Recruiting Experiment

Discussion in 'The Experience' started by Nautica, May 31, 2011.

  1. Nautica

    Nautica #whatisdrugs?

    Joined:
    Jul 27, 2009
    Messages:
    7,915
    I am trying out Edge's recruiting method, I did one year on Heisman with ND and the numbers are pretty decent. I think it being on Heisman may be a bit much, if we did it on AA I think it would work perfectly as I REALLY got screwed on recruiting on Heisman. Heisman would work if we were to do it the regular way, but I think I would like it if we did this way.

    I changed it a little to where you had to split the number of allowed schollies in the middle, with the 4 and 5 star rounded down, making that the max # of recruits that you could sign being that of that caliber. If you failed to fill all allowable schollies for that caliber, I allowed myself to fill it with 3 star and below recruits.

    So here is how year 1 went:

    I started with ND and had 16 allowable schollies. 8 max for :4stars:and:5stars: star recruits with the max amount of :5stars: being 4(half), and filling the rest of the allowed schollarships with :3stars:recruits up to 16 total.

    I was only able to sign 11 total out of the 16 that I needed to have a full 70 on the roster.

    ND #79th recruiting class consisted of 5:4stars:and 6:3stars:recruits. I was #1 on 3 :5stars: recruits in pipeline states, but because of how much of a bitch it is to recruit on Heisman with this, it only took one bad recruiting week to put me far behind on these guys.

    Here is how the top 10 looked at season's end

    Team-5-4-3-2-1

    1. USC 7-7-4-0-0
    2. UT 6-7-5-0-1
    3. UM 3-10-7-0-0
    4. OU 2-10-8-0-0
    5. UF 5-4-5-0-3
    6. PSU 2-10-4-0-2
    7. Bama 4-4-8-0-0
    8. TCU 0-7-17-1-0
    9. OSU 3-6-6-0-3
    10. NEB 0-11-8-1-0

    Now if I were to carry this on for the next 5 to 6 seasons, the CPU teams would dominate. I do not like Heisman setting for this at all.

    I'll restart my file and do the same thing, but on AA recruiting and then post the results.
     
  2. Nautica

    Nautica #whatisdrugs?

    Joined:
    Jul 27, 2009
    Messages:
    7,915
    Alright, I went and simulated through season 1 again. Only this time I put it on AA recruiting instead of Heisman. Here were my results:

    Notre Dame ended up with the 21st ranked recruiting class. I started out with 16 total scholarships at the start of the season, and by the end I had 14 players signed. I had 1 player transfer out, but had 1 incoming transfer, which kept my allowed scholarships @ 16. I used the same method as before, and here is how it came out.

    :5stars:=2
    :4stars:=6
    :3stars:=6

    Here is how the top 10 looked:
    (team-5-4-3-2-1)
    1. UF 9-5-6-0-0
    2. LSU 7-6-5-0-0
    3. USC 4-9-4-0-0
    4. OSU 6-4-3-0-6
    5. FSU 1-10-7-3-0
    6. V-Tech 3-6-7-2-0
    7. OU 3-9-1-0-3
    8. UGA 1-9-9-1-0
    9. Tenn 0-9-12-2-0
    10. Bama 3-4-11-0-0
    I'll post season 2, 3, 4, and 5 as I go on, and in season 5 I'll post OVR from 20 different teams so we can see how teams progressed through 4 seasons.
     
  3. Nautica

    Nautica #whatisdrugs?

    Joined:
    Jul 27, 2009
    Messages:
    7,915
    I really like this method so far, but I did modify it slightly.

    Here is what I am suggesting we use(mind you I'm still testing it out further).

    # of Scholarships(SR+(70-# on roster))/2(if odd number, round up)=# of recruits ≤ 3*
    # of Scholarships-# of ≤ 3*= # of recruits ≥ 4*
    # of recruits ≥ 4*/2(rounded down)= Max # of 5* recruits

    Because of the mass restrictions of this formula, I suggest that scholarships left remaining going into the offseason are free for guys to pursue whatever recruit they wish due to the fact that signing recruits in the offseason is extremely difficult. That being said, we should make a rule though that if a person has filled their max # of ≥ 4* recruits and they still have open scholarships that they cannot sign more than 2 ≥ 4* recruits over their limit in the offseason.

