Spotting a Couple of Forum Football Fallacies

[chipwhich]

The diminishing returns theme is playing over and over.

[IHOPSkins]

.....That's correct, but that fact only makes it more likely that the draft will offer more chances to improve on the defense. It doesn't justify planning on drafting for defense -- which was the point made by my OP. Regardless of the odds, the draft still might offer better chances to improve your offense.

Of course if the Draft Class is weighted for Offense you go with the flow

But last time I checked....there is the same amount of Offensive and Defensive players on every play....and by inference the same amount of Talent on both sides of the ball available during the draft

Given that we have scouted all talent.....if we can draft Equally Talented players for Both O and D....we should choose D....better chance of improvement

The diminishing returns theme is playing over and over.

Exactly

I defer to your concise explanation

:notworthy

[Mr. Nostril]

Let me make sure I understand. Let's say that the efficiency of our elite offense is rated at 35 while our lackluster offense is rated at 20 for a total team rating of 55.

You are saying that there is a limiting factor on growth at the 35 level. If that's right, is it possible for you to give an example of how that might happen? I can't come up with one on my own.

I can't give a practical version, as I don't know enough about the Xs and Os of football, but there's a limit built into this model, as a unit can't have a higher score than 50. So imagine, you have a choice between adding 3 offensive players who would normally have combined value of 16 or 3 defensive players who have the same value. Well, it would only make sense to take the defensive players since adding the offensive players would cap out the model and you wouldn't be able to access the full value of the players chosen.

It stands to reason from this that it works more like a singularity where as you add talent to a unit the value of that unit moves ever and ever closer to 50, but never actually reaches that level. So, the closer you are to the goal of a unit rated 50, the less value you get from adding more talent.

Look at it this way. If there was a defense with no players, it would have a score of zero. If you added 26 NFL defensive players at random (with respect to filling the proper roster spots) you would probably have a defense rated at lest 20, but if you replaced each of those 26 with someone better(with respect to salary cap concerns), you probably wouldn't have a defense rated at least 40. Building from bad to respectable probably requires less talent than from respectable to good, which itself probably requires less talent than building from good to elite. So, it probably makes more sense to add talent to your weaker units, not because some inherent value in having a balanced team, because talent added to the weaker units probably results in greater total value.

Of course, this is only a hypothesis that I'm asserting makes sense to me. In order to determine whether it's true or not someone would have to study what happened when teams tried to bolster already strong units vs when they tried to rebuild weak units. I'm just saying that the model presented doesn't entail that equal assets are equally valuable when you take into account to the context in which they're placed.

[Oldfan]

The diminishing returns theme is playing over and over.

Give me an example of one in this discussion.

[IHOPSkins]

Give me an example of one in this discussion.

Equal Talent Available.....go with D

You changed your argument by saying....If there is more talent in the draft on the offense.....

Well that might not be the case

Could be Equal....and if such.....Go D

ERGO

We should Draft D if

There is more talent available on D....Or Equal talent on D and O

Only IF there is an over abundance of talent on the O side...should we concentrate on O

I have spoken

:allhail:

[Oldfan]

...But last time I checked....there is the same amount of Offensive and Defensive players on every play....and by inference the same amount of Talent on both sides of the ball available during the draft.

That explanation isn't going to work.

Coin flips don't follow a 50/50 pattern: HTHTHTHT and neither will the the draft offer DODODODO. The randomness of chance can offer us better opportunities to improve on offense than on defense.

[IHOPSkins]

...... The randomness of chance can offer us better opportunities to improve on offense than on defense.

Lets say the draft is LOADED with Offensive talent.....but all this talent is picked before your selection at 32 overall....(since we won the Superbowl)....then on your board are 2 players rated the same........heck lets say the offensive player is rated BETTER then the defensive player.....you should still draft the defensive player because the slight advantage given to the offensive player will be MORE then offset by the OVERALL IMPROVEMENT....... shown by the formula below

1 = bad ......5 = great

O Draftee (3.1) - previous O starter (3.0) = 0.1 improvement

D Draftee (3.0) - previous D starter (2.0) = 1.0 improvement

1.0 > 0.1

0.9 overall improvement difference drafting D in this case

Heck I even allowed you that the offensive player replaced is ONLY average (3....on a 5 scale)

Again the point is made--------Its not only about what you get.....but ALSO what you replace

Do the math and draft D :pfft:

[Oldfan]

...Of course, this is only a hypothesis that I'm asserting makes sense to me. In order to determine whether it's true or not someone would have to study what happened when teams tried to bolster already strong units vs when they tried to rebuild weak units. I'm just saying that the model presented doesn't entail that equal assets are equally valuable when you take into account to the context in which they're placed.

