a steal in the third round ?

This post points out the problem with cbrad's approach. It's also a microcosm of the inherent problem with letting stats do all your thinking.

Whoever else he would have taken is irrelevant based on his argument because his argument has nothing to do with who was taken and only when they were taken.

 

I think the answer to this is right in your post Key, Vontae, he's one of the top 10-15 CBs in the league, regardless of how intelligent he is.

Tank has great coverage ability and ball skills, coverage ability and ball skills you do not see often in the 3rd round, and boundary CB is probably the least cerebral position on the field, especially if they are going to ask him to press.

I don't think they completely trust either CB in press, especially Max because he doesn't have recovery speed, but Tank is a pure press CB, him and X could be a very good duo.

 

Yeah.. I only said "that's not a good list". You keep implying that means every player on that list isn't "good" by the criterion of whether they are a starter or not (or something similar). I never said that. Look.. in that original post I'm responding to the posts above me where there's gushing about a player. I say I hope he works out but that other than OV the list of picks isn't good. The context is provided by the thread. If you can't see that.. fine, it's still proper writing.

Besides, I tried to make that clear to you in post #15 (which you responded to) and #17 which is right above your response in case there was confusion. Not saying you can't read what I wrote differently, but it's not like every post has to specify every context independent of what came before it. So yeah I can see how you can read into what I wrote something I didn't mean, but what I wrote is perfectly fine given the context.

 

You asked "but to correlate those facts with tankersley is pointless no?"

I said.. "In principle yes (it's pointless), in practice no (it isn't)".

I'm actually pretty good at betting.

 

Nope. Adding in that I mean they're a starter has not been stated by anyone other than yourself. That point has no bearing on the delineation of good/not good. Starters are not inherently good players simply due to the possibility of there being no other option on the roster; hence making player x a starter.

If you say only OV is good and the rest of the list is not good, then it's quite easy to understand; the one player was good is, and the rest are not. Hence the delineation; otherwise it would be pointless to delineate it.

No one has asked for each post to specify something different, you created the line of good/not good. It's that simple.

No one has said it is not proper English. That is not the crux of the issue. It is however a delineation that is the crux of the issue. A Venn diagram would provide a simple proof. Where the two overlap if at all would be where both sides of that delineation are true. Your statement only provides for one or the other. As such if they are not on the good side they are on the other as per your statement. It's that simple.

The delineation was created by yourself. It could be that what you said was not what you meant, but that is another matter entirely.

Outside of Langford I agree the list is full of crap. Although I have to say that Jerry continues to play for the Giants for what reason I don't know (but being honest the guy does play). Now if we went so deep as to quantify their respective stats as a means of providing a basis of good production versus bad so be it. In that aspect I would lean on your expertise as I don't have the background in stats as you - and I take it that you are typically fair in your analysis of such data. However, even then the matter of what the scheme would require of each player would have to be considered as all DE's are not asked to do the same thing. In many situations one end is asked to provide contain/ maintaining position while another is tasked to apply pressure etc. but I think going that deep likely is not a good use of time as the reward wouldn't justify the work involved.

My only contention in the end is that Langford is a better player than you originally gave him credit for. The intellectual level of involvement of this debate is stimulating - which I wholeheartedly enjoy.

So please understand that in no way am I saying that your argument is useless, nor are you. These types of debates are good for us and the board - which I think in general the board needs more of.

 

OK.. what you're not understanding is the concept of "good list". You keep thinking that separates players into "good players" vs. "bad players". Not the way it works in general. In general, context is required.

Let me try a different example. Suppose the conversation is about how often a team picks pro-bowlers in the 1st round. You have two teams A and B. Team A has great success with 1st round picks but none are pro-bowlers. Team B has less success on average with 1st round picks but has several pro-bowlers (across the entire drafting history).

OK.. which list of 1st round picks is a "good list" in the context? Meaning, which list suggests that team's drafting strategy (let's just suppose it remains the same over time) leads to a decent percentage of pro-bowlers?

Clearly, team A's list is NOT a "good list". That does NOT mean any or all players on that list are not "good" per se, which is where you keep making a faulty inference. Whether something is a "good list" or not depends on the context, which in this hypothetical case is about pro-bowlers and in our case here is about "steals".

Team B's list is a "good list" even if there are many bad players in it. So just because you label something a "good list" doesn't mean all players are "good" either. And doing the same thing I did (from a logical point of view), if you take away all pro-bowlers from team B's list, then it's not a "good list" anymore.

I hope that clears things up.

 

Possibly but removing the outliers would mean you would have to adjust either list to be equal no? Otherwise the data would have been unfairly altered without the same being done for the other - the data would be skewed as the total numbers would be different and therefore so would the final outcome based on the change.

Also to answer the question it would be list B as it has produced the pro bowlers in question. The other has none. So if we are looking at pro bowlers the answer is apparent. No?

 

If you're not doing any statistical analysis on the data and are just categorizing them you don't have to worry about outliers or sample size etc.. The question is purely whether one data point should belong in one of two possible categories. And if you do statistical analysis you can adjust for sample size differences.

Right.. so just to justify what I wrote in my first post, note that I said that other than OV it's not a good list. If the context is about how often you get steals (and for the moment just agree with me that OV is the only steal), then the list isn't good. Doesn't imply any or all players on that list (other than OV) couldn't be labeled "good" independent of the context.

