Skills, Combine values and the Draft..

[Dandelion]

!! Attention - this is still a work in progress.. all data found here might be changed drastically, if further research brings up different results !!

There already is a very nice thread by MalcPow here about maximum/minimum combine ratings and their influence on the players attributes/skills.. I tried to get some conclusions from a different approach.

MalcPow set up a draft class with 1000 players, giving them skill values in steps from 375 to 625 and listed the resulting combine values.

I made up a draft class, set the height and weight values to the best according to position and then set all skills at values of 375, 500 and 625 (like 0%, 50% and 100%) to get a basic line for each of the three percentages, then set one skill after the other to 625 (with all others remaining at 375) to check which combine value would be influenced by it. Even for skills, that don't appear in a position's skill "set". Least I set all the skills belonging to a position's skill set* to 625 to get a final base line.

I took this different approach, because after taking a look at a draft class made through TCY I noticed several players having skills outside of their position's skill set and I wanted to see if (and how) those would influence the combine values.
I'm not sure if the draft classes made up within FOF2k7 are done the same way (= with skills not belonging to a position's skill set), but I *think* they are.. any further information would be greatly appreciated..

* Skill sets
Each position has a certain number of skills that are kinda unique to that position. All passing skills fall into the skill set for QBs for example. Several skills in most sets are available to more than one position. A RB's position skill set derives from some of the running and passing skill sets, for example.
Good news is, that usually only the skills from the position's skill set have influence on the combine values.
Bad news is, that there are three positions where this isn't the case... These are QB, FB and ILB/OLB

So for most positions, it doesn't matter if any skills outside of the positions skill set have any value assigned to them unless the player changes his position. At least, I presume (!) the basic skill values given in a draft file stay unchanged through a players career, meaning he would swap skill sets when he goes to a new position. This would explain the usually very low values when one moves a player from RB to CB - he usually just lacks the skills from the DB's skill set..

Combine Values
Some (newer..) player might make the same (probable) false assumption that I made for quite a while, that combine values decide about a player's skill values. The opposite is true - the higher a player's (potential) skill values (in the draft file) are, the better he will be at the combine. And as the combine values only derive from certain skills in each skill set, even great combine ratings are no indication for a player to be good at different positions.
Most positions have more than one skill that influences a certain combine value. In these cases, a really great combine value can only be reached with all appropriate skills being in a high range. A good combine value doesn't necessary give indication about which of the several skills is maxxed and which isn't. And with QBs, FBs and ILB/OLBs there might be even some unexpected influence on the combine values.
Then again some skills don't seem to have any influence at combine values at all.

I'll try to give examples of what I think to have found below.. Like MalcPow I would love to see it as a starting point for discussion - especially with input from those guys who have a lot more experience with FOF than I have.. There's a good chance, that my view of all this is off target in some (... or all..) points.

Influence ratings
I give four different ratings to the influence of a skill on a combine value. From bottom to top these are:
low - medium - large - maximum

There aren't any fixed values applyable to these ratings, as minimum and maximum values are different from skill to skill and within the skills from position to position..
As a rule of thumb I've set the values for the influence to the ranges for the skill values..
Let me use Benchpress for QBs as an example:
With all skill values at 375 Benchpress ranges from 6 to 7, at 500 the range is 10 to 13 and at 625 it's 12 to 17

Low influence would take Benchpress to something around 9, halfway between the 375 and 500 ranges. Medium would be the 500 range. Large would fall in between 500 and 625 and maximum would at last be the 625 range.
In numbers:
low ~ 8 to 9
medium ~ 10 to 12
large ~ 13 to 14
maximum ~ 15+

As the ranges for the different combine values aren't very strict, all these influence ratings aren't set in stone..
So take a look at the values MalcPow listed here and You should be able to get an idea about the numbers involved..

