moleculo
Footballguy
I would like to do a study based on what can be learned from multiple pre-season rankings. I am not so much trying to see who is better at ranking or who was the closest. Instead, I am looking for a method to understand who might be under-ranked vs AVT and who might be overranked.
Motivation: Certain players every year far exceed their expectations from the experts in general. Of course, certain other players underachieve as well. Knowing who is risky (or, conversely, who might have considerable upside) would be invaluable heading into a draft. I will refer to these players as boom/bust players.
Hypothesis: boom/bust players can be identified by calculating the standard deviation (or some other statistical calculation) of pre-season rankings.
Anyone who cares enough to do rankings is looking at data. We are all looking at the same global set of data, but each ranking weighs the impact of various data attributes differently. Many people looking at the global data will identify different elements, and will thus rank accordingly. Comparing the deviation of independant rankings will show differences in opinions, which may be a strong indicator of boom/bust potiental.
Potiental issues:
1. as many independantly formulated pre-season rankings as possible are needed to lend statistical robustness.
2. impact of injury cannot be accounted for (or can it? is injury risk already accounted for?)
3. process would be unable to determine if a boom/bust player is likely to boom or bust. I may not be able to separate these two, further analysis may be needed, but we have until July to work further.
Discussion: Nearly every ranking list is similar for the most part. Nearly everyone had LT, Holmes, SA at the top of the list for RB's, Manning, Culpepper, etc. for QB, etc for QB, and so forth. Most pre-draft rankings will be nearly identical, with a few exceptions. Burleson might be high in some lists, Steve Smith might be higher in others.
A draft is not made and a league is not won in the first round. Yet, SA vs LT vs Holmes has been debated immensely. I say that LT vs SA is actually irrelevant. Anyone can draft in the first round by looking at any ranking on the web or in a magazine, and any actual difference in performance weighs heavily on variables that are beyond forcasting. The mid to late rounds are where the core of the debate should be, and this is where the differences in opinion will be the most dramatic.
Going into the draft, we all know the strategies of AVT, VBD, DVBD, tiering, gut feel, etc. Are these really the best strategies? I suspect that the most important aspect of a sucessful draft is managing risk. If you have a sure fire RB such as LT, you might be more willing to gamble with a boom/bust player such as Mike Anderson or Willis McGahee. Conversely, if you are depending on a risky RB as your RB1, you might be better of playing it safe with other future picks. Regardless, the only tool I know of for managing risk is intuition.
For a smaller league (6-8 teams), a sucessful team could be made by drafting many boom/bust players, and then using the WW to backfill for any losers. For a larger league (10+ teams), it might be better to avoid such players to minimize risk, or at the very least to minimize the impact of potiental busts. Hence, drafting with an eye on risk management may be as important (if not more important) as the above mentioned strategies. Note that risk management is not a drafting strategy in itself, rather it is a supplement to accepted strategies.
Factors that may make a player boom/bust might include coaching changes, coming off of injuries (steve Smith), uncertainty as to who will start (Bell vs Anderson), increased compentancy (Palmer and Eli Manning), player movement (TO, Plaxico, Moss, Moss), rookie production (Ronnie Brown), etc.
What I need from you: Please, please, please send me your pre-season rankings. Again, I need as many as possible. I am looking for rankings as complex as the FBG, as arcane as Sports Illustrated, as personal as your own projections, and as basic as a list w/ no projections. Anything you have I would be happy to use. Of course, when I have the data compiled, anyone providing rankings will be credited.
Please e-mail any 2005 pre-season rankings to: moleculo@bellsouth.net
Thanks in advance for any help. As always, comments/suggestions are always appreciated.
Motivation: Certain players every year far exceed their expectations from the experts in general. Of course, certain other players underachieve as well. Knowing who is risky (or, conversely, who might have considerable upside) would be invaluable heading into a draft. I will refer to these players as boom/bust players.
Hypothesis: boom/bust players can be identified by calculating the standard deviation (or some other statistical calculation) of pre-season rankings.
Anyone who cares enough to do rankings is looking at data. We are all looking at the same global set of data, but each ranking weighs the impact of various data attributes differently. Many people looking at the global data will identify different elements, and will thus rank accordingly. Comparing the deviation of independant rankings will show differences in opinions, which may be a strong indicator of boom/bust potiental.
Potiental issues:
1. as many independantly formulated pre-season rankings as possible are needed to lend statistical robustness.
2. impact of injury cannot be accounted for (or can it? is injury risk already accounted for?)
3. process would be unable to determine if a boom/bust player is likely to boom or bust. I may not be able to separate these two, further analysis may be needed, but we have until July to work further.
Discussion: Nearly every ranking list is similar for the most part. Nearly everyone had LT, Holmes, SA at the top of the list for RB's, Manning, Culpepper, etc. for QB, etc for QB, and so forth. Most pre-draft rankings will be nearly identical, with a few exceptions. Burleson might be high in some lists, Steve Smith might be higher in others.
A draft is not made and a league is not won in the first round. Yet, SA vs LT vs Holmes has been debated immensely. I say that LT vs SA is actually irrelevant. Anyone can draft in the first round by looking at any ranking on the web or in a magazine, and any actual difference in performance weighs heavily on variables that are beyond forcasting. The mid to late rounds are where the core of the debate should be, and this is where the differences in opinion will be the most dramatic.
Going into the draft, we all know the strategies of AVT, VBD, DVBD, tiering, gut feel, etc. Are these really the best strategies? I suspect that the most important aspect of a sucessful draft is managing risk. If you have a sure fire RB such as LT, you might be more willing to gamble with a boom/bust player such as Mike Anderson or Willis McGahee. Conversely, if you are depending on a risky RB as your RB1, you might be better of playing it safe with other future picks. Regardless, the only tool I know of for managing risk is intuition.
For a smaller league (6-8 teams), a sucessful team could be made by drafting many boom/bust players, and then using the WW to backfill for any losers. For a larger league (10+ teams), it might be better to avoid such players to minimize risk, or at the very least to minimize the impact of potiental busts. Hence, drafting with an eye on risk management may be as important (if not more important) as the above mentioned strategies. Note that risk management is not a drafting strategy in itself, rather it is a supplement to accepted strategies.
Factors that may make a player boom/bust might include coaching changes, coming off of injuries (steve Smith), uncertainty as to who will start (Bell vs Anderson), increased compentancy (Palmer and Eli Manning), player movement (TO, Plaxico, Moss, Moss), rookie production (Ronnie Brown), etc.
What I need from you: Please, please, please send me your pre-season rankings. Again, I need as many as possible. I am looking for rankings as complex as the FBG, as arcane as Sports Illustrated, as personal as your own projections, and as basic as a list w/ no projections. Anything you have I would be happy to use. Of course, when I have the data compiled, anyone providing rankings will be credited.
Please e-mail any 2005 pre-season rankings to: moleculo@bellsouth.net
Thanks in advance for any help. As always, comments/suggestions are always appreciated.
Last edited by a moderator:
