AdamJT13 and the trade value chart

[Patsfanin Philly]

Does anyone remember the specifics of the 1995 trade when we lost 3 spots? (trading a higher pick for a lower pick straight up)

I suppose if I weren't lazy I could just look it up

Official New England Patriots - History - All-Time Draft/Trade.

Read more at: Official New England Patriots - History - All-Time Draft/Trade​

7th round traded #231 to Cleveland for #234

231 A.C. Tellison -- Miami (Fla.) Cleveland Browns









234 Carlos Yancy DB Georgia New England Patriots

 

[MaineMan]

Official New England Patriots - History - All-Time Draft/Trade.

Read more at: Official New England Patriots - History - All-Time Draft/Trade​

7th round traded #231 to Cleveland for #234

231 A.C. Tellison -- Miami (Fla.) Cleveland Browns









234 Carlos Yancy DB Georgia New England Patriots

IOW, the Pats gave up something like 2.2 points and got 2.15 points? OMG!! Shades of Ditka!!!

 

[patchick]

IOW, the Pats gave up something like 2.2 points and got 2.15 points? OMG!! Shades of Ditka!!!

Sounds like a "dang, dropped my notecards -- you pick while i sort out my board" trade.

 

[ctpatsfan77]

Bingo!
Biggest media myth: #33 has so much extra value. Ridiculous. If #32 can be signed for an extra year, the value of #33 drops significantly. And this concept of "an extra night to digest the draft and reorganize and assess"....must be so confusing for teams that have had to make ONE whole pick. How can they possible see the draft clearly now that 32 players are gone.

While I agree with a lot of what you say, there is one key issue you're overlooking. The draft does not involve what in economics would be called "rational actors." Factors unrelated to the actual values of the players (e.g., GM job security) often come into play, and that's where the extra value for #33 comes in—especially for a team like NE that doesn't desperately need that pick (the way that Carolina would if they still had it).

 

[MaineMan]

I like this analysis, I wish I thought about it myself. Now if someone could calculate all the averages for each pick and put that in a grid it make it that much better.

Going from AdamJT's list of 1995 - 2009, there are several picks that have only been traded down twice (and #2, not at all), so that's a pretty small sample size for getting meaningful results just doing direct averages pick-by-pick.

I'm still "translating" his pick numbers to standard chart values and have worked through the #28, so far. Once I've gotten through, say, #35, I'm going to see if I can work out some averages for 3-pick ranges. For example, picks #16 (1000), #17 (950) and #18 (900) average 950 on the chart. The average compensation in the eleven "trades down" from those picks is 948 with a high of 1058 (for a #17) and a low of 824 (for a #16).

I might even do a "rolling" 3-pick average (e.g. #15-#17, #16-#18, #17-#19), but I seriously doubt that the results will be radically different, at least after the first 5 picks or so. Starting somewhere around the middle of round one, it appears that the average compensation difference will be roughly equivalent to a 7th round compensatory pick.

IOW, what this boils down to is an exercise that will likely end up demonstrating that the "approximations" of the standard draft trade value chart (which is all they were intended to be in the first place) are pretty close after all.

 

[patchick]

Going from AdamJT's list of 1995 - 2009, there are several picks that have only been traded down twice (and #2, not at all), so that's a pretty small sample size for getting meaningful results just doing direct averages pick-by-pick.

I'm still "translating" his pick numbers to standard chart values and have worked through the #28, so far. Once I've gotten through, say, #35, I'm going to see if I can work out some averages for 3-pick ranges. For example, picks #16 (1000), #17 (950) and #18 (900) average 950 on the chart. The average compensation in the eleven "trades down" from those picks is 948 with a high of 1058 (for a #17) and a low of 824 (for a #16).

I might even do a "rolling" 3-pick average (e.g. #15-#17, #16-#18, #17-#19), but I seriously doubt that the results will be radically different, at least after the first 5 picks or so. Starting somewhere around the middle of round one, it appears that the average compensation difference will be roughly equivalent to a 7th round compensatory pick.