    Another thing I would suggest is that we do away with the swaying back restrictions. This would severely hurt lower prestige programs as they are trying to build, plus it is predetermined whether or not a player is going to leave. The amount of time you use to talk a player back has no bearing on whether or not he stays.

    So what do you guys think so far?
     
  4. dakota7

    dakota7 Former Blue Chip Recruit

    Joined:
    Sep 23, 2009
    Messages:
    12,443
    So you want to be able to sway everyone into staying? Does the cpu sway their players back?

    I would add one layer to this, how about a couple of freebies? A few random "bonus" guys that don't count toward your total? Maybe numbers 1, 50, 100, 150, 200? The OFFL had these. It would create more competition between users and force even more strategic recruiting.
     
    • Like Like x 1
  5. dakota7

    dakota7 Former Blue Chip Recruit

    Joined:
    Sep 23, 2009
    Messages:
    12,443
    Otherwise, I like it.
     
  6. jello1717

    jello1717 "Those who stay will be champions." -- Bo

    Joined:
    Jun 21, 2009
    Messages:
    13,617
    Assuming that you guys have kept your recruiting results from the year, a better experiment (or maybe just an additional one) would be to look @ your results and apply the rules to them and see how many extras you got. IE. Florida had 12 schollies and signed 5 :5stars:, 6 :4stars: and 1 :3stars: you know that he signed 2 too many :5stars: and 3 too many :4stars: (I think that's what'd be with the formula). You can then add them up for all 12 teams and see how many extra guys you 12 kept from the CPU teams. It'll be tough to see what the CPU teams would look like however. The problem with doing it your way is that you're only 1 guy and you can't really see how the changes would affect the CPU teams (which is the entire goal I think) with only 1 user instead of 12 users.

    Of course, the ultimate test would be to start a new 12 user dynasty and use these rules for 4 or so seasons, but that's a ton of work.

    As a general observation, this formula makes half of our recruits :3stars: and below which sounds really low for most of the teams we're looking at. Also, not only does this increase the parity between user and CPU, but it'll also increase the parity between user teams (I'm not sure if that's the intention or not). By forcing Oregon, USC, LSU, Alabama, UofM, PSU to fill half of their roster with :3stars: this will bring them down to the level of Arizona, SCar, and others in a few years.
     
  7. GoGators

    GoGators GT: KSherm

    Joined:
    May 19, 2009
    Messages:
    7,893
    The last part is the thing I hate about recruiting restrictions like this. They create parity between users which isn't right. Take the example that Dakota's Wisconsin team becomes the best in the dynasty while the rest of us regress to 4 and 5 star schools. Why should his recruiting, which is a consequence of his success on the field, get capped? Limits like this only affect the best teams. It's essentially creating a handicap and I'm not a fan to be honest.
     
    • Like Like x 1
  8. blLL flo

    blLL flo BTFU!

    Joined:
    Aug 13, 2009
    Messages:
    2,655
    This is what concerns me with this formula. I do not want half of our teams to be filled with :3stars:. If this becomes a rule, it is going to become a race for all the JUCO guys.

    All this clamoring for the CPU to be more competitive, but when we've chosen 12 top 25 teams then obviously the we are gonna be the best when the dust settles after a few seasons.

    And finally, EA said they made changes to the AI recruiting logic in '12, so all this testing and rule making may not even matter come 7/12.
     
  9. Nautica

    Nautica #whatisdrugs?

    Joined:
    Jul 27, 2009
    Messages:
    7,915
    Since last night I have been thinking of ways to tune it, like 2/3 of your allowable scholarships being allowed for 4* or greater recruits. Would that be more of a middle ground that everyone can agree on?

    See the problem is that the recruiting was supposed to be more difficult this past title release, and at first it was, but we all got the hang of it pretty quickly and 4-5 seasons in guys start complaining because the CPU is irrelevant because the users are eating up all the recruits. I thought I would try out edge's method and we could tweak it.

    Your right about the JUCO thing bill, I noticed myself going straight for JUCO's with my 3* scholarships, but a lot of the time those 3 star guys are mainly on your roster as "filler" and by the time they actually see some playing time, they are in the 80's and will never start. In RMG when I was playing as Rutgers, most of those non-juco 3* guys by the time they ended up as a JR or SR, they were around the mid 80 mark, and a couple ended up in the low 90's. We shouldn't be afraid of the 3 star recruits, as some of them that are not juco's do come in rated decently.

