rolyaTy
You're Heinous
First off, I'm not sure if this has been discussed here or not, so if it has feel free to disregard it, but it's new to me. I'm not sure it's useful or informative, so I'll bounce it off you guys first and I'm sure you'll tell me what you think.
Our league just finished drafting, and as usual, I post my analysis of the draft, complete with insults, jokes, and I try to include a little factual information in there too. Part of my analysis is based on how people did with getting players with low ADP's, part of it is how well they filled key roles, my takes on players, etc.
This year, I got to thinking more about the concept of ADP of players, and your draft position's inherent value. In strict serpentine drafts, with no trading, with 12 teams and 16 rounds, each team ends up with an average (team) pick of 96.5. In other words, if you drafted strictly according to your draft position (DP), you should end up with a team comprised roughly of players with an average ADP of 96.5. Of course, this isn't practically true, as you draft different players for different leagues, ADP's from the web aren't ideal (they don't represent true value), etc...however, I think there is some value in looking at things like this.
One example which backs up this idea is when you consider drafts of uneven numbers of rounds. If your draft has an uneven number of rounds, the DP of those at the top of the draft will be better than the DP of those at the bottom of the draft. This is obvious, when you think about it. Look at a 1 round draft, with 10 teams. The difference in DP between each position is 1, and from 1 to 10 the difference is 9. When you get up to say, 13 rounds and keep 10 teams, the difference between 1st pick and 2nd pick is only 1/13 or 0.08 and from 1 to 10 it's .8 or so. So the more rounds you have, the more the differences don't matter to the type of team you can draft.
What this goes to show is that by looking at the DP of a team, one can get information that is able to be compared to other teams in the league, and a value judgement can be made based on numbers alone.
So I was thinking that after the draft is complete, you can calculate the DP of a team, and then compare that DP to the ADP of the players drafted on that team, and then compare these values to evaluate how the draft went. This would also be interesting to correlate to final season results (ignoring FA acquisitions and injuries) to determine whether some players are better drafters, or are better projectors, than the average FF player.
Average Draft Positions are determined by a large sample size, and when its large enough, it represents the collective FF group's average projection of value for a FF player. It is where some concepts of value in drafts come from, and it's the best approximation of a players value at any given time that I know of.
So, when drafts come around, one of the goals is to not take players "too early", in reference to the ADP. And if a player "falls" to you, in reference to ADP again, you might do well to take him. Ideally, you want to maximize the quality of players you get during a draft, and minimize the number of times you have to "reach" for a player. Most every time you reach, you are placing your opinion of a players value above the collective FF valuation of the player (unless you're filling a roster spot, or have too many at one spot, etc...assume i'm talking about people of the same position).
To me, this means that one can analyze a draft based on the total ADP you accumulate on your team (TADP), assuming you also fill key positions, have depth, cover byes, etc. If you average your teams ADP, compare it to your DP, and then compare how you did to other teams, you can get an idea of how well your draft went, and how well others drafts went.
For instance, if your DP was 96.5 (12 team 16 round) and your average TADP was 96.63, then that means you drafted pretty close to the average, and by all accounts, had an average quality draft as you didn't acquire a lot of players at a bargain.
If your DP was 96.5 and your TADP was 140, then you likely reached for a LOT of players, and you would have to have extremely good insights to be able to do well in the league, as you have gone far away from mainstream FF thinking about the value of players.
If your DP was 96.5 and your TADP was 80, then it seems that you acquired a good number of highly ranked players at a bargain.
In analyzing my league this year, these are the results, by draft pick #:
Draft# DP TADP
1st 95.81 102.88
2nd 95.94 95.25
3rd 96.06 146.94
4th 96.19 102.81
5th 96.31 92.88
6th 96.44 98.94
7th 96.56 83.31
8th 96.69 101.19
9th 96.81 104.25
10th 96.94 98.44
11th 97.06 81.25
12th 97.19 96.63
A little information about our league. Three (optional) keeper league, which will skew this towards the bottom as the bottom teams kept more players as they had better keepers. The DP's should've all been 96.5, but we do our order a little differently, as the first round is mirrored with the 4th round...makes it seem uneven, even though we have 12 teams with 16 rounds.