[Dandelion]

QB
Skills in skill set - influenced combine value (grade of influence):
Screen Passes - Agility (maximum)
Short Passes - none
Medium Passes - Broad Jump (maximum)
Long Passes - Benchpress (medium)
Deep Passes - Benchpress (medium)
3rd Down Passing - none
Scramble - 40 Time (maximum)
Avoid Interceptions - Solecismic (maximum)
Accuracy - Agility (maximum), Position Drill (medium)
Timing - Position Drill (medium)
Avoid Fumble - none

Skills from other skill sets with influence on combine values:
RB: Hole Recognition - 40 Time (maximum)

Usefull skill sets for possible position change (with no influence on combine values):
RB, FB, TE, WR, CB, S

Important paired skills:
Screen Passes/Accuracy - Agility
Long Passes/Deep Passes - Benchpress
Accuracy/Timing - Position Drill
Scramble/Hole Recognition - 40 Time

When You look for the values given in MalcPow's post, the maximum values can only be found on players with great values in BOTH paired skills.
Take Long Passes/Deep Passes for example.. both influence Benchpress. If either of the two skills is at maximum value, Benchpress will reach a value between 10 to 14. If both are maxxed out, Benchpress will reach 12 to 21.
From this result several possibilities:
Take an QB with Benchpress 13 at combine for example. With that value he falls into different ranges.
- a player with Long AND Deep Passes at 500 (50%) can have a high Benchpress of about 13.
- a player with either Long OR Deep Passes at 625 (100%) and the other at 375 (0%) would have a good chance for Benchpress 13.
- a player with either Long OR Deep Passes at 625 and the other at 500 could end up with a low 13 Benchpress value.

So Benchpress 13 could indicate Long/Deep Passing skills anywhere from 375 to 625 (0 to 100%). To be halfway sure BOTH skills are better then mediocre, Benchpress should at least be in the 15+ Range.

The combine values I got for a QB with all his skill set values at 625 are (max/min):
Solecismic - 48 / 28
40 Time - 4.32 / 4.84
Benchpress - 21 / 12
Agility - 6.93 / 7.80
Broad Jump - 119" / 108"
Position Drill - 98 / 81

[Dandelion]

RB
Skills in skill set - influenced combine value (grade of influence):
Breakaway Speed - 40 Time (medium)
Power Inside - Benchpress (maximum)
3rd Down Running - Agility (low), Broad Jump (medium)
Hole Recognition - Solecismic (maximum)
Elusiveness - Agility (low)
Avoid Drops - none
Get Downfield - Agility (low)
Route Running - Position Drill (large)
3rd Down Catch - none
Punt Return - none
Kick Return - none
Endurance - Broad Jump (medium)
Special Teams - none

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
FB, TE, WR, CB, S

Important paired skills:
3rd Down Running/Elusiveness/Get Downfield - Agility
3rd Down Running/Endurance - Broad Jump

The combine values I got for a RB with all his skill set values at 625 are (max/min):
Solecismic - 28 / 18
40 Time - 4.29 / 4.57
Benchpress - 26 / 17
Agility - 6.86 / 7.05
Broad Jump - 131" / 118"
Position Drill - 39 / 19

[Dandelion]

FB
Skills in skill set - influenced combine value (grade of influence):
Power Inside - Benchpress (maximum)
3rd Down Running - Agility (maximum), Broad Jump (medium)
Hole Recognition - Solecismic (maximum)
Avoid Drops - none
Route Running - Position Drill (maximum)
3rd Down Catch - none
Run Blocking - Broad Jump (medium)
Pass Blocking - none
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
RB: Elusiveness - Agility (maximum)
WR: Get Downfield - Agility (maximum)

Usefull skill sets for possible position change (with no influence on combine values):
RB, TE, WR, CB, S

Important paired skills:
3rd Down Running/Elusiveness/Get Downfield - Agility
3rd Down Running/Run Blocking - Broad Jump

The combine values I got for a FB with all his skill set values at 625 are (max/min):
Solecismic - 30 / 20
40 Time - 4.88 / 4.95
Benchpress - 33 / 22
Agility - 7.09 / 7.47 *
Broad Jump - 120" / 108"
Position Drill - 49 / 24

* with 3rd Down Running at 625 (100%) and Elusiveness and Get Downfield at 375 (0%): 7.63 / 7.96

[Dandelion]

TE
Skills in skill set - influenced combine value (grade of influence):
Avoid Drops - Position Drill (large)
Get Downfield - 40 Time (medium), Agility (maximum)
Route Running - Solecismic (maximum)
3rd Down Catch - Broad Jump (medium)
Courage - none
Big Play Receiving - none
Run Blocking - Broad Jump (low)
Pass Blocking - Benchpress (low)
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
RB, FB, WR, CB, S

Important paired skills:
3rd Down Catch/Run Blocking - Broad Jump

The combine values I got for a TE with all his skill set values at 625 are (max/min):
Solecismic - 32 / 21
40 Time - 4.45 / 4.71
Benchpress - 36 / 23
Agility - 7.25 / 7.56
Broad Jump - 125" / 108"
Position Drill - 53 / 36

The TE Position seems a bit weird, as the influence of the skills on the combine values seems lower than for other positions..

