Like all of you, I was bored and just waiting for the lockout to end so I could have my "fix" of football. And then I remembered 2 Fan posts on this site. One was by "joeklecko" and the other was by "Nu York Fan" and they both referred to a posting by "Waldo" which dealt with mathematically evaluating OLB's before the draft. http://www.footballsfuture.com/phpBB2/viewtopic.php?t=439601 . I would like to thank them for planting a seed that grew into a puzzle that chased away the boredom, and will keep my mind occupied for a very long time. But I can not solve this puzzle by myself, since I am a Jets fan and not a football fan. Outside of the Jets I am not competent to make any judgments, so I need your help and guidance. In return for your help, you might find the answer to a few questions that puzzle you. Why did Vernon Golshen fail? What is the formula the Jets use in selecting their Defensive Linemen and others? What should be the expectations for this years top two draft choices? Do the Combine Test really tell us anything useful about a prospect's chance of success in the NFL? Are there any undiscovered diamonds in this year's undrafted free agent class? At the very least, there should be a few lively and entertaining debates.
While numbers will not tell you everything, you can not measure heart, commitment, football instinct, and character with numbers, but they can tell you about the raw physical clay, with which a coach can mold and shape a player. One advantage of going by the numbers is that they are a constant, which can not be changed, they are what they are, not highlights or lowlights which are edited, or varying according to the competition faced. Some tell you that numbers can be deceiving, that only film study will show you how a player plays and the level of competition he faced, but film study will not tell you about the competence of his college coaches, and other subjective factors which also affects his play. Film study is a vitally important second step, before the final step of actually meeting the player and watching him work out. However, these two steps require resources and time which is limited.
Numbers have another advantage, formulas can be found that yield results that were, in the past, significant guides to success in the NFL. Thus, a large number of potential prospects can be processed quickly and in batches, so that those deserving further analysis can be separated from those who don't. This can greatly enhance the ability of teams to find players in small school programs, non-NCAA programs, non-collegians, and Semi-pros, who might otherwise be overlooked: Andy Studebaker (Wheaton College), Pierre Garcon (Mount Union), Joe Greene (North Texas State), Terry Bradshaw (Louisiana Tech), Jerry Rice (Miss. Valley St.), Kurt Warner (Northern Iowa), Jason Taylor (Akron), DeMarcus Ware (Troy), Wayne Chrebet (Hofstra), Otis Sistrunk (None), Charley Powell (None), Eris Swan (None), Vince Papale (None), and Antonio Gates (Didn't play football in college). A quick and simple way to stretch the limited resources of any team.
The numbers are relevant, that is why they have the tests at the combine, and Pro-days for those who were not invited to the combine. But how relevant are these individual tests to predicting success in the NFL. That is the first task I will tackle. But we must remember that each position on the football field has its own set of physical requirements, so each position ideally should be analyzed individually. That will come in the future, for the time being I will analyze individual groupings, Defensive Linemen, Linebackers, etc. I will begin with Defensive Linemen because they were the first draft choices of the Jets, and I was curious to evaluate them.
I will try to be systematic in my study, and take you step by step as it evolved, but I am aware of its shortcomings. Rather then just using High Achievement players in evaluating the Combine tests, I should have also indicated generally accepted above average players, and starters, as well as busts, as "Waldo" did in his original posting. I should have also used many more undrafted prospects, rather than only using the prospects listed in "NFL.com"; these are both areas you can help me with and I invite your participation. I will begin with Defensive Linemen, and the next grouping will be devoted to the Linebackers, the Jets greatest need.
I welcome and invite your participation, this will only be as useful as a team effort makes it. Thanks.
In my Database of Defensive Linemen drafted in the last five years there are 269 players with more or less complete data. These include 32 High Achiever players such as 18 Pro-bowlers (Yellow) AND 14 who were chosen as "top 100 players of 2010" (Light Blue) as selected by NFL Network (incomplete at this time) and/or "Football Outsider." Or they were selected as part of 2009 or 10 All Rookie Team (Dark Blue) by Pro Football Weekly and the Professional Football Writers of America.
Except for the two Defensive Linemen drafted by the Jets, this year's draft class (drafted and undrafted) was not included in this first analysis since I do not know what their level of achievement will be, as I set up my formulas. They will be returned to the database when I evaluate this year's draft class. This reduced my Database of Defensive Linemen to 210 players. When a rookie was selected for the Pro-Bowl, they were only counted in the pro-bowl grouping. There are Pro-Bowl Defensive Linemen of the past five years who are not in the database because they were drafted before 2003 and Combine or their Pro Day data was not available.
These High Achiever players were used in order to help evaluate the effectiveness of the formulas used to evaluate the players. In total, the 32 High Achiever players comprise 15.2% of the database. A random sample in any test where data was available for all High Achiever players, being rigorous, would be 4 High Achievement players in the top 10%, and 8 in the top quartile (25%), being less rigorous.