Intuitively, you're right. It is probably much easier to move from very bad to mediocre than it would be to move from mediocre to excellence regardless of whether we are talking about offense, defense or the whole team. That is especially true given the NFL rules aimed at producing parity.

But, I think you are missing the same point as some others. Even if you are right, it does not change the point made in my OP:

The truth is that, if you can draft a player who can significantly improve your team, it doesn't matter which side of the ball he plays on.

What it would change is the odds of finding players on the offense v. odds of finding players on the defense who would improve your team when your turn comes up to draft. Even if the odds diminish on finding offensive players, it still remains true that the team should take them when they represent significant improvement to the team.

---------- Post added December-22nd-2012 at 03:35 PM ----------

...Again the point is made--------Its not only about what you get.....but ALSO what you replace

Of course it is.

When I wrote about significant improvement, that factor was already in the calculations. I covered that point earlier, but I don't recall which post.

[Est.1974]

But, I think you are missing the same point as some others. Even if you are right, it does not change the point made in my OP:

The truth is that, if you can draft a player who can significantly improve your team, it doesn't matter which side of the ball he plays on.

The thing is OF, to me that point reads in a slightly different context when you add in the previous line from the OP ;

Our defense is weaker than our offense, therefore we should focus on defense in the draft.

The truth is that, if you can draft a player who can significantly improve your team, it doesn't matter which side of the ball he plays on.

[Oldfan]

The thing is OF' date=' to me that point reads in a slightly different context when you add in the previous line from the OP ;[/quote']The previous line was a fallacy recommending a plan for drafting. Are you saying it didn't read that way to you?

[Mr. Nostril]

Intuitively, you're right. It is probably much easier to move from very bad to mediocre than it would be to move from mediocre to excellence regardless of whether we are talking about offense, defense or the whole team. That is especially true given the NFL rules aimed at producing parity.

But, I think you are missing the same point as some others. Even if you are right, it does not change the point made in my OP:

The truth is that, if you can draft a player who can significantly improve your team, it doesn't matter which side of the ball he plays on.

What it would change is the odds of finding players on the offense v. odds of finding players on the defense who would improve your team when your turn comes up to draft. Even if the odds diminish on finding offensive players, it still remains true that the team should take them when they represent significant improvement to the team.

If you're not, in any practical sense, arguing that a team should focus only on finding the best players rather than filling needs and strengthening weak units, then how is this not a tautology? It's basically a complicated way of saying, "The player that improves the team the most is the best player for the team," which is obviously true. But my counter-argument can't be considered at fault because it assumed you were making a point that has some application in reality.

[Hubbs]

Let me make sure I understand. Let's say that the efficiency of our elite offense is rated at 35 while our lackluster offense is rated at 20 for a total team rating of 55.

You are saying that there is a limiting factor on growth at the 35 level. If that's right, is it possible for you to give an example of how that might happen? I can't come up with one on my own.

The limiting factor is the number of touchdowns that a really good offense already scores. The value of an offensive player isn't ultimately measured by the number of touchdowns he personally scores, it's measured by the total number of points the offense scores with that player on the team that wouldn't have been scored without him. For example, the true value of Calvin Johnson isn't the number of touchdowns that show up on his stat line at the end of the year, it's a combination of the touchdowns he scores that would have otherwise been field goals or no points at all, the field goals he generated by making plays to get within field goal range that otherwise would not have been made, and the total combined points that were ultimately scored by other players specifically because the opposing defense was paying extra attention to Johnson that would not have otherwise been paid to an average replacement player. Obviously it's impossible to know these things with 100% certainty, as we would need a bunch of parallel universes in which we could control for absolutely every other factor except one single player at a time. But I feel pretty confident in saying that Calvin Johnson ultimately provides more true value to Detroit as the team is currently composed than he would to, say, New Orleans as the team is currently composed. If you take Calvin Johnson away from Detroit, that's probably a bottom-10 offense. If you add Calvin Johnson to New Orleans... yes, he puts up a ridiculous stat line. Yes, he increases the total number of points they score in a season. But does he ultimately help them score as many points that they otherwise would not score as he does in Detroit? I say no, because New Orleans already scores so many, and therefore an enormous chunk of his huge stat line is merely production that would still have been generated by other players.