The interesting thing to me is that a few teams jump out right away. The first being the 146. This poor drafting can be explained by the owner not being at the draft, and not submitting a cheatsheet. His draft was conducted by the MFL autodraft software, and he really came out bad, not to mention his keepers were Favre, Fred Taylor and Thomas Jones (cost him 1st, 2nd and 3rd round picks to keep them). This is flagged as the worst of the draft, and obviously so.
Team 1, is a NO homer, and got reggie bush #1 overall, and made a few homer picks as well. This drew down his TADP. #9 was a Buf homer, and reached for a few of them, but nothing bad, had a good draft overall but started with questionable picks after round 10. Still, not too bad.
Aside from the autodraft guy, we all did pretty well. #11 was also autodrafted by MFL, but started with keepers of Portis, Chad Johnson, and Larry Fitz, which really boosted his stock, as well as the computer autodrafting Frank Gore before the trade...that really helped him (ADP's updated today). He had the best overall draft. #7 did well, coming in 13 points below his DP, and consistantly drafted people who were overlooked by other teams.
So this is just one example of how the numbers can be used. You can compare the TADP to the DP and see how you did with respect to "normal", where normal is the DP. As ADP's change closer to the season, one could rerun this analysis with updated ADP's, while storing the previous TADP, to see how well an owner picked with respect to the rankings at the start of the season. This can go on until ADP's are no longer valid. At the end of the season, the final standings can be correlated with the difference between the DP and the TADP's, to see which owner was truly the best estimator of talent this year (or the luckiest, a distinction which matters little in FF).
In addition, when trying to decide between various projections from people you want to use, such as footballguys here, compare their rankings to the ADP's using this method, and then at the end of the year, determine if anyone had any surprising insight into the value of players that the rest of the FF community overlooked.
So, that was long, and i'm sure few got through it. I'm not touting this as being anything special, just something I was thinking about and put a little effort into looking into. If it's nothing more than stating the obvious, I could see someone saying that, but it's quantifying the obvious that's cool to me. It's easy to see that a person who consistantly reaches for players had a bad draft, but how does his bad draft compare to another player in the league who also reached. Can you correlate qualitative information like that with quantitative year end results, etc.
Our league just finished drafting, and as usual, I post my analysis of the draft, complete with insults, jokes, and I try to include a little factual information in there too. Part of my analysis is based on how people did with getting players with low ADP's, part of it is how well they filled key roles, my takes on players, etc.
This year, I got to thinking more about the concept of ADP of players, and your draft position's inherent value. In strict serpentine drafts, with no trading, with 12 teams and 16 rounds, each team ends up with an average (team) pick of 96.5. In other words, if you drafted strictly according to your draft position (DP), you should end up with a team comprised roughly of players with an average ADP of 96.5. Of course, this isn't practically true, as you draft different players for different leagues, ADP's from the web aren't ideal (they don't represent true value), etc...however, I think there is some value in looking at things like this.
One example which backs up this idea is when you consider drafts of uneven numbers of rounds. If your draft has an uneven number of rounds, the DP of those at the top of the draft will be better than the DP of those at the bottom of the draft. This is obvious, when you think about it. Look at a 1 round draft, with 10 teams. The difference in DP between each position is 1, and from 1 to 10 the difference is 9. When you get up to say, 13 rounds and keep 10 teams, the difference between 1st pick and 2nd pick is only 1/13 or 0.08 and from 1 to 10 it's .8 or so. So the more rounds you have, the more the differences don't matter to the type of team you can draft.
What this goes to show is that by looking at the DP of a team, one can get information that is able to be compared to other teams in the league, and a value judgement can be made based on numbers alone.
So I was thinking that after the draft is complete, you can calculate the DP of a team, and then compare that DP to the ADP of the players drafted on that team, and then compare these values to evaluate how the draft went. This would also be interesting to correlate to final season results (ignoring FA acquisitions and injuries) to determine whether some players are better drafters, or are better projectors, than the average FF player.
Average Draft Positions are determined by a large sample size, and when its large enough, it represents the collective FF group's average projection of value for a FF player. It is where some concepts of value in drafts come from, and it's the best approximation of a players value at any given time that I know of.