 

Also the post you quoted most recently was altered as I added
More to it albeit while you had quoted mine. The alteration was made at the same time nt afterwards - I mean that in all honesty and I am thoroughly enjoying this debate.

 

Read my post above. As we are saying something similar and would provide more validation as to my leaning towards an analysis etc.

 

Hey.. I respect your football knowledge. I stay away from player evals for the same reason some might stay away from arguments about stats so respect goes both ways.

 

I hope others are getting something out of this as well. It's a debate without insults - but it does have two strong willed, I'd like to say intelligent individuals having a good ****ing debate.

 

Like I told Finster in post #33, I'm totally open to a different threshold for when you'd call a player a steal in the 3rd round. So if you agree with his viewpoint you'll get no argument from me. Personally I use that word more for guys at OV's level.

 

Could do because Langford wasn't asked to produce in the same format as OV as KL was a 3-4 DE and then locked inside on a 4-3. So the requirements for each with regards to production are different.

 

Yeah, chances are he won't pan out. That's independent of whether there are any players on that list you'd consider decent, good or whatever else you want to call them as long as they don't cross the "it's a steal of a pick" threshold.

Playing the odds isn't pointless. Whether we're talking about how the Dolphins have drafted in the 3rd round, how other teams or GM's have drafted, etc.. odds aren't great for getting a player of OV's caliber (or if you want to also add Langford in fine.. doesn't change things much).

Look.. all I'm saying with who knows how many posts is that I'm going to wait and see what happens to Tankersley before joining you in celebration about picking him.

 

Odds game applies to any situation where you have similar enough circumstances, in this case trying to predict NFL success from college performance.

That last sentence of yours is cool though dj. If you have a list of predictions (must be complete so it's not biased) you've made about NFL success from college then we can see how often you're correct. And yes I'd definitely use that data instead of population stats. Question is.. do we have unbiased (meaning including all the failures) data on how often your evals are correct?

 

DJ's point, which is correct, is that it doesn't matter who is on the list, that has no bearing on Tank one way or the other.

It's like trying to predict the weather a year in advance, because it has rained 6 times in the last 10 years on Nov 16th, doesn't mean there's a 60% chance of rain on the next Nov 16th, each Nov 16th, or pick, is unique unto itself.

 

Weather models do precisely what you're suggesting we shouldn't: use past data to predict the future even though the particular circumstances for the weather you're trying to predict are unique.

That's true whether you're trying to predict the weather for the next day, 5 days from now, or use seasonal averages. In every case the circumstances on the day of prediction will be unique, yet we use "population" stats as the basis for prediction. It even goes further than that: they use an ensemble of models and use the average of that ensemble. In other words it's not just "population data" = "past historical data" but a population of models they average among!!

So while in principle what happened in the past has no bearing on the present (if seen in a non-deterministic sense) in practice it does due to statistics.

 

Totally agree with all that.

What justifies the odds argument I'm making is that the best you find on this board are probably not consistently better than the average GM either, meaning we can look at what pro's do and say you're probably not a huge outlier until proven otherwise.

 

I agree...in some situations he will shine. I'm not against the pick at all, I just wouldn't call it a steal since there's a lot of limitations there. If you compare him to someone like Grimes...who had limited abilities but a knack for reading the QB and flushing out plays....they're two completely different types of corners. Grimes was extremely well-rounded, while Vontae and Tank are one trick ponies.

Again, there's nothing wrong with that- Tankersley could thrive in Miami and he'll be essential for shutting down big play receivers. But for run support, trick plays, crossing routes, etc. he's probably going to be a non-factor. That's why teams paused on him.

 

How did this thread go from a nice decent conversatiom and analysis about our rookie CB Cordrea Tankersly into yet another masturbatory argumentative debate over 3rd round vs 4th rounds steals?

Yes, the thread topic mentions 3rd round steal, but nothing in DJ's original post starts the conversation into historical draft comparisons.

I swear every thread here turns into a divisive debate that is rarely ever related to the actual original post. Way too many people here seem to turn every thread this way.

<sarcasm>If only the internet had more fighting. </sarcasm>

 

Past drafts have no bearing on this year's drafts. You can't look at Ireland's failures and apply that to Tannenbaum and Gase.

 

Dude, all cbrad said was he wasn't into premature celebration. Jeez, you guys are able to get worked up over the smallest things. Tankersly hasn't played a snap yet. Could be the next Revis. Maybe never plays a snap in his career. He even said the history of previous picks has no bearing on Tankersly, just that he isn't prematurely calling it a steal.

It's just repeating what he said in his original post.

It's like a team that goes 2-14 for four years, changes coaches and FO, and many players, and you jump on a guy for not declaring it a championship team. Hey, that's his prerogative.

I will throw my hat into the not declaring someone a steal before they ever play a down ring. When we traded up for Dion Jordan it was considered by some a steal due to the fact we gave up so few resources. If Tankersly is a solid player after a couple years, he was good value in the 3rd. If hes a probowler he's a steal. Like Reshad Jones.

 

Jee, I wonder who said that in their very first post in the thread? Hmm, I'll give you a hit, it rhymes with tee-sad.

It's like you guys went all red before you got to the last line his his first post

 

Why is it ok on this site for people to question something, but in the same post say they aren't questioning it, and somehow that makes sense to people?

 

That's not what happened.