These numbers will be adjusted according to the number of players and High Achievers for whom data is available in each test. These adjustments will be mentioned in the introduction to each chart. The more important number statistically is the top 10% since it is more rigorous, and less subject to random noise. The Charts show the top 50 players in any test. I also included the Jets (10), just because I'm just curious, and I am a Fan. Jets are Green and High Achiever Jets are Olive. This chart includes in addition to the standard tests at the Combine, a computed constant called Mass which is the players weight divided by his hight, it is the measure of how massive a player is. The 32 High Achiever players are:
Pro-Bowlers and Jets not included in analysis because of incomplete data:
The various tests at the Combine are supposed to measure:
One would think that only for WRs and CBs should the 40 yard dash time even be analyzed, since few players travel 40 yards in one play. But, it is possible that the 40 yard dash measures energy level or endurance over the short period of time a typical play takes. As with all the tests the results will speak for themselves.
It would appear the first 10 yards, known as the 10 yard split, is a far more important number. This shows how quickly a player can get off the line of scrimmage. It doesn't factor in top speed, but it it is thought to show burst and acceleration, universally important measures.
The 20 yard Short Shuttle: The short shuttle asks players to redirect. While the 10 Yard and 40 Yard numbers measure straight-line speed, the short shuttle's job is testing how players will react on the field with a change in direction. It is primarily run to evaluate the quickness and change-of-direction ability of players. Linebackers are consistently forced to read plays and change direction throughout just a single play. A Defensive End may be called on to do the same as he chases down QB's and RB's.
These first three tests test the agility and speed of prospective players in various ways, the next three try to measure the strength of a player. The KEI is the better known name for the Kirwan Explosion Index. According to Pat Kirwan, it's developer, it is the raw number found by adding Bench Press Reps to the Vertical Jump and Broad Jump. A score above 70 is a good number.
The vertical jump is a measure of how high an individual can jump off the ground from a standstill. From bending their knees and exploding straight up, it measures the strength of one's legs.
Like the vertical jump, the broad jump is done from a standing position, but this drill measures how far forward a player can jump. The broad jump is very important, because rather than just measuring the power of the legs, it is using more of your whole lower body. It's a standardized test of lower body power.
ALL STATISTICS REPRESENT THE PHYSICAL POTENTIAL OF THE PLAYERS AT THE TIME THEY WERE DRAFTED. Once they are in the NFL they will undergo conditioning and strength training, as well as being coached on technique. Therefore, a players physical being is subject to change once they are in the NFL. But it all starts with the physical potential, the "Raw Material", as measured before the draft. These are objective measurements, but just as important are the subjective measurements of character, technique, football instinct, mental and emotional make-up which can not be measured in numbers. That is what coaches and scouts are paid to do.
MOST OF THE PLAYERS IN THE DATABASE ARE WITHIN THE FIRST FIVE YEARS OF THEIR CAREERS IN THE NFL AND MAY NOT HAVE YET ACHIEVED THE HONORS THEY MAY ULTIMATELY ACHIEVE. So the formulas, can and will be refined as time goes on. Also, once the Lockout is completed, and Undrafted Free Agents have joined a team, I will improve the database before the next draft season. Any suggestions you may have will be helpful.
All raw player statistics are from nfldraftscout.com. My computed statistics, except for mass, have been normalized so that the top Player is in the upper part of the 99 percentile range. Different constants have been used to bring about this desired effect, however this means that analysis should be done only within columns and not between columns, except to the extent they share factors. In all computations the raw data was used, and not the results of previous computations.
Since I am a Jets Fan and not a Football Fan, I have not felt competent or tried to delineate who was a starting players for their teams, above average players, or widely accepted but not honored High Achievement players. This is something you, the general football community, could help me with. This information would help improve the validity of my analysis. The positions played were those played in college, according to nfldraftscout.com and they may have been changed in the NFL, this information would also be helpful. I hope this becomes a community project and I am open to any comments or criticisms you may have.
I am completely aware of the difficulties presented by drawing conclusions based on small data sets, so I will do so in a conservative matter. . My parameters are if something produces 2.5 or more times the expected result in the top 10%, it will be deemed significant. If it produces between twice the expected result and 2.5 times, it will be considered somewhat significance. 1.5 to 2 times the expected result will be considered only marginally significance. If the top 25% is within one level of significance of the top 10% (significant + somewhat significant) the 10% reading will be considered confirmed. In those cases where there is a reverse divergence between the 10% and 25% groupings the lower level of significance will govern, unless there are more the 50% of High Achiever players in the top 25% or there are special circumstances, which I will take note of.
THE DEFENSIVE LINEMEN IN THIS YEAR'S DRAFT CLASS. It is impossible to know who is possibly a high achievement player, such as being chosen for an all rookie team, without them having been on a team for a year. Therefore these players would serve no useful function in trying to assess formulas which attempt evaluate players for their potential. First the formulas must be evaluated', and only then can the players be evaluated. Therefore for my first set of tests, which evaluates formulas, this year's draft class of defensive lineman, as in the chart below, were not included in the database. Once I have evaluated my formulas I will then evaluate the players, looking for undrafted jewels. The undrafted players are in purple and those who were drafted are brown. Please note, the position played is based on the position played in college.