So, when drafts come around, one of the goals is to not take players "too early", in reference to the ADP. And if a player "falls" to you, in reference to ADP again, you might do well to take him. Ideally, you want to maximize the quality of players you get during a draft, and minimize the number of times you have to "reach" for a player. Most every time you reach, you are placing your opinion of a players value above the collective FF valuation of the player (unless you're filling a roster spot, or have too many at one spot, etc...assume i'm talking about people of the same position).
To me, this means that one can analyze a draft based on the total ADP you accumulate on your team (TADP), assuming you also fill key positions, have depth, cover byes, etc. If you average your teams ADP, compare it to your DP, and then compare how you did to other teams, you can get an idea of how well your draft went, and how well others drafts went.
For instance, if your DP was 96.5 (12 team 16 round) and your average TADP was 96.63, then that means you drafted pretty close to the average, and by all accounts, had an average quality draft as you didn't acquire a lot of players at a bargain.
If your DP was 96.5 and your TADP was 140, then you likely reached for a LOT of players, and you would have to have extremely good insights to be able to do well in the league, as you have gone far away from mainstream FF thinking about the value of players.
If your DP was 96.5 and your TADP was 80, then it seems that you acquired a good number of highly ranked players at a bargain.
In analyzing my league this year, these are the results, by draft pick #:
Draft# DP TADP
1st 95.81 102.88
2nd 95.94 95.25
3rd 96.06 146.94
4th 96.19 102.81
5th 96.31 92.88
6th 96.44 98.94
7th 96.56 83.31
8th 96.69 101.19
9th 96.81 104.25
10th 96.94 98.44
11th 97.06 81.25
12th 97.19 96.63
A little information about our league. Three (optional) keeper league, which will skew this towards the bottom as the bottom teams kept more players as they had better keepers. The DP's should've all been 96.5, but we do our order a little differently, as the first round is mirrored with the 4th round...makes it seem uneven, even though we have 12 teams with 16 rounds.
The interesting thing to me is that a few teams jump out right away. The first being the 146. This poor drafting can be explained by the owner not being at the draft, and not submitting a cheatsheet. His draft was conducted by the MFL autodraft software, and he really came out bad, not to mention his keepers were Favre, Fred Taylor and Thomas Jones (cost him 1st, 2nd and 3rd round picks to keep them). This is flagged as the worst of the draft, and obviously so.
Team 1, is a NO homer, and got reggie bush #1 overall, and made a few homer picks as well. This drew down his TADP. #9 was a Buf homer, and reached for a few of them, but nothing bad, had a good draft overall but started with questionable picks after round 10. Still, not too bad.
Aside from the autodraft guy, we all did pretty well. #11 was also autodrafted by MFL, but started with keepers of Portis, Chad Johnson, and Larry Fitz, which really boosted his stock, as well as the computer autodrafting Frank Gore before the trade...that really helped him (ADP's updated today). He had the best overall draft. #7 did well, coming in 13 points below his DP, and consistantly drafted people who were overlooked by other teams.
So this is just one example of how the numbers can be used. You can compare the TADP to the DP and see how you did with respect to "normal", where normal is the DP. As ADP's change closer to the season, one could rerun this analysis with updated ADP's, while storing the previous TADP, to see how well an owner picked with respect to the rankings at the start of the season. This can go on until ADP's are no longer valid. At the end of the season, the final standings can be correlated with the difference between the DP and the TADP's, to see which owner was truly the best estimator of talent this year (or the luckiest, a distinction which matters little in FF).
In addition, when trying to decide between various projections from people you want to use, such as footballguys here, compare their rankings to the ADP's using this method, and then at the end of the year, determine if anyone had any surprising insight into the value of players that the rest of the FF community overlooked.
So, that was long, and i'm sure few got through it. I'm not touting this as being anything special, just something I was thinking about and put a little effort into looking into. If it's nothing more than stating the obvious, I could see someone saying that, but it's quantifying the obvious that's cool to me. It's easy to see that a person who consistantly reaches for players had a bad draft, but how does his bad draft compare to another player in the league who also reached. Can you correlate qualitative information like that with quantitative year end results, etc.