[Dandelion]

WR
Skills in skill set - influenced combine value (grade of influence):
Avoid Drops - Position Drill (large)
Get Downfield - 40 Time (low), Agility (large)
Route Running - Solecismic (large), Position Drill (low)
3rd Down Catch - none
Courage - Bench Press (maximum)
Big Play Receiving - 40 Time (medium)
Punt Return - none
Kick Return - none
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
RB, FB, TE, CB, S

Important paired skills:
Avoid Drop/Route Running - Position Drill
Get Downfield/Big Play Receiving - 40 Time

The combine values I got for a WR with all his skill set values at 625 are (max/min):
Solecismic - 31 / 18
40 Time - 4.25 / 4.45
Benchpress - 21 / 12
Agility - 6.73 / 7.05
Broad Jump - 131" / 102"
Position Drill - 69 / 49

[Dandelion]

C
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (medium)
Pass Blocking - Benchpress (medium), Agility (maximum)
Endurance - Broad Jump (maximum)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
G, T, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a C with all his skill set values at 625 are (max/min):
Solecismic - 17 / 13
40 Time - 4.93 / 5.23
Benchpress - 38 / 25
Agility - 7.52 / 7.94
Broad Jump - 106" / 91"
Position Drill - 0 / 0

[Dandelion]

G
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (low)
Pass Blocking - Benchpress (low), Agility (maximum)
Endurance - Broad Jump (large)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, T, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a G with all his skill set values at 625 are (max/min):
Solecismic - 16 / 11
40 Time - 4.90 / 5.15
Benchpress - 43 / 27
Agility - 7.49 / 7.74
Broad Jump - 109" / 91"
Position Drill - 0 / 0

[Dandelion]

T
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (low)
Pass Blocking - Benchpress (low), Agility (maximum)
Endurance - Broad Jump (large)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, G, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a T with all his skill set values at 625 are (max/min):
Solecismic - 13 / 9
40 Time - 4.98 / 5.22
Benchpress - 43 / 28
Agility - 7.34 / 7.72
Broad Jump - 113" / 96"
Position Drill - 0 / 0

[Dandelion]

P
Skills in skill set - influenced combine value (grade of influence):
Punt Power - 40 Time (large)
Punt Hang Time - Benchpress (maximum)
Directional Punts - Solecismic (maximum)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
K

Important paired skills:
none

The combine values I got for a P with all his skill set values at 625 are (max/min):
Solecismic - 33 / 22
40 Time - 4.64 / 4.96
Benchpress - 18 / 13
Agility - 7.13 / 8.40
Broad Jump - 119" / 93"
Position Drill - 0 / 0

[Dandelion]

K
Skills in skill set - influenced combine value (grade of influence):
Kick Accuracy - Solecismic (maximum)
Kick Power - Benchpress (medium), Broad Jump (maximum)
Kick Distance - 40 Time (maximum)
Kick Hang Time - Benchpress (medium)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
P

Important paired skills:
Kick Power/Hang Time - Benchpress

The combine values I got for a K with all his skill set values at 625 are (max/min):
Solecismic - 31 / 23
40 Time - 4.91 / 5.05
Benchpress - 20 / 12
Agility - 7.14 / 8.40
Broad Jump - 117" / 104"
Position Drill - 0 / 0

[Dandelion]

DE
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Benchpress (medium), Agility (maximum)
Pass Rushing - 40 Time (maximum), Benchpress (low)
Endurance - Broad Jump (maximum)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, G, T, DT

Important paired skills:
Run Defense/Pass Rushing - Benchpress

The combine values I got for a DE with all his skill set values at 625 are (max/min):
Solecismic - 24 / 9
40 Time - 4.58 / 4.81
Benchpress - 38 / 27
Agility - 7.07 / 7.37
Broad Jump - 127" / 104"
Position Drill - 0 / 0