IOW, what this boils down to is an exercise that will likely end up demonstrating that the "approximations" of the standard draft trade value chart (which is all they were intended to be in the first place) are pretty close after all.

I suspect you're right about the outcome...but regarding the scant data, maybe the goal is to fit a curve rather than compute pick-by-pick history?

 

[borg]

While I agree with a lot of what you say, there is one key issue you're overlooking. The draft does not involve what in economics would be called "rational actors." Factors unrelated to the actual values of the players (e.g., GM job security) often come into play, and that's where the extra value for #33 comes in—especially for a team like NE that doesn't desperately need that pick (the way that Carolina would if they still had it).

Looking at the last 5 drafts, only one 1st pick of day 2 or day 3 involved a trade. It just doesn't appear that GMs are getting all lathered up after a nights rest.

 

[WhiZa]

Going from AdamJT's list of 1995 - 2009, there are several picks that have only been traded down twice (and #2, not at all), so that's a pretty small sample size for getting meaningful results just doing direct averages pick-by-pick.

I'm still "translating" his pick numbers to standard chart values and have worked through the #28, so far. Once I've gotten through, say, #35, I'm going to see if I can work out some averages for 3-pick ranges. For example, picks #16 (1000), #17 (950) and #18 (900) average 950 on the chart. The average compensation in the eleven "trades down" from those picks is 948 with a high of 1058 (for a #17) and a low of 824 (for a #16).

I might even do a "rolling" 3-pick average (e.g. #15-#17, #16-#18, #17-#19), but I seriously doubt that the results will be radically different, at least after the first 5 picks or so. Starting somewhere around the middle of round one, it appears that the average compensation difference will be roughly equivalent to a 7th round compensatory pick.

IOW, what this boils down to is an exercise that will likely end up demonstrating that the "approximations" of the standard draft trade value chart (which is all they were intended to be in the first place) are pretty close after all.

Nice work MaineMan! I can't wait to see the results

I suspect you're right about the outcome...but regarding the scant data, maybe the goal is to fit a curve rather than compute pick-by-pick history?

This is my thought as well. It's probably easier to average each pick out instead of getting it exact. Though both would be nice.

 

[ctpatsfan77]

Looking at the last 5 drafts, only one 1st pick of day 2 or day 3 involved a trade. It just doesn't appear that GMs are getting all lathered up after a nights rest.

Well, this is only the second year of the current format, and all we know for sure is that St. Louis didn't get an offer they considered better than getting the player they wanted.

A lot of it comes down to who's still left on the board and whether or not the Bills have taken a QB when day one ends.

 

[supafly]

Official New England Patriots - History - All-Time Draft/Trade.

Read more at: Official New England Patriots - History - All-Time Draft/Trade​

7th round traded #231 to Cleveland for #234

231 A.C. Tellison -- Miami (Fla.) Cleveland Browns









234 Carlos Yancy DB Georgia New England Patriots

Thank you kindly, much obliged.

 

[MaineMan]

Okay, here ya go - "rolling average" for three-pick ranges for the entire first round history provided by AdamJT.

These are all "trades down".

I've "normalized" some of the actual trades - those trades in which the team trading down included a second pick. This appears, in every case so far, to have been done to actually better balance the ledger WRT the standard trade values. For example, there was a trade in which one team gave up the #6, #37 & # 132 (2222 points) for the #17, #18 & #54 (2210 points). The point difference (12) is equivalent to the #201 (early 7th). The percentage variation from SCV is 0.5%.

Trades down from the #1 overall pick are pretty clearly anomalous compared to the vast majority of the other 1st round picks, although there are a few "Ditka Debacle" anomalies later, too (9, actually).

SVC = standard chart value
============================

PICK (# trades) /avg SCV / comp range / AVG COMP / DIFF (%) / VALUE
-
#1 (3) / 3000 / 1957-2683+ / 2293 / -707 (-24%) / #26
[NOTE: Trades were all down to #4-#6, picking up at least an extra 3rd, plus some.]
["+" denotes compensation that included a next year's 1st, figured at the minimum value of 590. All of these deals were of the "Ditka Debacle" variety.]

