A Realistic Model for Draft Pick Value in Madden

[DCEBB2001]

One of the areas where the Madden video game series can improve upon their existing CFM model is within the draft pick trade logic. Often users are compelled to ask their peers about advice for "realistic" draft trades that would mimic the real-life value placed upon draft picks.

For pick-to-pick transactions (trades), most NFL teams still refer to the draft pick trade chart that correlates an estimated value for each pick to the pick number. For those unfamiliar with this chart, you can find a version of it here:

http://profootballtalk.nbcsports.com...t-trade-chart/

One of the more nebulous interactions occurs when attempting to trade draft picks for players. All too often one can find it too easy to acquire starter-caliber talent via trading away draft picks, yet, many NFL teams tend to hang onto those picks as if they were worth billions in revenue. Most notably, the Green Bay Packers' GM Ted Thompson has been notorious for refusing to part with draft picks for players since taking over the position in 2005. He values his draft picks so much, that he refused to trade a 4th round pick for Randy Moss prior to his record-setting season in 2007.

In an effort to find the true value of draft picks in the NFL, and create a model that could be incorporated into Madden, I have analyzed over 118 pick-for-player trades going back to 2000 in the NFL. Using FBG Ratings historical scouting data, I was able to assign a value to each player at the time of their trade, to the day they were traded. These values were set equal to the amount of draft pick points used to acquire the player.

Trades involving multiple players were not admitted into this study as it limits the ability to isolate the value of each individual player and compare it to the number of draft pick points needed to make the acquisition. It is assumed that the trade was of equal value for each team.

Once the player grades were mapped to the draft pick value points needed to acquire the player, an exponential regression was applied to the values. This regression yielded an accurate model that represents how many draft pick points are needed to acquire a player of a particular value. Using intervals of one-tenth of a grade point for the player overall values, you find the following chart that correlates player value to draft pick points. Also included is a current NFL player graded at the same grade level from this past July along with a possible draft picks that would be needed to acquire the player. This list is not to say that a team WILL trade these picks for the player, but that an equal value of player for picks is indicated below:

Grade Points Player Picks
10.00 3181.4 JJ.Watt #1, #81
9.40 1456.6 A.Rodgers #7
8.80 876.1 D.Bryant #18
8.50 709.4 A.Brown #25
8.30 622.1 D.Thomas #29
8.10 548.6 H.Ngata #35
7.90 485.8 N.Suh #41
7.70 431.6 A.Jeffrey #46
7.50 384.2 J.Thomas #51
7.30 342.4 M.Stafford #55
7.10 305.3 KJ.Wright #59
6.90 272.1 C.Kaepernick #63
6.70 242.2 E.Manning #69
6.50 215.1 A.Ellington #75
6.30 190.4 O.Scandrick #80
6.10 167.8 E.Ansah #84
5.90 147.1 B.Quick #88
5.70 127.9 D.Butler #93
5.50 110.2 K.Golston #97
5.30 93.7 M.Rivera #101
5.10 78.4 O.Bolden #107
4.90 64.0 C.Gray #115
4.70 50.6 N.Robey #121
4.50 38.0 P.Hunt #136
4.30 26.1 S.Mason #163
4.10 14.8 A.Bonner #191
3.90 4.2 D.Lynch #218

Based upon this valuation, elite players would require at least high 1st round picks. Perennial all-pros would also require at least a late 1st/early 2nd round pick to acquire. Solid starters could be had for a 2nd or 3rd round pick, and solid backups could be traded for 4th and 5th rounders. It is usually found that developmental players are acquired for late round picks.

What was newly found to me was how much value the draft picks had for all-pro-type players. I would have figured that it would take at least a mid-first-round pick to acquire them, but players like Ngata and Suh could be had for a 2nd round pick. That may indicate that draft pick value outweighs current player value more than one would anticipate. The fact that a top 100 player (like Eli Manning) could be had for only a 3rd round pick is surprising.

In conclusion, a model similar to this one could be developed and incorporated into the Madden video game series to better enhance the trade logic, especially regarding draft picks. This could better replicate the true value of draft picks as they are compared against OVR values of the current players in the game.