[Dandelion]

DT
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Benchpress (medium), Agility (large)
Pass Rushing - 40 Time (maximum), Benchpress (medium)
Endurance - Broad Jump (maximum)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, G, T, DE

Important paired skills:
Run Defense/Pass Rushing - Benchpress

The combine values I got for a DT with all his skill set values at 625 are (max/min):
Solecismic - 21 / 8
40 Time - 4.85 / 5.04
Benchpress - 41 / 28
Agility - 7.14 / 7.69
Broad Jump - 116" / 101"
Position Drill - 0 / 0

[Dandelion]

ILB
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Benchpress (low), Agility (maximum)
Pass Rushing - 40 Time (maximum), Benchpress (low)
Man-to-man Defense - Broad Jump (large)
Zone Defense - Position Drill (maximum)
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
Interceptions - Position Drill (maximum)

Usefull skill sets for possible position change (with no influence on combine values):
DE, DT, OLB, CB, S

Important paired skills:
Run Defense/Pass Rushing - Benchpress
Zone Defense/Interceptions - Position Drill

The combine values I got for a ILB with all his skill set values at 625 are (max/min):
Solecismic - 35 / 20
40 Time - 4.53 / 4.82
Benchpress - 35 / 21
Agility - 7.09 / 7.51
Broad Jump - 122" / 104"
Position Drill - 49 / 27 *

* without Interception 625 skill: 35 / 27

[Dandelion]

OLB
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Benchpress (low), Agility (maximum)
Pass Rushing - 40 Time (maximum), Benchpress (medium)
Man-to-man Defense - Broad Jump (large)
Zone Defense - Position Drill (maximum)
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
Interceptions - Position Drill (maximum)

Usefull skill sets for possible position change (with no influence on combine values):
DE, DT, ILB, CB, S

Important paired skills:
Run Defense/Pass Rushing - Benchpress
Zone Defense/Interceptions - Position Drill

The combine values I got for a OLB with all his skill set values at 625 are (max/min):
Solecismic - 33 / 15
40 Time - 4.52 / 4.68
Benchpress - 29 / 18
Agility - 6.97 / 7.34
Broad Jump - 125" / 112"
Position Drill - 45 / 27 *

* without Interception 625 skill: 35 / 27

[Dandelion]

CB
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Benchpress (low), Agility (maximum)
Man-to-man Defense - 40 Time (large)
Zone Defense - 40 Time (medium), Position Drill (medium)
Interceptions - Position Drill (medium)
Punt Returns - Broad Jump (large)
Kick Returns - Broad Jump (large)
Endurance - none
Special Teams - none

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
RB, FB, TE, WR, ILB, OLB, S

Important paired skills:
Zone Defense/Interceptions - Position Drill
Punt/Kick Returns - Broad Jump

The combine values I got for a CB with all his skill set values at 625 are (max/min):
Solecismic - 27 / 12
40 Time - 4.30 / 4.49
Benchpress - 17 / 12
Agility - 6.77 / 7.12
Broad Jump - 128" / 102"
Position Drill - 59 / 43

[Dandelion]

S
Skills in skill set - influenced combine value (grade of influence):
Run Defense - Agility (large)
Man-to-man Defense - 40 Time (low), Agility (low)
Zone Defense - 40 Time (low), Position Drill (medium)
Interceptions - Position Drill (medium)
Punt Returns - Broad Jump (large)
Kick Returns - Broad Jump (large)
Endurance - Broad Jump (low)
Special Teams - none

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
RB, FB, TE, WR, ILB, OLB, CB

Important paired skills:
Run/Man-to-man Defense - Agility
Zone Defense/Interceptions - Position Drill
Punt/Kick Returns/Endurance - Broad Jump
Man-to-man/Zone Defense - 40 Time

The combine values I got for a S with all his skill set values at 625 are (max/min):
Solecismic - 38 / 18
40 Time - 4.33 / 4.56
Benchpress - 23 / 15
Agility - 6.86 / 7.20
Broad Jump - 127" / 94"
Position Drill - 59 / 42

[Dandelion]

Another nice tidbit I just noticed:

Height, weight and age are responsible for %-developed

Easy to check: set up a draft class with players of perfect height/weight for their position, all , born in the same year.. and all players from one position end up with the same %-developed..