#2 (2600) [no trades]

#3-5 (5) / 1920 / 1675-2247+(*) / 2004 / +124 (+7%) / #94 (late 3rd)
[* = the actual "Ditka Debacle"]

#4-6 (8) / 1650/ 1476-2247+(*) / 1766 / +116 (+7%) / #96 (end of 3rd)

#5-7 (12) / 1575 / 1473-2247+(*) / 1685 / +110 (+7%) / #98 (start of 4th)

#6-8 (11) / 1536 / 1127-2030 / 1584 / +48 (+3%) / #124 (late 4th)

#7-9 (7) / 1464 / 1127-1736 / 1522 / +58 (+4%) / #118 (late 4th)

#8-10 (4) / 1338 / 1127-1500+ / 1366 / +28 (+2%) / #161 (start of 6th)

#9-11 (5) / 1290 / 1275-1500+ / 1382 / +92 (+7%) / #102 (early 4th)

#10-12 (7) / 1243 / 1170-1500+ / 1320 / +77 (+6%) / #109 (mid-4th)

#11-13 (9) / 1189 / 1044-1420 / 1210 / +21 (+2%) / #178 (mid 6th)

#12-14 (10) / 1150 / 1041-1420 / 1167 / +17 (+1%) / #188 (late 6th)

#13-15 (9) / 1111 / 904-1293 / 1093 / -18 (-2%) / #186 (late 6th)

#14-16 (8) / 1050 / 824-1124 / 1012 / -38 (-4%) / #136 (early 5th)

#15-17 (10) / 985 / 824-1058 / 965 / -20 (-2%) / #181 (late 6th)

#16-18 (11) / 950 / 824-1058 / 948 / -2 (0%) / ("8th" rounder)

#17-19 (14) / 907 / 814-1270+ / 964 / 57 (6%) / #118 (late 4th)
[NOTE: Two trades involving #19 were clearly trades up from the 2nd by teams that didn't have a 1st and that included a "next 1st".]

#18-20 (12) / 875 / 814-1270+ / 935 / 60 (7%) / #117 (late 4th)

#19-21 (11) / 855 / 710-1270+ / 916 / 61 (7%) / #117 (late 4th)

#20-22 (11) / 803 / 710-1130+ / 884 / 81 (10%) / #106 (early 4th)

#21-23 (12) / 777 / 710-1130+ / 868 / 91 (12%) / #102 (early 4th)

#22-24 (12) / 767 / 719-1130+ / 859 / 92 (12%) / #102 (early 4th)

#23-25 (9) / 742 / 719-939+ / 801 / 59 (8%) / #117 (late 4th)

#24-26 (8) / 718 / 705-844+ / 757 / 38 (5%) / #136 (early 5th)

#25-27 (8) / 706 / 705-844+ / 779 / 73 (10%) / #110 (mid-4th)

#26-28 (8) / 680 / 645-860+ / 736 / 56 (8%) / #119 (late 4th)

#27-29 (8) / 658 / 645-860+ / 759 / 101 (15%) / #100 (early 4th)

#28-30 (9) / 640 / 608-812+ / 719 / 79 (12%) / #107 (early 4th)

#31 (600) [no trades]
#32 (590) [no trades]

There were 84 total trade down transactions involving picks #3-#30. Of those, nine were of the anomalous "Ditka Debacle" variety, IOW way over 20% variation from SCV (as much as 45%). The remaining 75 transactions averaged within 7% of SCV.

Of the 65 transactions involving picks #11-30, 65% were within 10% of SCV and 51% were with 5% of SCV.
=========================================

#33-35 (5) / 570 / 594-690 / 656 / +86 (15%) / #104 (early 4th)

And, futhermore, WHEW!

I'll try to do more with the 2nd round tomorrow, but I'm not promising anything.

 

[supafly]

Okay, here ya go - "rolling average" for three-pick ranges for the entire first round history provided by AdamJT.