Or, to compare two physical freaks, if I was Sean Payton (without the suspension) and I was given the choice of adding either Calvin Johnson or Sean Taylor to this year's New Orleans team, I would pick Taylor in a heartbeat. You seem to be suggesting that if their talent could be quantified—if their Madden ratings, so to speak, were both 99, and they were both replacing players with ratings of 75—they would essentially represent the same amount of improvement, just on opposite sides of the ball. And you're correct in that they would represent the same amount of improvement at their respective positions. But the New Orleans offense as a whole already scores so frequently that there aren't nearly as many fruitless drives for Calvin Johnson to turn into points with his play than there would be on an average offense. On the flip side, Sean Taylor would have opportunities to prevent the New Orleans defense from giving up points-producing drives damn near every time he would be on the field, because they already give up so many.

[Mr. Nostril]

Also, I think the "you have to take the superior talent regardless of where he plays" argument really only applies to top draft picks, because the talent starts to cluster together the more players who come off the board, as is the nature of the bell curve. So, the talent disparity is no longer as great once we get to later rounds of the draft, and certainly not great enough to warrant ignorance of our strength and weaknesses. We don't have a first round pick and we're probably making the playoffs, so we're not going to be making a pick until late in the second round. How can this argument possibly apply to the Redskins?

[Est.1974]

The previous line was a fallacy recommending a plan for drafting. Are you saying it didn't read that way to you?

I think my interpretation differs because I don't consider that line to be a fallacy. Well, not when applied to this Redskins roster as things stand today.

[Oldfan]

If you're not, in any practical sense, arguing that a team should focus only on finding the best players rather than filling needs and strengthening weak units, then how is this not a tautology?

The argument clears the clutter away from a popular forum fallacy. It isn't a tautology.

It's basically a complicated way of saying, "The player that improves the team the most is the best player for the team," which is obviously true.

That might be obvious to you now that I have cleared the matter up for you, but it hasn't been to the many posters who have argued that "Our defense is weaker than our offense, therefore we should focus on defense in the draft." That statement wasn't obviously false because it has the ring of truth which only disappears when it is given a closer inspection.

My argument can't be considered at fault because it assumed you were making a point that has some application in reality.

Well, what I read is that you had a hunch there might be a counter-argument.

You missed the point as stated in the OP because it didn't sound obviously true to you. Consequently, you were looking for a way to find the error.

---------- Post added December-22nd-2012 at 04:30 PM ----------

I think my interpretation differs because I don't consider that line to be a fallacy. Well' date=' not when applied to this Redskins roster as things stand today.[/quote']Mr. Nostril just faulted me for arguing the obvious. Now you are telling me you don't agree.

I should just back out and let you two argue this.

[Jumbo]

OF, let me check to see if I am can interpret your personal application of your premise correctly in the following example:

Our team is as it is (O doing notably better than d--vague and general, I know, but lets go with it) and we are grading talent (which can include all things you find important in a player) on a scale of 1-10 with 10 being top rated:

We have the chance for our first selection to be either a #8 level TE or a #5 level safety (and we've made no FA changes at either position). I'm thinking points of separation in talent rating is not a "small" difference in this scale nor is it huge. I'd call it notable/meaningful/serious/substantial etc.

In our current situation, you'd go TE.

Now...either way you answer that (and I have no presumptive or invested agenda--this is just interesting discourse to me), let me ask you if you see this as a complication to your premise:

If in FA you've potentially changed the level of the weaker squad but have yet to see it on the field, how does that work in your decision making?