[NateDogPack12]

If someone called up the Green Bay Packers and offered merely the seventh pick in the draft for Aaron Rodgers, Ted Thompson would get a great chuckle and then hang up the phone. For the most part, this approximate value calculation plays out exactly as I would expect, but elite Quarterbacks are another matter entirely.

[Ueauvan]

i also believe the draft trade chart is out of date, iirc it was created to figure out the financial implications of a trade before rookie max % age of total cap and fixed rookie salaries came into being. realistically it needs to be updated and refer to the rookie fixed salary scale. all of which is irrelevant, if you have a need for qb and theres one in the draft, you will do what you can to get that pick or not trade if you have the pick

[DCEBB2001]

Quote: Originally Posted by NateDogPack12

If someone called up the Green Bay Packers and offered merely the seventh pick in the draft for Aaron Rodgers, Ted Thompson would get a great chuckle and then hang up the phone. For the most part, this approximate value calculation plays out exactly as I would expect, but elite Quarterbacks are another matter entirely.

Right, and I made it quite clear that these are merely what the equal value would be based upon 15 years of historical data of player grades vs. pick value to acquire the player. It has nothing to do with the asking price. There are several examples from the sample data where the actual price it took to acquire the player lied above the line for which the draft point value was set.

In 2013, Darrelle Revis had a grade of 8.42 and should have commanded an equal value of 672 draft points; basically a #27 overall pick. However, what he went for was a #13 and #104, a total of 1236 draft points. This illustrates that his "market value" was set higher than his "true trade value" if you will. The same happened for Carson Palmer in 2011 where his 6.01 grade was far surpassed by by a value of 1480 draft points whereas he should have only been set equal to 158 points, or #86 overall.

So once again, this doesn't really have anything to do with the asking value, but rather, is based upon the true value of the player over the course of a sample of 118 trades. The study does have its limitations, as I pointed out in the original post.

[DCEBB2001]

Quote: Originally Posted by Ueauvan

if you have a need for qb and theres one in the draft, you will do what you can to get that pick or not trade if you have the pick

This is more of a preference. Some GMs will pick for need, while others will pick for value. It just depends on your GM's preference.

[jhigginsluckow]

This is tough for me to give an opinion on because in my opinion, 90%+ of people don't understand quite how valuable draft picks are. It's not merely about the quality of the player you draft, but also their low cap number. With a low cap number, not only do you get that potentially talented player, but you can also sign another quality player with the extra cap money you saved.

Very simply, take the Dez Bryant value. It says he should be worth the #18 pick in the draft. The 18th pick in the draft makes what, probably $5m a year in salary on average? Dez just signed an extension giving him $14m a year (I believe). So while I think most people will rush to judgment and say no way would Dallas trade Dez for only the 18th pick, you have to think about it a little deeper. It's not just Dez Bryant for the 18th pick, it's Dez Bryant for the 18th pick AND $9m in cap space you can use toward someone else. Torrey Smith just signed with San Francisco recently at about $8m a year. Torrey Smith + the 18th overall pick for Dez Bryant suddenly doesn't sound so ridiculous.

It's the very premise Chip Kelly made the LeSean McCoy trade on. It wasn't just McCoy for Kiko Alonso - it was McCoy for Alonso plus a ton of Cap savings (which he blew on crap players, but that's besides the point).

HOWEVER, I do think that the sniff test needs to be done here. The idea that JJ Watt is worth more than Aaron Rodgers, for example, is absurd. Quarterbacks need to have a scale made for them all on their own if this is to work, in my opinion.

[DCEBB2001]