These are all "trades down".

I've "normalized" some of the actual trades - those trades in which the team trading down included a second pick. This appears, in every case so far, to have been done to actually better balance the ledger WRT the standard trade values. For example, there was a trade in which one team gave up the #6, #37 & # 132 (2222 points) for the #17, #18 & #54 (2210 points). The point difference (12) is equivalent to the #201 (early 7th). The percentage variation from SCV is 0.5%.

Trades down from the #1 overall pick are pretty clearly anomalous compared to the vast majority of the other 1st round picks, although there are a few "Ditka Debacle" anomalies later, too (9, actually).

SVC = standard chart value
============================

PICK (# trades) /avg SCV / comp range / AVG COMP / DIFF (%) / VALUE
-
#1 (3) / 3000 / 1957-2683+ / 2293 / -707 (-24%) / #26
[NOTE: Trades were all down to #4-#6, picking up at least an extra 3rd, plus some.]
["+" denotes compensation that included a next year's 1st, figured at the minimum value of 590. All of these deals were of the "Ditka Debacle" variety.]

#2 (2600) [no trades]

#3-5 (5) / 1920 / 1675-2247+(*) / 2004 / +124 (+7%) / #94 (late 3rd)
[* = the actual "Ditka Debacle"]

#4-6 (8) / 1650/ 1476-2247+(*) / 1766 / +116 (+7%) / #96 (end of 3rd)

#5-7 (12) / 1575 / 1473-2247+(*) / 1685 / +110 (+7%) / #98 (start of 4th)

#6-8 (11) / 1536 / 1127-2030 / 1584 / +48 (+3%) / #124 (late 4th)

#7-9 (7) / 1464 / 1127-1736 / 1522 / +58 (+4%) / #118 (late 4th)

#8-10 (4) / 1338 / 1127-1500+ / 1366 / +28 (+2%) / #161 (start of 6th)

#9-11 (5) / 1290 / 1275-1500+ / 1382 / +92 (+7%) / #102 (early 4th)

#10-12 (7) / 1243 / 1170-1500+ / 1320 / +77 (+6%) / #109 (mid-4th)

#11-13 (9) / 1189 / 1044-1420 / 1210 / +21 (+2%) / #178 (mid 6th)

#12-14 (10) / 1150 / 1041-1420 / 1167 / +17 (+1%) / #188 (late 6th)

#13-15 (9) / 1111 / 904-1293 / 1093 / -18 (-2%) / #186 (late 6th)

#14-16 (8) / 1050 / 824-1124 / 1012 / -38 (-4%) / #136 (early 5th)

#15-17 (10) / 985 / 824-1058 / 965 / -20 (-2%) / #181 (late 6th)

#16-18 (11) / 950 / 824-1058 / 948 / -2 (0%) / ("8th" rounder)

#17-19 (14) / 907 / 814-1270+ / 964 / 57 (6%) / #118 (late 4th)
[NOTE: Two trades involving #19 were clearly trades up from the 2nd by teams that didn't have a 1st and that included a "next 1st".]

#18-20 (12) / 875 / 814-1270+ / 935 / 60 (7%) / #117 (late 4th)

#19-21 (11) / 855 / 710-1270+ / 916 / 61 (7%) / #117 (late 4th)

#20-22 (11) / 803 / 710-1130+ / 884 / 81 (10%) / #106 (early 4th)

#21-23 (12) / 777 / 710-1130+ / 868 / 91 (12%) / #102 (early 4th)

#22-24 (12) / 767 / 719-1130+ / 859 / 92 (12%) / #102 (early 4th)

#23-25 (9) / 742 / 719-939+ / 801 / 59 (8%) / #117 (late 4th)

#24-26 (8) / 718 / 705-844+ / 757 / 38 (5%) / #136 (early 5th)

#25-27 (8) / 706 / 705-844+ / 779 / 73 (10%) / #110 (mid-4th)

#26-28 (8) / 680 / 645-860+ / 736 / 56 (8%) / #119 (late 4th)

#27-29 (8) / 658 / 645-860+ / 759 / 101 (15%) / #100 (early 4th)

#28-30 (9) / 640 / 608-812+ / 719 / 79 (12%) / #107 (early 4th)

#31 (600) [no trades]
#32 (590) [no trades]

There were 84 total trade down transactions involving picks #3-#30. Of those, nine were of the anomalous "Ditka Debacle" variety, IOW way over 20% variation from SCV (as much as 45%). The remaining 75 transactions averaged within 7% of SCV.