Say in our above example, we have added two FAs who are 7-8 level talent on O, the 8 at RG and the 7 at KR, but have yet to see them play a game in B & G, does that sway your use of your premise in the example? Do you evaluate their likelihoods as helping the O (perhaps notably) and let that determine any influence, or do you leave that in "doesn't count in this decision" land, or what regarding the example given. It seems to me that it also matters how pick the talent difference is of course in our draft example, that's why I picked the difference of "3" that I did, to make it more complicated (if it does).

Hope that all made sense ---have to go but will be back tonight.

[IHOPSkins]

....... Even if the odds diminish on finding offensive players, it still remains true that the team should take them when they represent significant improvement to the team......

And since the significance of improvement is much more likely with the current D then the current O....we should draft D

[Est.1974]

Mr. Nostril just faulted me for arguing the obvious. Now you are telling me you don't agree.

I should just back out and let you two argue this.

Ha, nice one OF......I'm done for the day. Approaching 10pm, checking out time for now. take it easy buddy.

[Oldfan]

The limiting factor is the number of touchdowns that a really good offense already scores. The value of an offensive player isn't ultimately measured by the number of touchdowns he personally scores, it's measured by the total number of points the offense scores with that player on the team that wouldn't have been scored without him....

I agree with your conclusion. However...

Let's suppose that we had two grade A wide-receivers already on the roster. Adding a third wouldn't be a significant improvement. So, if such an opportunity came up in the draft. We would try to trade down instead. Or, we might draft the rookie and trade a veteran allowing us to get younger and better. The bottom line is that this grade A rookie WR is the BPA who has value we can trade or use one way or another.

Now, in the highly unlikely event that our offense is filled with grade A talent, we do the same thing when the opportunity arises because the BPA has value we can use one way or another.

[if we get a shot at another Calvin Johnson, our offense would improve so much that our defense would find it hard to lose games.]

---------- Post added December-22nd-2012 at 05:20 PM ----------

And since the significance of improvement is much more likely with the current D then the current O....we should draft D

No...for the 18 reasons already stated. [slight exaggerations are allowed under forum rules]

---------- Post added December-22nd-2012 at 05:52 PM ----------

OF, let me check to see if I am can interpret your personal application of your premise correctly in the following example...

1) I think you have to project that your FA selections are good ones; we don't have enough money or picks to take backups;

2) You gave me a lousy choice for our first pick, but I have to take the TE unless there's a chance to trade down in the draft; there's too much difference in the grade ( eight v five ) to do otherwise; If I can trade down, I might be able to get the safety in a lower round. If Fred Davis comes back healthy, I'd scheme to use two TEs more often.

3)I look at both FA and the draft the same; I'd take the BPA in FA also; I'm not just looking for gap fillers; Garcon was a good pick.

Did I answer your questions?

[Jumbo]

I think you have to project that your FA selections are good ones; we don't have enough money or picks to take backups;

2) You gave me a lousy choice for our first pick, but I have to take the TE unless there's a chance to trade down in the draft; there's too much difference in the grade ( eight v five ) to do otherwise; If I can trade down, I might be able to get the safety in a lower round. If Fred Davis comes back healthy, I'd scheme to use two TEs more often.

3)I look at both FA and the draft the same; I'd take the BPA in FA also; I'm not just looking for gap fillers; Garcon was a good pick.

Did I answer your questions?

Yes, and I hope by "lousy" you meant "tough" as that was my intention---to gauge how committed you were to the premise and how well you felt it held up in such situations. The answers satisfied, and I pretty much agree with the thinking. While I try to always take big picture (well, in almost everything, by default), I find in about 8 times out of 10, it's hard to beat a mantra of "top talent tops most other considerations."

[Est.1974]

You gave me a lousy choice for our first pick, but I have to take the TE unless there's a chance to trade down in the draft; there's too much difference in the grade ( eight v five ) to do otherwise; If I can trade down, I might be able to get the safety in a lower round. If Fred Davis comes back healthy, I'd scheme to use two TEs more often

Funnily enough, Fred Davis was selected very much under the same BPA remit.