Quote: Originally Posted by jhigginsluckow

This is tough for me to give an opinion on because in my opinion, 90%+ of people don't understand quite how valuable draft picks are. It's not merely about the quality of the player you draft, but also their low cap number. With a low cap number, not only do you get that potentially talented player, but you can also sign another quality player with the extra cap money you saved. Very simply, take the Dez Bryant value. It says he should be worth the #18 pick in the draft. The 18th pick in the draft makes what, probably $5m a year in salary on average? Dez just signed an extension giving him $14m a year (I believe). So while I think most people will rush to judgment and say no way would Dallas trade Dez for only the 18th pick, you have to think about it a little deeper. It's not just Dez Bryant for the 18th pick, it's Dez Bryant for the 18th pick AND $9m in cap space you can use toward someone else. Torrey Smith just signed with San Francisco recently at about $8m a year. Torrey Smith + the 18th overall pick for Dez Bryant suddenly doesn't sound so ridiculous. It's the very premise Chip Kelly made the LeSean McCoy trade on. It wasn't just McCoy for Kiko Alonso - it was McCoy for Alonso plus a ton of Cap savings (which he blew on crap players, but that's besides the point). HOWEVER, I do think that the sniff test needs to be done here. The idea that JJ Watt is worth more than Aaron Rodgers, for example, is absurd. Quarterbacks need to have a scale made for them all on their own if this is to work, in my opinion.

I don't think the the word "absurd" is really the right way to describe it. I think that within the limitations of the test, it works. Once again, this wasn't done down to the positional level, as I wouldn't even have enough data points to conduct a real SRS (30+ samples) for each position group. We can see by looking at some of the individual grades that most QBs went for far fewer draft points than anticipated by the model.

Let's look at all of the QBs from this sample data set:

1. Matt Schaub was traded for a mere 19 points when he had a grade of 7.35. In this model, he should have been traded for 352 points. He was undervalued.

2. Alex Smith had a grade of 6.25 and was traded for 900 draft points. His value indicated that he should have been traded for 185 points. He was overvalued.

3. Carson Palmer (2011) had a grade of 6.01 and was traded for 1480 points when he should have been traded for 158 points. He was overvalued.

4. Matt Cassel (2015) had a grade of 5.95 and was traded for 34 points. He should have been traded for 152 points. He was undervalued.

5. Carson Palmer (2013) had a grade of 5.81 and was traded for 17.3 points. He should have been traded for 138 points. He was undervalued.

6. Colt McCoy had a grade of 5.53 and was traded for 4.7 points. He should have been traded for 112 points. He was undervalued.

7. Blaine Gabbert had a grade of 5.51 and was traded for 9.4 points. He should have been traded for 111 points. He was undervalued.

8. Matt Flynn had a grade of 5.38 and was traded for 41.2 points. He should have been traded for 100 points. He was undervalued.

9. Case Keenum had a grade of 5.31 and was traded for 8.2 points. He should have been traded for 94 points. He was undervalued.

10. Ryan Mallett had a grade of 5.30 and was traded for 8.2 points. He should have been traded for 93 points. He was undervalued.

11. Ryan Fitzpatrick had a grade of 5.21 and was traded for 8.2 points. He should have been traded for 86 points. He was undervalued.

12. Matt Barkley had a grade of 5.04 and was traded for 8.2 points. He should have been traded for 73 points. He was undervalued.

13. Matt Cassel (2015 again) had a grade of 5.04 and was traded for 29.5 points. He should have been traded for 73 points. He was undervalued.

14. Joey Harrington had a grade of 4.87 and was traded for 21 points. He should have been traded for 61 points. He was undervalued.

15. Matt Schaub (2007) had a grade of 4.58 and was traded for 1030 points. He should have been traded for 42 points. He was overvalued.

16. Drew Stanton had a grade of 4.36 and was traded for 18.6 points. He should have been traded for 29 points. He was undervalued.

17. Cody Pickett had a grade of 3.96 and was traded for 8.2 points. He should have been traded for 7 points. He was overvalued.



Of the 17 QBs traded for picks only since 2000, only 4 were acquired for more points than they were worth based upon their grade. The other 13 QBs were undervalued compared to the rational function line. Over-evaluation occurred at both the low end and high end of the grading scale. To this point, it appears that QBs were actually undervalued far more often than they were overvalued. Another scale could be made for QBs altogether, but I would need a larger sample size to do it. Perhaps I can revisit that in a few more years when more QBs are traded.

[RogueHominid]

Love this data.

It'd be great to have Madden's trade logic for players and picks governed by data gleaned from historical analysis.

Obviously, once you enter your CFM, you're entering a self-contained, dynamic universe, but I'd love it if the general rules governing trading had some 10-15 year data from longitudinal studies behind it.

Maybe it does already, IDK, but this data is great as a conversation starter.