Of the 65 transactions involving picks #11-30, 65% were within 10% of SCV and 51% were with 5% of SCV.
=========================================

#33-35 (5) / 570 / 594-690 / 656 / +86 (15%) / #104 (early 4th)

And, futhermore, WHEW!

I'll try to do more with the 2nd round tomorrow, but I'm not promising anything.

Thank you Rainman...I mean MaineMan

 

[Off The Grid]

While I agree with a lot of what you say, there is one key issue you're overlooking. The draft does not involve what in economics would be called "rational actors." Factors unrelated to the actual values of the players (e.g., GM job security) often come into play, and that's where the extra value for #33 comes in—especially for a team like NE that doesn't desperately need that pick (the way that Carolina would if they still had it).

Strong point, per usual, sir.

I don't see us reaping a CRAZY Premium for #33's Nominal Value, but I DO see a Premium.

 

[Off The Grid]

Okay, here ya go - "rolling average" for three-pick ranges for the entire first round history provided by AdamJT.

These are all "trades down".

I've "normalized" some of the actual trades - those trades in which the team trading down included a second pick. This appears, in every case so far, to have been done to actually better balance the ledger WRT the standard trade values. For example, there was a trade in which one team gave up the #6, #37 & # 132 (2222 points) for the #17, #18 & #54 (2210 points). The point difference (12) is equivalent to the #201 (early 7th). The percentage variation from SCV is 0.5%.

Trades down from the #1 overall pick are pretty clearly anomalous compared to the vast majority of the other 1st round picks, although there are a few "Ditka Debacle" anomalies later, too (9, actually).

SVC = standard chart value
============================

PICK (# trades) /avg SCV / comp range / AVG COMP / DIFF (%) / VALUE
-
#1 (3) / 3000 / 1957-2683+ / 2293 / -707 (-24%) / #26
[NOTE: Trades were all down to #4-#6, picking up at least an extra 3rd, plus some.]
["+" denotes compensation that included a next year's 1st, figured at the minimum value of 590. All of these deals were of the "Ditka Debacle" variety.]

#2 (2600) [no trades]

#3-5 (5) / 1920 / 1675-2247+(*) / 2004 / +124 (+7%) / #94 (late 3rd)
[* = the actual "Ditka Debacle"]

#4-6 (8) / 1650/ 1476-2247+(*) / 1766 / +116 (+7%) / #96 (end of 3rd)

#5-7 (12) / 1575 / 1473-2247+(*) / 1685 / +110 (+7%) / #98 (start of 4th)

#6-8 (11) / 1536 / 1127-2030 / 1584 / +48 (+3%) / #124 (late 4th)

#7-9 (7) / 1464 / 1127-1736 / 1522 / +58 (+4%) / #118 (late 4th)

#8-10 (4) / 1338 / 1127-1500+ / 1366 / +28 (+2%) / #161 (start of 6th)

#9-11 (5) / 1290 / 1275-1500+ / 1382 / +92 (+7%) / #102 (early 4th)

#10-12 (7) / 1243 / 1170-1500+ / 1320 / +77 (+6%) / #109 (mid-4th)

#11-13 (9) / 1189 / 1044-1420 / 1210 / +21 (+2%) / #178 (mid 6th)

#12-14 (10) / 1150 / 1041-1420 / 1167 / +17 (+1%) / #188 (late 6th)

#13-15 (9) / 1111 / 904-1293 / 1093 / -18 (-2%) / #186 (late 6th)

#14-16 (8) / 1050 / 824-1124 / 1012 / -38 (-4%) / #136 (early 5th)

#15-17 (10) / 985 / 824-1058 / 965 / -20 (-2%) / #181 (late 6th)

#16-18 (11) / 950 / 824-1058 / 948 / -2 (0%) / ("8th" rounder)

#17-19 (14) / 907 / 814-1270+ / 964 / 57 (6%) / #118 (late 4th)
[NOTE: Two trades involving #19 were clearly trades up from the 2nd by teams that didn't have a 1st and that included a "next 1st".]