Every inch or pound either adds or substracts from a base value..
Example: QBs
Height / Weight
6'2" / 217 = 27% developed
6'2" / 218 = 44%
6'2" / 219 = 22%
6'2" / 220 = 39%

Example: RBs
Height / Weight
5'11" / 215 = 40%
5'11" / 216 = 57%
5'11" / 217 = 35%

For a year a player is younger than his draft class, he seems to lose 2% and he gains 2% for a year that he is older..

[Celeval]

Excellent information.

[path12]

Interesting stuff, thanks for posting it. I've got a draft coming up so this may be useful.

[Dandelion]

Update 4-29-2010

- some minor glitches and errors ironed out..
- added ratings for the influence of skills on combine values

[OldSchool]

For the defensive front seven, does "Pass Rushing" mean PR Strength or Technique?

Thanks for posting this, great info.

[Dandelion]

Originally Posted by OldSchool

For the defensive front seven, does "Pass Rushing" mean PR Strength or Technique?

I'm not sure - the helpfile to the Draft file generator only says "Pass Rush Rating"..

A look at the bars of a draft class with only Pass Rushing as maximum Skill seems to give PRT bars higher than PRS for most players. As soon as the same players suddenly learn how to stop the run, too, the bar goes further up and quite often all three bars are equally long..

!ATTENTION - PURE GUESSWORK!
Could be that PRS, as one of those Skills without a column of it's own in the draftfile.csv, is made up by some algorithm using PRT/RDef - or even worse: (some of) the combine values made from PRT/RDef.. if this is true, a high PRS should always be found in combination with high values in 40 Time, Agility and Benchpress..
This could also explain why PRS is considered a skill seldom "masked"..

[Celeval]

Originally Posted by Dandelion

!ATTENTION - PURE GUESSWORK!
Could be that PRS, as one of those Skills without a column of it's own in the draftfile.csv, is made up by some algorithm using PRT/RDef - or even worse: (some of) the combine values made from PRT/RDef.. if this is true, a high PRS should always be found in combination with high values in 40 Time, Agility and Benchpress..
This could also explain why PRS is considered a skill seldom "masked"..

I can say for certain that PRS is its own value, separate from the others.

[jeffrey]

These match up relatively well with my data mining results, for the most part. There are some big differences, though.

Are the results you're reporting after correcting for within-category variance or are they from eyeballing the data?

[Dandelion]

Originally Posted by Celeval

I can say for certain that PRS is its own value, separate from the others.

Anything about a source for PRS then? Or is it a purely randomised value?

Originally Posted by jeffrey

Are the results you're reporting after correcting for within-category variance or are they from eyeballing the data?

A bit of both, I'd say..
I tested all skills seperately with the same basic draftclass - one after the other for each position, with 375, 500 and 625 values set, even when the skill wasn't in that positions skill set - and then all from each position skill set together, but only with 625 set as the value for each skill..

Started a new line of tests today based on 5 runs with the same draftclass made up from the same 50 players for each position. This time I include the values for the complete set of skills of those players. That'll take a while to complete.. change values in draft.csv, make draft.faf, induce in game, extractor run for the players at 3 seconds per player (... for 800 players in the complete set that would be.. 40 minutes or so... rinse, repeat.. getting the data from the extractor's result .csv is the quick part )

Seems I made at least one mistake in the first attempt, because I had intelligence set to 1 for all players on all tests. During the first line of new tests I set Int to 100 and it seems as if that would greatly influence Solecismic.

[jeffrey]

What I mean to say is, when you fix an attribute and look for how it affects the combine value, there will be some variation between samples (either differences in player, other attributes, or just another iteration). When you look for differences between the impacts of attribute levels (i.e. 500 and 695 in your examples, whatever that means. I don't use the draft generator), you want to make some comparisons between the magnitude of the difference and the magnitude of variations just within the same group (i.e. all the 500s). Random variations within these groups are bound to lead to false impressions of differences between them in some cases.

You can do a web search for "significance testing" if you are interested in more information.