    But yes, come '12 all of this may not even matter.
     
  10. Nautica

    Nautica #whatisdrugs?

    Joined:
    Jul 27, 2009
    Messages:
    7,915
    What would be a more acceptable method to try?

    2/3 rounded up or 3/4 rounded up?

    With 16 scholarships here is how it would look:

    2/3 method

    11 4* or greater recruits with 5-:5stars: max
    5-:3stars:or lower

    3/4 method

    12 4* or greater recruits with 6:5stars: max
    4-:3stars:or lower

    Obviously the more recruits you need the larger the amount of 5 and 6 star recruits will be allowed.

    I'm really just spit-balling here
     
  11. Hova

    Hova Live Action. YiYiYi!

    Joined:
    Feb 21, 2009
    Messages:
    12,136
    If you allow even 4 5 stars PER user team the CPU is garbage in 4 years, almost all of them except 3 teams. Just do the math, if users are getting 90% of the 5 stars its not going to be pretty for the CPU.
     
  12. JCspartan2

    JCspartan2 Sparty Party

    Joined:
    Feb 21, 2009
    Messages:
    4,967
    Bingo.
     
  13. edge7771

    edge7771 Walk On

    Joined:
    Apr 27, 2011
    Messages:
    1,944
    I know most are gonna be against it everyone was in our league to. But in real life just about every team in college footbal is made up of half of 3* are less players. The issue we were running into is people having 14-17 kids and they all were 4/5 start kids. When you have 12 schools taking on avg 15 4/5* kids that's leaves only 70 left for the other 100 and so D1 schools since on avg there are around 250 4/5* kids a year. I'm for what ever, but it made our league so much better an more balanced. Here is last year's rival recruiting rankings.