[brandymac27]

OF, I found this link with a little bit of info. You may have already run across it, but I figured I'd post it in case you haven't.

http://www.advancednflstats.com/2011/01/top-offenses-top-defenses.html

[Jumbo]

Funnily enough' date=' Fred Davis was selected very much under the same BPA remit.[/quote']

And TK and I were in a very minority group supporting the move that day on this board.

I was very enthused about it, and posted such, and was very much thought off-base, no little part due to the "thinking" that Cooley was going to be God Forever (stand down Cooley heads---that's not a diss to him).

Majorities of the posters here are often far more NOT BPA, and are far more of the "upgrade that position that needs it most on the unit that needs it most with that pick" school, which is the fundamental trait OF is challenging, I believe.

[Ryko]

Sorry Oldfan, Look at it this way. If every player gets an individual score based on how good they are (a higher number = better valued player) and their individual number contributes to the whole units score, than your argument would be false. If a better offense (like our current one here in DC) has a total score of 35 (to use the same arbitrary number I saw you use in one of your posts) and our defense has a number lets of 30, then yes adding a player with a quality of "5" to either side of the ball would be equal. (the following are guesses based on the redskins current starters, EX, RG3 is the qb)

So lets say that on offense, our starters look like this QB - 8, RB - 6, FB - 2, WR - 4, WR - 3, TE - 2, OT - 3, OT - 2, OG - 2, OG -2, C - 2. (total 35)

And the defense: OLB - 5, OLB - 4, ILB - 7, ILB - 4, DT - 3, DE - 2, DE - 1, CB - 1, CB - 1, FS - 1, SS - 1. (total 30)

So yes, then adding a 5 point increase wouldn't matter which side of the ball it goes on. But lets say you have the same high draft pick choice and the BEST AVAILABLE player that you can select who is left on the board is a "6" to go at any position. You stick him at QB, does that help? No. he wouldnt play right now. If you stick him at OG, then yes you will get a 4 point increase. If you stick him at CB, then you will get a 5 point increase. Which seems better? of course adding the talent to the defense where you are weaker!

Now your argument for BPA, lets say that the best player available is at QB or RB or OLB where the increase would be 0 or maybe even 1. Then the obvious thing to do would be trade down, but in the NFL a trade isnt always possible to get the value you want in return so why would you draft (EX Matt Barkley, with a score of 6, +0 to the offense) in the second round because he was the BPA instead of maybe a CB with a score of 3 that would atleast play and give you some improvements (+2 to the defense)

The NFL is too non-linear to say it doesnt matter where it goes, and that you can trade at any given time for value, or any other argument. BPA available works at positions you dont already have a superstar at. And BPA is a difficult concept because VERY VERY RARELY will the BPA have a rating substantially higher than the second BPA, when the second BPA fills and need and the first doesnt. If you have a QB of 5 and a WR of 1, and the BPA are QB - 6 and WR - 5, who are you going to take? It is never a sure thing that you will be able to trade down to get that 1 point value loss that you think you would be losing. You have to take both concepts (BPA and draft by need) into consideration, as well as the individuals. AND also that some positions (QB!) are more valuable than others (TE) in the NFL today.

[Est.1974]

I was very enthused about it, and posted such, and was very much thought off-base, no little part due to the "thinking" that Cooley was going to be God Forever (stand down Cooley heads---that's not a diss to him).

Majorities of the posters here are often far more NOT BPA, and are far more of the "upgrade that position that needs it most on the unit that needs it most with that pick" school, which is the fundamental trait OF is challenging, I believe.

I think the Davis / Cooley scenario is a great example -

As well as taking the BPA option via the draft, the FO needs to be smart enough to evaluate its existing assets at the same position and be smarter in 'moving on when the times right'.

You take BPA, then move on an asset at the same position before his worth to anyone else has gone. You basically upgrade at the positoin and get something else back ( player / draft pick) to utilise elsewhere. Unless of course you can gain full value buy keeping both on the roster for a prolonged period of time.

Davis had something like 3 receptions as a rookie and I'd argue he wasn't fully utilised through a fair period of his early career here.

I'd suggest that following the BPA approach and making it work to its full value entails more than just picking the BPA in the draft....as explained above.