[Dandelion]

I'm neither a math student nor a native english speaker, so both my testing and my description of what I did might be a bit amateurish..

I already gave an explanation how I gathered my data. And - except the error I made by excluding Intelligence from the testing, which lowered the Solecismic score greatly - it seems to fit into the ranges that MalcPow had given in his thread here.

As already mentioned, I started a second line of tests with a slightly different approach.
For starters I took only DEs as they have only few skills in their position's skill set.

A draft class of 50 players per position was made. All have the same size, weight, year of birth, birthplace and college.
Then all skills - even those not in their skill set, for a total of 41 skills - were set to 625 in the draftfile.csv. 625 equals a (potential) skill value of 100. Intelligence was set to 100 (maximum possible value) for all of them. And in case it might have an unwanted influence on the skills, the volatility rating was set to the lowest value (1).
Then I used that draft class 5 times, reloading the game between the attempts, so that every time the data would be randomized again. From that came a total number of 240 players listed, as I excluded the combine skippers without any physical combine values.

I used a spreadsheet to get the maximum and minimum values per combine value. And also a medium value - adding up all data for each combine value, dividing it by the number of players - and the median value - through a formula given by OpenOffice, which I use for the spreadsheet.

For those 240 DEs the data came up as follows:

Solecismic: Max 46 / Min 30 / Medium 38 / Median 38
40 Time: Max 4.56 / Min 4.81 / Medium 4.71 / Median 4.71
Benchpress: Max 40 / Min 27 / Medium 33 / Median 33
Agility: Max 6.99 / Min 7.49 / Medium 7.23 / Median 7.24
Broadjump: Max 129 / Min 110 / Medium 120 / Median 120

As Extractor also gathers the Scout's "sight" of the players skills, I also included these in the spreadsheet. That gave me the lowest and highest values the players might reach, according to my scout. These were sorted to find out the range from the highest and lowest "low" values, the highest and lowest "high" values, and the medium/media values..
For Run Defense these would be:

Low bracket: low 6 / high 65 / medium 51 / median 59
High bracket: low 41 / high 100 / medium 86 / median 94

So the worst DE had a 6 to 41 range and the best a 65 to 100 range for his potential Run Defense..

This procedure was then repeated with the skill values set to 500 (equalling 50 Skill) and 375 (equalling 0 Skill) and Intelligence set accordingly to 50 and 1.

The three different results are the baseline for my test.

Next I did one draftrun each for every skill, using a value of 625, again excluding players coming up with 0 in 40 Time, Benchpress etc..
I'm not done with all skills yet, but I can give the result for the Run Defense test (46 players).

Solecismic: Max 21 / Min 9 / Medium 13 / Median 14
40 Time: Max 5.09 / Min 5.15 / Medium 5.14 / Median 5.15*
Benchpress: Max 31 / Min 19 / Medium 26 / Median 27
Agility: Max 7.06 / Min 7.36 / Medium 7.23 / Median 7.23
Broadjump: Max 103 / Min 94 / Medium 98 / Median 98

Then I compare to which of the three baselines this data fits..
Solecismic = Min, Medium and Median are about equal to the 375 Baseline, Max is 3 points higher than the 375 baseline's max reaching the 500 baseline minimum, but is still 5 points lower than the 500 medium/median.. so Solecismic is in the 375 (0 Skill) baselin
40 Time = fits nicely into the 375 baseline values
Benchpress = comes close to the 500 baseline (max is 3 points too low, min 2 points, medium 2 is points lower and median is 1 point lower, but all a 5+ points higher than the 375 baseline)
Agility = all but max fit in nicely into the 625 baseline. And as the maximum value from the 46 players testrun is still better than the minimum value from the 625 baseline, it's a fit..
Broad Jump = 375 baseline all the way..

From this, I draw the conclusion, that two combine values are influence by the Run Defense Skill: Agility, to a high extent, and Benchpress, to a lower extent.

The data I got from the scouts view for these 46 DEs was nearly the same as from the baseline test (as could be expected..)
Low bracket: low 8 / high 65 / medium 56 / median 65
High bracket: low 43 / high 100 / medium 65 / median 100*

* the high values for the median come from the fact that from 46 players 27 reached a 40 Time of 5.15 and 24 had a possible high Run Defense of 100 in the scout's opinion..