    total kids 5* 4* 3* avg*




    #1

    [​IMG]Alabama

    22

    3

    14

    5

    3.91

    2,632


    #2

    [​IMG]Florida State

    29

    2

    13

    13

    3.55

    2,365


    #3

    [​IMG]Texas

    22

    1

    15

    6

    3.77

    2,322


    #4

    [​IMG]USC

    30

    1

    14

    13

    3.47

    2,315


    #5

    [​IMG]Georgia

    26

    2

    12

    11

    3.58

    2,309


    #6

    [​IMG]LSU

    22

    3

    9

    9

    3.64

    2,224


    #7

    [​IMG]Auburn

    24

    1

    13

    10

    3.63

    2,194


    #8

    [​IMG]Clemson

    29

    4

    6

    15

    3.34

    2,173


    #9

    [​IMG]Oregon

    23

    2

    9

    12

    3.57

    2,048


    #10

    [​IMG]Notre Dame

    23

    2

    8

    12

    3.48

    1,976


    #11

    [​IMG]Ohio State

    23

    1

    9

    12

    3.43

    1,927


    #12

    [​IMG]Florida

    19

    0

    11

    8

    3.58

    1,851


    #13

    [​IMG]Tennessee

    27

    0

    12

    14

    3.41

    1,811


    #14

    [​IMG]Oklahoma

    17

    2

    7

    8

    3.65

    1,722


    #15

    [​IMG]Nebraska

    19

    0

    11

    7

    3.53

    1,698


    #16

    [​IMG]North Carolina

    25

    0

    9

    14

    3.28

    1,661


    #17

    [​IMG]California

    22

    0

    10

    12

    3.45

    1,646


    #18

    [​IMG]South Carolina

    32

    1

    6

    25

    3.25

    1,638


    #19

    [​IMG]Mississippi

    27

    1

    5

    21

    3.26

    1,573


    #20

    [​IMG]Texas Tech

    27

    0

    6

    19

    3.15

    1,457


    #21

    [​IMG]Michigan

    20

    0

    6

    13

    3.25

    1,314


    #22

    [​IMG]Stanford

    19

    0

    7

    11

    3.32

    1,311


    #23

    [​IMG]Washington

    23

    0

    5

    15

    3.09

    1,300


    #24

    [​IMG]Arkansas

    30

    0

    4

    24

    3.07

    1,189


    #25

    [​IMG]Virginia

    26

    0

    5

    17

    3.04

    1,144


    #26

    [​IMG]TCU

    25

    0

    2

    23

    3.08

    1,122


    #27

    [​IMG]Texas A&M

    22

    0

    3

    15

    2.95

    1,098


    #28

    [​IMG]Oklahoma State

    27

    0

    2

    22

    2.96

    1,096


    #29

    [​IMG]Louisville

    20

    0

    4

    15

    3.15

    1,086


    #30

    [​IMG]Iowa

    24

    0

    4

    16

    3.00

    1,065


    #31

    [​IMG]Michigan State

    21

    0

    2

    17

    3.00

    1,028


    #32

    [​IMG]Rutgers

    24

    0

    4

    15

    2.96

    1,014


    #33

    [​IMG]Virginia Tech

    21

    0

    4

    14

    3.05

    928


    #34

    [​IMG]Kansas

    27

    0

    2

    21

    2.93

    899


    #35

    [​IMG]Penn State

    16

    0

    5

    8

    3.13

    890


    #36

    [​IMG]Miami (FL)

    16

    0

    2

    13

    3.06

    855


    #37

    [​IMG]Utah

    19

    0

    2

    14

    2.95

    853


    #38

    [​IMG]Boston College

    23

    0

    4

    9

    2.74

    834


    #39

    [​IMG]UCF

    29

    0

    3

    19

    2.86

    778


    #40

    [​IMG]Wisconsin

    20

    0

    3

    15

    3.05

    767


    #41

    [​IMG]Georgia Tech

    22

    0

    1

    18

    2.91

    759


    #42

    [​IMG]Illinois

    27

    0

    1

    21

    2.85

    711


    #43

    [​IMG]Maryland

    21

    0

    1

    15

    2.81

    692


    #44

    [​IMG]Mississippi State

    22

    0

    1

    21

    3.05

    673


    #45

    [​IMG]UCLA

    16

    0

    3

    10

    3.00

    636


    #46

    [​IMG]Baylor

    19

    0

    1

    16

    2.95

    628


    #47

    [​IMG]West Virginia

    22

    0

    3

    15

    2.95

    624


    #48

    [​IMG]Missouri

    17

    0

    1

    16

    3.06

    607


    #49

    [​IMG]Cincinnati

    24

    0

    1

    20

    2.92

    605


    #50

    [​IMG]Southern Methodist

    27

    0

    1

    20

    2.81

    504
     
  14. GoGators

    GoGators GT: KSherm

    Joined:
    May 19, 2009
    Messages:
    7,893
    Edge makes some good points and I guess we may need these for next year. It's pretty evident how jacked the recruiting is when you look at rivals and top schools only get 2 or 3 five stars in a great year. Yet ea has it where a CPU controlled Florida can grab 9. There's such a disconnect between some of their programming and reality and I have no idea how.
     
  15. JCspartan2

    JCspartan2 Sparty Party

    Joined:
    Feb 21, 2009
    Messages:
    4,967
    Here is the issue:

    Sure a lot of schools make up their classes with 3* recruits, but some of those guys become great players in the future. See Cousins or Javon Ringer for example. In the game, this doesn't happen often. I think I had 1 player rated 3 stars or lower that wasn't a JUCO turn out to be phenomenal. In the game, if you don't get 4 star recruits, your team ratings are going to drop off the side of the earth.
     
  16. GoGators

    GoGators GT: KSherm

    Joined:
    May 19, 2009
    Messages:
    7,893
    This is one thing they said they fixed though. Incoming recruit ratings will be more in line with original roster players. I assume this means that 3 stars will at least be able to see the field by the time they are upper classmen.
     
  17. JCspartan2

    JCspartan2 Sparty Party

    Joined:
    Feb 21, 2009
    Messages:
    4,967
    True, they also said they adjusted the cpu recruiting AI, so we may not have to make adjustments at all.
     
  18. dakota7

    dakota7 Former Blue Chip Recruit

    Joined:
    Sep 23, 2009
    Messages:
    12,443
    If we want the CPU to be more competitive after year 3, there are only a couple of options. Take crappy teams and build up, use crazy hard (and probably unrealistic) sliders or attempt to keep the talent levels more balanced with recruiting restrictions and/or reduced ability to sway players back.
     
