a steal in the third round ?

I think we've wandered off the point a bit, they can't predict what will happen on Nov 16th, and they don't try because it's impossible, collecting data of annual rainfall and so forth is a lot different than predicting the weather on a particular day a year from now.

The same goes for draft picks, especially when there are different people picking, that pretty much wipes out any relevance of any prior pick having any bearing.

 

I think the problem people are having is that last line is kind of contradictory;

"Not that any of this tells you anything about the current draft/regime or player,
but it does give you reason to pause before premature celebration."

Now what he's saying is that the last 10 years can't tell you about this regime, but at the same time is a reason for pause, so the obvious hole here is how can what another guy did give any clue, good or bad, on what this guy will do?

The drafting this team has done in the past gives no clue on how these guys will draft, Brad's final sentence infers it does, but that is false.

 

Just as a general point, if anyone has an issue with where I took the debate between Cbrad and myself, I offer this:

The happiest people are typically the least concerned with what other people think of them. I'm so happy that even my as.shole is smiling.

 

It may be smiling, but its breath is terrible.

 

Probably. But not shi.t shall be given.

I knew you'd be all over that comment.

 

Yeah.. this is an issue that keeps coming up on these boards that somehow you can't make any inferences about a specific event unless absolutely everything is the same (which it NEVER is) to a comparable event. Even when you're looking at the same person at different points in time things aren't exactly the same.

Statistics works because we have distributions that specify the probability of an event occurring. It doesn't matter if you don't know precisely where within that distribution a person is, you can still specify the probability of events occurring.

So obviously some GM's (whether through luck or skill) will have greater ability in finding "steals" in the 3rd round than other GM's. There's some distribution for that which I don't have for all GM's in history but I at least collected some data to see whether we're looking at 80% probability of success, 50%, 20% or whatnot (just to get an idea of the range).

We don't have to know where a particular GM lies in that distribution in order to specify the odds of success, either for the average GM or for the top 10 percentiles of GM's etc..

This is why looking at a distribution with completely de-identified data so you don't know who's who can give you some indication of the probability of success without telling you anything about precisely where a particular individual lies within that distribution.

Hope that makes sense.

In any case, to say that because you're not looking at the same person you know absolutely nothing is provably false in tons of cases. It's not like the distribution of passer ratings for rookies in the last 10 years tells you NOTHING about the probability of a 2017 rookie having a certain passer rating. Of course it does, and from engineering to medicine to tons of other applications, population data is used to make inferences that have a higher probability of success than just a "we have no clue" approach for specific situations or specific people/patients.

 

I don't find it contradictory. Celebration is a purely emotional thing. We could win the Super Bowl, and some fans will applaud in a reserved manner, maybe hi five their spouse, and some will roam the streets, tip cars, and light police cruisers on fire.

It's a reason for HIM to pause. And why not, He's allowed that right. Not everyone needs to get in line and agree. I dont even need our history of 3rd round picks to give pause. The draft is such a crap shoot you can't even count on the best in the business. I think Stringer or someone had stats where a GM who has a hot streak usually reverts to the mean. Nobody has sustained greatness in draft picks.

And I don't find it all that far fetched. Streaks are weird. They sometimes seem to transcend. Look at the Browns and their draft misery, their QB misery, that spans across different regimes. Hell remember our bad streak of 2nd round picks (some draft picks, some trades) that spanned several front offices.

2002 - No pick. traded to move up in 2001 for Morlon Greenwood.
2003 - Eddie Moore.
2004 - Wade Smith.
2005 - AJ Feeley.
2006 - Daunte Culpepper.
2007 - John Beck (original pick), Samson Satele (trade from Welker)
2008 - Merling (original) Chad Henne (trade from Chambers)
2009 - Pat White (we did good by trading JT for a 2nd, then get White who we killed), Sean Smith
2010 - Brandon Marshall. traded down in the first to get it back, selected Odrick and Misi 1st and 2nd (and passing up a few great players)
2011 - Brandon Marshall. traded up to select ... Daniel Thomas.
2012 - Jonathan Martin.
2013 - Jamar Taylor.

2014 is when the bleeding stopped with Jarvis, and Jordan Philips the jury is still out. Couple of alright players like Wade Smith, Sean Smith (who was better AFTER he left here), Misi but not without costs.

That is a miserable, miserable 2nd round history spanning 5 administrations. And yes, none of the prior admin picks had any bearing on the next pick, but At any point cbrad would have been right to say, hey let's wait just one minute here. When I see it I'll believe it!

 

You did both my friend. You asked a question, and made a statement.

Which is fine. I don't think Cbrad, nor I, am saying you can't think the way you do. He gave his reason for why he won't be joining you. But he didn't say in his OP you were wrong for doing so, nor that you shouldn't think that way.

Some people like giving grades after the draft, and some don't. Both are fine!

 

The part that people seem to miss about statistics is that IT PROVES NOTHING.

For instance, let's say that 62% of all 1st round quarterbacks selected overall turn into franchise players (I just made that up as an example, so don't get hung up on if the math is right. It's not). In other words, 6.2 out of 10 will be great, 3.8 won't. But that tells us nothing about the next 1st round quarterback taken in the draft though- one thing has zero to do with the other. It's still a prediction and by itself, it's not strong enough to be based on fact.

For instance, how many of the 62% started four years in college? How many won a championship? How many threw for 5000+ yards? There could be a thousand factors that made that 62% happen....just grabbing a general stat doesn't make it a true predictor.