So, from my amateurish point of view, the second line of testing should get some solid data..
Of course I'd be happy to get advice on how to do some finetuning to the procedure.. and I'd love to see results from tests that other people have done in the same direction..

[TheMeat]

I like the way you're approaching it Dandelion. I have yet to do any extended draft testing but I'll be following this thread to see your further results, and also jeffrey's view on it since he seems to know a thing or two about statistics

[MIJB#19]

Originally Posted by Dandelion

As Extractor also gathers the Scout's "sight" of the players skills, I also included these in the spreadsheet. That gave me the lowest and highest values the players might reach, according to my scout. These were sorted to find out the range from the highest and lowest "low" values, the highest and lowest "high" values, and the medium/media values..
For Run Defense these would be:

Low bracket: low 6 / high 65 / medium 51 / median 59
High bracket: low 41 / high 100 / medium 86 / median 94

So the worst DE had a 6 to 41 range and the best a 65 to 100 range for his potential Run Defense..

You managed to get me confused here.

Is this saying that a run defense of 375 can range between 6 and 65 as scouted, or between 6 and 41? Ditto for the 675 value, is that between 41- and 100 or between 65 and 100?

[Dandelion]

Originally Posted by MIJB#19

You managed to get me confused here.

Is this saying that a run defense of 375 can range between 6 and 65 as scouted, or between 6 and 41? Ditto for the 675 value, is that between 41- and 100 or between 65 and 100?

Sorry, didn't want that to happen..
The example given was just to show the diversity within the 625 pts (100% for all skills) baseline.
The player(s) my scout saw as worst had a possible Run Defense Skill in the range from 6 to 41.
The player(s) my scout saw as best had a possible Run Defense in the range from 65 to 100.

[Rizon]

Originally Posted by Dandelion

C
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (medium)
Pass Blocking - Benchpress (medium), Agility (maximum)
Endurance - Broad Jump (maximum)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
G, T, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a C with all his skill set values at 625 are (max/min):
Solecismic - 17 / 13
40 Time - 4.93 / 5.23
Benchpress - 38 / 25
Agility - 7.52 / 7.94
Broad Jump - 106" / 91"
Position Drill - 0 / 0

Originally Posted by Dandelion

G
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (low)
Pass Blocking - Benchpress (low), Agility (maximum)
Endurance - Broad Jump (large)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, T, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a G with all his skill set values at 625 are (max/min):
Solecismic - 16 / 11
40 Time - 4.90 / 5.15
Benchpress - 43 / 27
Agility - 7.49 / 7.74
Broad Jump - 109" / 91"
Position Drill - 0 / 0

Originally Posted by Dandelion

T
Skills in skill set - influenced combine value (grade of influence):
Run Blocking - 40 Time (maximum), Benchpress (low)
Pass Blocking - Benchpress (low), Agility (maximum)
Endurance - Broad Jump (large)

Skills from other skill sets with influence on combine values:
none

Usefull skill sets for possible position change (with no influence on combine values):
C, G, DE, DT

Important paired skills:
Run/Pass Blocking - Benchpress

The combine values I got for a T with all his skill set values at 625 are (max/min):
Solecismic - 13 / 9
40 Time - 4.98 / 5.22
Benchpress - 43 / 28
Agility - 7.34 / 7.72
Broad Jump - 113" / 96"
Position Drill - 0 / 0

This might be answered in another thread ... but for these, if the Solecismic score doesn't correlate with any skill set, does this mean that score is useless, or does it have some type of use ... like say a higher score means better development? Can it be ignored altogether?

[RedKingGold]

It's still a strong indicator of the player's intelligence on the left side of the player card.

Outside of a QB's ability to learn formations, I'm not sure what, or even if, a player's intelligence has a meaningful impact on player performance in the game engine.

However, "Gridiron Greta" does rank the most intelligent players regardless of position in one of her 22 weekly ratings. Also, the "secret tournaments" found within the game seem to imply there might be some effect of personality traits, such as intelligence, in the game engine. Lastly, I suspect there might be some value to an OL who score off the charts on Solecismic if it's an oulier per MalcPow's thread.

Despite those unsubstantiated ideas...I usually ignore OL intelligence.

[RedKingGold]

So....yeah. Ignore it.