  19. jello1717

    jello1717 "Those who stay will be champions." -- Bo

    Joined:
    Jun 21, 2009
    Messages:
    13,617
    Making Edge's table more legilbe:

    Rank School Total :5stars: :4stars: :3stars: avg ??
    #1 Alabama 22 3 14 5 3.91 2,632
    #2 Florida State 29 2 13 13 3.55 2,365
    #3 Texas 22 1 15 6 3.77 2,322
    #4 USC 30 1 14 13 3.47 2,315
    #5 Georgia 26 2 12 11 3.58 2,309
    #6 LSU 22 3 9 9 3.64 2,224
    #7 Auburn 24 1 13 10 3.63 2,194
    #8 Clemson 29 4 6 15 3.34 2,173
    #9 Oregon 23 2 9 12 3.57 2,048
    #10 Notre Dame 23 2 8 12 3.48 1,976
    #11 Ohio State 23 1 9 12 3.43 1,927
    #12 Florida 19 0 11 8 3.58 1,851
    #13 Tennessee 27 0 12 14 3.41 1,811
    #14 Oklahoma 17 2 7 8 3.65 1,722
    #15 Nebraska 19 0 11 7 3.53 1,698
    #16 North Carolina 25 0 9 14 3.28 1,661
    #17 California 22 0 10 12 3.45 1,646
    #18 South Carolina 32 1 6 25 3.25 1,638
    #19 Mississippi 27 1 5 21 3.26 1,573
    #20 Texas Tech 27 0 6 19 3.15 1,457
    #21 Michigan 20 0 6 13 3.25 1,314
    #22 Stanford 19 0 7 11 3.32 1,311
    #23 Washington 23 0 5 15 3.09 1,300
    #24 Arkansas 30 0 4 24 3.07 1,189
    #25 Virginia 26 0 5 17 3.04 1,144
    #26 TCU 25 0 2 23 3.08 1,122
    #27 Texas A&M 22 0 3 15 2.95 1,098
    #28 Oklahoma State 27 0 2 22 2.96 1,096
    #29 Louisville 20 0 4 15 3.15 1,086
    #30 Iowa 24 0 4 16 3.00 1,065
    #31 Michigan State 21 0 2 17 3.00 1,028
    #32 Rutgers 24 0 4 15 2.96 1,014
    #33 Virginia Tech 21 0 4 14 3.05 928
    #34 Kansas 27 0 2 21 2.93 899
    #35 Penn State 16 0 5 8 3.13 890
    #36 Miami (FL) 16 0 2 13 3.06 855
    #37 Utah 19 0 2 14 2.95 853
    #38 Boston College 23 0 4 9 2.74 834
    #39 UCF 29 0 3 19 2.86 778
    #40 Wisconsin 20 0 3 15 3.05 767
    #41 Georgia Tech 22 0 1 18 2.91 759
    #42 Illinois 27 0 1 21 2.85 711
    #43 Maryland 21 0 1 15 2.81 692
    #44 Mississippi State 22 0 1 21 3.05 673
    #45 UCLA 16 0 3 10 3.00 636
    #46 Baylor 19 0 1 16 2.95 628
    #47 West Virginia 22 0 3 15 2.95 624
    #48 Missouri 17 0 1 16 3.06 607
    #49 Cincinnati 24 0 1 20 2.92 605
    #50 Southern Methodist 27 0 1 20 2.81 504
     
  20. jello1717

    jello1717 "Those who stay will be champions." -- Bo

    Joined:
    Jun 21, 2009
    Messages:
    13,617
    The thing to remember when looking @ real life #s is that they're recruiting 85 players, not 70 like us. It should be safe to assume that if they could only sign 70 players, they'd have roughly 15 fewer :3stars: instead of fewer :5stars: and :4stars: so their :3stars: numbers are higher than ours would be.

    As Gators stated, reducing the number of studs we can sign won't increase the overall talent level of CPU teams as the Floridas, Texases, Oklahomas, and Notre Dames of the league will just horde the guys we don't sign. It'll make those teams stronger, but the non elite programs will still only have scraps to choose from.

    I've said it before, but I think that we shouldn't be able to sway anyone into staying. If they want to leave early, then we need to let them go. The logic in the game will have some guys leaving earlier than we think they should (which happens in real life) and there'll still be guys that have no reason whatsoever to return, but do anyway (just like what happens now). What I mean is that we won't have all of our studs bugging out early if we choose to not sway anyone back.
     
