is already built in. If I know its not then I can impart that bias myself. But, if I know it has been then I would have to feel very strongly 1 way or the other to impart additional bias on top of the projections built in bias.
Re: Projection Theory Poll #3
As with your other observations in this thread, you're spot on.
Re: Projection Theory Poll #3
This is exactly what I have said since joining my first Fantasy sport.....projections are just what the word means. My only expectation is for a professional "expert" to be the one to do the hard research and go totally on "projections" not gut predictions. I can do that bad enough on my own . Something unique about sports and the fantasy world....I play every year in a college basketball bracket(as I am sure a lot of people reading this does) in the past 4 years, I have finished first, 3 times and third place once. Am I college basketball expert? HA! I don't even watch it! EVER! Never have, prolly never will. I go thru the brackett and simply look at there season records and where they "PROJECTION" is for that season. Also, the fact that is seems the same 4-8 teams are in it every year.....if I see a team I think sounds kewl that day, I pick em......If I watched it and researched it, then I would be all over the place and probably never win again.....and yes I know there is BIG difference between that and baseball....this DOES take knowledge and stats...thats why I am in this is AWESOME forum of knowledge!!!