Teams ranked by number of player arrests

[Largent80]

Wait. Hawk players arreseted?...It's a conspiracy, and Pete says they will all play Sunday...... :lol:

 

[Jville]

I think that presentation is just a space filler for a publication. It lacks insight. As has been alluded to, arrests seem to come in bunches. Documenting and presenting clusters of arrest might actually provide some enlightenment.

 

[Popeyejones]

By the way, when I say conviction, IMO, the statistical analysis would have to include anyone who's conviction got set aside due to plea deals that would put them in probationary status or diversion type programs. I'd also like the list to include alongside the conviction, exactly what the circumstance was in which they were arrested. This would give us a more realistic picture as to what types of players are on current rosters and what NFL organizations are willing to forgive and look past.

IIRC convictions are in the dataset and you could run them if you wanted.

As for the "circumstances" to me that sounds like slipping into the inherently subjective type of coding that can make formal analysis beside the point.

It also sounds like you're interested in a different question than these data are tackling.

I suppose you could get at it from available data, but it would be really big project, a pain in the @ss to code, and you probably wouldn't have enough data to get beyond the autocorrelation problems if you're just interested in CURRENT rosters and teams (players nested in teams).

My suspicion is that folks who have run this aren't separating the series into two discrete time periods because they're lazy, but because the data aren't robust enough to treat them in a more nuanced way. (I'm just guessing about that though).

 

[Popeyejones]

Edit: also worth saying that if it WAS the coach who mattered, we'd expect a renowned player's coach (e.g. Pete Carroll) to have more arrests on his team than a strong, father figure disciplinarian type (e.g. Mike Singletary). We could investigate that question if we could get players to rank all the NFL coaches in terms of their role as disciplinarians, but it might be a waste of time given that we already have a reasonable suspicion that it's not the coach that matters.

I realize this isn't your main point but don't confuse players coach for being someone that allows his players to walk all over him.

Oh, I'm not. I'm just spit-balling how you could actually go after the problem of trying to measure a coaching effect.

I think a " disciplinarian" hypothesis is probably a pretty reasonable one, and it's actually something that you could survey for.

So, I'm not suggesting that "players coach" == "walk all over him", but IMO that's one of those fine-grained distinctions that sends us down the rabbit hole of increasingly subjective classifications that ultimately make the whole endeavor pointless. And likewise, a disciplinarian scale would have some people at the far end who DO let their players do whatever they want*, and people more towards the middle of the distribution who aren't strong disciplinarians but who also don't let their players walk all over them. Wouldn't it be interesting if we found more arrests at BOTH ENDS of the disciplinarian distribution?

In any case though, mostly moot IMO, beecause the data we DO have gives the impression that any possible coaching effect might be small, if present at all.


*Am I poking the bear too much by thinking of just letting "Marsawn be Marshawn" right now? If so, that can't be celebrated out of one side of someone's mouth while also pulled back on out of the other side. Not accusing you of doing this, but a spade is a spade, and I think Pete Carroll is the subtext of a fair number of the posts in this thread, and reasonably so and for obvious reasons.

 

[kidhawk]

By the way, when I say conviction, IMO, the statistical analysis would have to include anyone who's conviction got set aside due to plea deals that would put them in probationary status or diversion type programs. I'd also like the list to include alongside the conviction, exactly what the circumstance was in which they were arrested. This would give us a more realistic picture as to what types of players are on current rosters and what NFL organizations are willing to forgive and look past.

IIRC convictions are in the dataset and you could run them if you wanted.

As for the "circumstances" to me that sounds like slipping into the inherently subjective type of coding that can make formal analysis beside the point.

It also sounds like you're interested in a different question than these data are tackling.

I suppose you could get at it from available data, but it would be really big project, a pain in the @ss to code, and you probably wouldn't have enough data to get beyond the autocorrelation problems if you're just interested in CURRENT rosters and teams (players nested in teams).

My suspicion is that folks who have run this aren't separating the series into two discrete time periods because they're lazy, but because the data aren't robust enough to treat them in a more nuanced way. (I'm just guessing about that though).

I'm sure it could be done, and if the time was taken to make this original list, it wouldn't be much more difficult to come up with a list of players currently on rosters and their records.

I just think that a list of arrests by teams is only half the story. It doesn't say if they were actually convicted of anything or if they actually did anything. It also doesn't show if the team gets rid of the player. It just seems like a waste of time to make that list when there are more substantive lists they could have made.

I'd be willing to bet that at l east a third of the players (and I'm thinking substantially higher) aren't still on the team they were on when they were arrested, and many may not even be in the league any longer.

 

[mikeak]

I think a better and more telling stat about an organization, is not the number of arrests, but the number of players with convictions on their current roster and what those convictions are for.

This is what I am trying to say. Number arrests is basically meaningless what type of crime commited is better. The the most meaningful is who is on your roster now with either. For instance anything before 2010 is meaningless for Seattle same goes for San Francisco and 2011.

This completely ignores how the owner impacts both the hiring of a GM, Coach and how the owner communicates with players, how the owner allows / doesn't allow players with known issues in their background to be drafted and finally how the owner handles players that gets arrested / convicted.

Example. Supposedly Harbaugh wanted Rice off the Baltimore team. Supposedly the owner shut it down.....

 

[Popeyejones]

I just think that a list of arrests by teams is only half the story.

FWIW I don't think it's half the story. I think it's a DIFFERENT story (i.e. it's a different question entirely).

The question being asked and answered is if there's some stability in arrest rates across teams across time (e.g. across different regimes and short term spikes and dips for tems). It turns out there is. IMO it's interesting because it's somewhat counter-intuitive.

 

[Popeyejones]

This completely ignores how the owner impacts both the hiring of a GM, Coach and how the owner communicates with players, how the owner allows / doesn't allow players with known issues in their background to be drafted and finally how the owner handles players that gets arrested / convicted.

Example. Supposedly Harbaugh wanted Rice off the Baltimore team. Supposedly the owner shut it down.....

Good point. You'd also have to have some insight into the internal organizational structure of teams over time and with individual decisions. You could get closer to it by looking at GMs with GM authority, but there's still a ton of problems.

Twp years ago were the Hawks' drafting decisions the decisions of the GM Schneider? The decisions of the coach Carroll? The decisions of McCloughlan as the senior personnel executive? It's likely all three, but we don't know who was making the case for each pick, if they're were pick swapping, etc., etc.

Just to use the 9ers (the team I know the most) as an example, Baalke is in charge of things but it's not an iron fist: picks are sometimes left to position coaches, or made because someone feels very strongly (e.g. WR coach John Morton really, really wanted Ronald Johnson when he was available in the 6th in 2011. Harbaugh wanted ADB, but it was late in the draft so Baalke went with Morton's wants as the position coach, which everyone was okay with at the time -- if we're modeling a dataset, WHOSE pick does that count as though?).