Here's what I mean- this is a true stat. Of the last 5 inmates executed in South Carolina, all of them asked for M&M's as part of their last meal. Therefore, it's factual to say that M&M's lead to inmate death. But is it a real predictor? Does it teach us anything at all...or is it just rubbish to prove a point?

 

The way it's done in science is to validate predictions that are made based on one set of data on another. For example.. say your model of QB's turning into franchise players is to simply look at historical rates. Then you should test your model predictions using only 1970's data on the 1980's, or vice versa, or using predictions based on 1970's-1990's data on the 2000's, etc..

Once it's validated on data that is "novel" to the original data you made the predictions from, you have evidence the model works well. So yes you can "prove" things (as in show the model works to some confidence level - that's a technical term in stats that tells you the probability new data will be within the predicted range).

Also.. your M&M example only shows correlation. No scientist or statistician with half a brain (which occasionally it seems like they don't have lol) would just jump to the conclusion that correlation = causation. You need to test for causality if there's uncertainty in that (sometimes the underlying mechanism is not in doubt and you can dispense with such a test).

 

Saying i'll believe it when I see it is a lot different than saying the odds are bad because past regimes have failed, but I have no problem with anyone saying they'll believe it when they see it based on the principal standing.

You're right, disaster can continue from 1 regime to another, however, as you also stated there is still no predictors from 1 regime to another, Eddie Moore was not a predictor for Jamar Taylor etc, so where Brad may have been right if he had predicted it, it wouldn't have been based on reliable knowledge, he just would have been lucky.

 

Yes, but NFL analysts aren't scientists- they're much closer to the M&M example than the unbiased scientific analysis that you're describing. Anyone in an NFL booth is using those models to prove a point, not valid a valid conclusion. Very big difference.

 

Your M&M example only shows that it's a fact that they asked for it as part of their meal. It has nothing whatsoever to do with their death. That is just a nonsensical stattement and is not based in fact on any level other than they asked for it as part of their last meal.

The method by which they were executed would lead to their death. As would the manner of crime committed which landed them on death row; which will lead to their death.

 

Actually the M&M example iof something that is facially unrelated, but on deeper examination there could be some kind of effect, or they both may be be correlated to a causitive factor. The Freakonomic series of books explore these kind of relationships.

Hypothetically there might be something in the brain of psycopaths (and it is well established clinically diagnosed psychopaths have different brain structures to regular people) to have a strong liking for M&Ms or similar multi colored candy. So if psycopaths are more likely to commit capital crimes and M&Ms are the candy of choice for psycopaths there is an indirect relationship between M&Ms and executions. The relationship iwould not be causal but 2 different indicators of the underlying cause.

 

That makes more sense if produced in that way than what he said. As there, there is little to link death and M&M's

 

I know. I was just pointing out that sometimes there are links between things that appear absurd to be linked at first glance.

 

Sigh.....

Everyone has the right to have whatever opinion they want about anything....including opinions about other opinions.

Saying they worried about X pick because of reasons that have literally not one single thing to do with X pick deserves to get questioned, though. It's about the same as saying I'm worried about Adam Gase as a coach because there's been little to no successful HCs named Adam. It is a stupid qualifier.

 

The beauty of all this is, we will all find out in 2-4 years

I only consider it a steal if he plays balls to the wall on our team. If he leaves and becomes an all pro it doesn't count

 

I disagree but again that's okay. If he's a solid starter he's value in the 3rd. If he's good starter he's a very good value maybe even a steal. If he's a probowler he's definitely a steal.

Solid starters to me are steals in the 5/7 rounds. Mayyybbe 4th. But those are just my own definitions in this semantic game

If he's just okay and then becomes an all pro on another team we'll that just sucks but it happens to us

 

That was the point- it wasn't scientific at all and a ridiculous assumption...exactly like a lot of people do with NFL statistics to predict the future.

 

Wow...I just read through this thread and have to comment on a few things. lol


First, I think Tank has the ability to be a really good corner. He may not be the brightest bulb in the box (from what I've read), but I don't think he's as dumb as some here have tried to insinuate.

Second, cbrad, you kind of got piled on, huh? I didn't take cbrad's post as anything negative about Tank. I took it as a "The Dolphins haven't done really well with drafting in the 3rd round over the last decade so I'm going to wait to give my judgment on this years 3rd round pick" statement. Hell, he even went so far as to leave out the last 2 years of 3rd round picks because he doesn't feel it's fair to judge them yet.

Third, I've got to agree with jdang on this one. Hell can freeze over now.... lol

 

The problem with cbrad's assessment is that it has nothing to do with Tank the player. Deej made a thread about Tank the player and cbrad has doubts not because of the player which would be valid, but because of a whole host of stuff that has absolutely nothing to do with the player. Zero. Zilch. Nada. Zip. His point doesn't even really have anything to do with this FO. As I said earlier, it is like questioning if Gase is a good coach because there's a dearth of good HCs named Adam. JD jumping in, is typical JD where he changes the actual argument and defends a poster based solely on the people arguing with the poster. Had you or resnor, for example, taken cbrad's stance instead and I defended you, he'd have argued against you.

 

I get it. However, I don't think cbrad was really talking about Tank specifically.. I took it more as he was stating his concerns on ANY 3rd round draft pick of the Dolphins because of their history.

Anyhow, that's how I took it.

 

That is how he took it and it's specifically the problem we're talking about.

 

No wonder my thang shrunk this morning.

I mean, wait what.