#18-20 (12) / 875 / 814-1270+ / 935 / 60 (7%) / #117 (late 4th)

#19-21 (11) / 855 / 710-1270+ / 916 / 61 (7%) / #117 (late 4th)

#20-22 (11) / 803 / 710-1130+ / 884 / 81 (10%) / #106 (early 4th)

#21-23 (12) / 777 / 710-1130+ / 868 / 91 (12%) / #102 (early 4th)

#22-24 (12) / 767 / 719-1130+ / 859 / 92 (12%) / #102 (early 4th)

#23-25 (9) / 742 / 719-939+ / 801 / 59 (8%) / #117 (late 4th)

#24-26 (8) / 718 / 705-844+ / 757 / 38 (5%) / #136 (early 5th)

#25-27 (8) / 706 / 705-844+ / 779 / 73 (10%) / #110 (mid-4th)

#26-28 (8) / 680 / 645-860+ / 736 / 56 (8%) / #119 (late 4th)

#27-29 (8) / 658 / 645-860+ / 759 / 101 (15%) / #100 (early 4th)

#28-30 (9) / 640 / 608-812+ / 719 / 79 (12%) / #107 (early 4th)

#31 (600) [no trades]
#32 (590) [no trades]

There were 84 total trade down transactions involving picks #3-#30. Of those, nine were of the anomalous "Ditka Debacle" variety, IOW way over 20% variation from SCV (as much as 45%). The remaining 75 transactions averaged within 7% of SCV.

Of the 65 transactions involving picks #11-30, 65% were within 10% of SCV and 51% were with 5% of SCV.
=========================================

#33-35 (5) / 570 / 594-690 / 656 / +86 (15%) / #104 (early 4th)

And, futhermore, WHEW!

I'll try to do more with the 2nd round tomorrow, but I'm not promising anything.

Tremendous work, Brother Maine!! :rocker:

And bless you for using the correct term ~ "Rolling" Average ~ I thought I was the only one in the GALAXY who used that...I need to get out!!

 

[convertedpatsfan]

Thanks MaineMan. That's awesome work.

It's interesting how the late first round is where teams are overpaying more compared to the draft chart. The total values are lower, but overall pretty consistently 10 to 12% more.

And the early 2nd round work you've already done shows that #33 may indeed be a valuable pick.

Odd that no trades were made for #31 or #32. Guess everyone is tired by that point.

 

[MaineMan]

Thanks MaineMan. That's awesome work.

It's interesting how the late first round is where teams are overpaying more compared to the draft chart. The total values are lower, but overall pretty consistently 10 to 12% more.

And the early 2nd round work you've already done shows that #33 may indeed be a valuable pick.

Odd that no trades were made for #31 or #32. Guess everyone is tired by that point.

You're welcome.

Yeah, that there appear to be certain ranges with more trades than others or with more "overpayment" than others is no surprise in itself. But where those ranges appear is a bit off from what I might have expected. Not a lot, just shifted a bit. Interesting to speculate about the "why" in trend terms, but such a small data set might be "tainted" if it turned out that several of those trades were made by Al Davis, Millen, or other similar crazies.

 

[Off The Grid]

- Trading down out of the #1 and #3 has historically involved giving a steep discount, as much as 35%.

Sister Patricia actually expected this, but I must confess that this by FAR the most surprising revelation, to me.

It makes sense, too: The "Book Value" on #1-#3 IS pretty lofty.

Great stuff, man.