  21. jello1717

    jello1717 "Those who stay will be champions." -- Bo

    Joined:
    Jun 21, 2009
    Messages:
    13,617
    Added a couple more categories to this. The 1st is the weighted average (total *70 (EA)/85(NCAA)) with the assumption that they keep all :5stars: and :4stars: and just didn't sign as many :3stars:. The last group has the weighted average applied to all columns (total, :5stars: :4stars: and :3stars:). I also added a column for each grouping for the % of the signings that were :3stars:.

    I'm not drawing any conclusions from this, just adding more data to Edge's post (and playing around with google docs).

     
    • Like Like x 1
  22. edge7771

    edge7771 Walk On

    Joined:
    Apr 27, 2011
    Messages:
    1,944
    No because in real life each school only gets 25 scholarships a year so if you base it off 70 each school would get about 5 or 6 less scholarships. So in turn more kids are evenly placed because they want to play and you can't over recruit. Also you can't just stock pile talent and hoard 3 or 4 players at on position, you have to recruit your depth chart and do your homework. Most on here can crush the cpu no matter what, but when we went to this plan it changed. We didn't see five or six cpu programs just grab everyone they tended to spread out, and the cpu teams were much better across the board.
     
  23. jello1717

    jello1717 "Those who stay will be champions." -- Bo

    Joined:
    Jun 21, 2009
    Messages:
    13,617
    I find this surprising because in my offline dynasty (where there's only 1 user taking good players) Alabama, Florida, Oklahoma, Texas, Ohio State, USC, and Notre Dame consistently hoarded :5stars: and :4stars: players (they almost always got more than I did) and the rest of the schools just grabbed the scraps that those big CPU schools left behind.
     
  24. edge7771

    edge7771 Walk On

    Joined:
    Apr 27, 2011
    Messages:
    1,944
    I'm just speaking from our league and we all are big programs and we have 12 users, not sure about any other league. But honestly I don't think I've ever seen a cpu sign more than 3 5* and if they do sign 3 it's usually one school. Most of the 5* kids are spread out except when it comes to users. Now the 4* kids you'll 8 or 9 but that is the norm. You also have to remember your not going to be as strong so all those A+ won't be there. I know most don't like as we didn't it either, it's just an idea and everyone in our league really likes it and were on our 7th season.
     
  25. GoGators

    GoGators GT: KSherm

    Joined:
    May 19, 2009
    Messages:
    7,893
    I think there's a few problems with recruiting that cause the disconnect between users and CPU:

    1 - Users basically recruit 10 guys at a time using 10 hours/recruit since we typically only sign 15 guys due to our cap. The CPU on the other hand recruits to 25 and therefore they aren't investing the same amount of time and therefore only the CPU big dogs can compete, i.e. Ohio St, Texas, Florida, USC, etc.

    2 - The cpu's recruiting logic in the past was jacked. Look at Ohio St now, they have 3 :5stars: QBs on their roster while there's holes everywhere else. They were spending max time on two guys they didn't need. I know cause they out recruited me for them....bastids!

    3 - Redshirting - I put together some data earlier this year showing users redshirt around 65-70% of players while the CPU is only redshirting about 55% of their players. That 10% difference constitutes 7 guys per season and is almost an entire offense! So, if you take into consideration we are signing all the 5 stars and the top 4 stars since all users have top 20 classes, and then we redshirt all of them, the CPU has no chance at all! It's an effed system that we really need to change.

    4 - Players leaving - like Jello said, possibly going to a system where you can't sway players would be better but at the same time I'm not sure it helps a ton because that's just another recruit we're going to grab.

    5 - Users schedule all recruiting visits for rivalry games against CPU opponents they know they'll beat. The CPU unfortunately doesn't have that luxury. Tennessee probably had tons of recruits visit our games and since they could never beat me they were screwed. I think creating custom rivalries where users play users will help this a bit in `12. Then you need the risk of winning a user vs user game in order to gain the rivalry bonus.

    Honestly in `12, I think we need to start with recruiting on HEISMAN and modify it down to AA if we start getting killed on the recruiting trail. We CANNOT start on AA and move to HEISMAN because if we kill the CPU on AA, moving to HEISMAN will be too little too late.
     

Share This Page